NordicCNC's build log

[NordicCnc]

Change log
10.3.2020 - Z-axis ratio changed to 2:1. Removed 48V power supply from component list. Added Delta 400W servo motors to component list.
4.3.2020 - Ballscrew dimensions changed.
4.3.2020 - Motor and power supply alternatives added.


Hello,

I’ve decided to start a build log in order to collect all the advice I receive in one single place. I found myself having to go back to all my posted questions, to refresh my memory quite often.
The idea is to build a good CNC milling machine for myself that I will use for my own projects. I plan to take a modular design approach when building this machine. What this means I can explain in another post upon request! Note that this just an approach that I will use for fun and it will not be a deal breaker if it turns out to be too complicated or doesn’t succeed for other reasons.

Machine requirements
The machine will be designed based on following requirements:
- The budget is undefined but I prefer to stay below 8000€.
- Stiff enough to machine aluminium with good tolerances and surface finish. What I mean with good finish is that the finished product should not have a terrible surface. If I can achieve the same surface finish as in this video, I would be more than happy: https://www.youtube.com/watch?v=xyVr35pgx7s
- Fast feed rates for machining large wooden sheets.
- Auto tool change spindle is preferred. Possibility to downgrade to a manual tool change spindle if the budget requires it.
- The footprint is not critical but is preferred to be as small as possible for the following working area: Y-axis (base frame): 1400mm, X-axis (gantry): 1000mm, Z-axis (spindle axis): 200mm.

Machine control and software
- I already own a Mach3 license and will continue using it. In the future I can upgrade to Mach4 if I find it necessary.
- Interface cards Ethernet SmoothStepper (ESS) https://www.cncroom.com/interface-cards/ess-mb3. This I already own. Breakout board will be MB3.
- MB3 breakout board: https://www.cncroom.com/interface-cards/ess-mb3.
- The MB3 and ESS will be powered with a 24V, 150W & 6.3A power supply: https://www.sorotec.de/shop/Power-supply-24V-150Watt-6-3A-Heng-Fu.html
- Delta servo motors (400W, 3000rpm) with brake (for Z-axis): https://www.aliexpress.com/item/32671356577.html?spm=2114.12010612.8148356.2.23d15 2a8ZJcwMZ. - Was JMC 180W servo motor.
- Delta servo motors (400W, 3000rpm) without brake (for X- and Y-axis): https://www.aliexpress.com/item/32671356516.html?spm. - Was JMC 180W servo motor.
- 3kW Auto tool change spindle with BT30 tool interface combined with a 3.7kW VFD: https://www.ebay.com/itm/3KW-220V-ATC-Automatic-Tool-Change-Water-Cooled-Spindle-BT30-3-7KW-VFD-Inverter/264448370955

Z-axis design & components
The solution I will be using to allow for the heavy ATC spindle is to keep the Z-axis weight as low as possible. Linear rails, ballscrew & bearings, and servo motor will all be mounted on the rear plate. The ballscrew nut housing and guide carriages will be mounted on the front plate. This is based on the advice that there is not much difference between front/rear mounting when it comes to stiffness. Yes, it is slightly stiffer with front mounted linear rails, but this is a compromise I need to take. The gantry clearance will be 200mm. Anyway I will still be able to surface components up to a height of 240mm and with a diameter of around 160mm, since it will be possible to raise the tool above the bottom of the gantry. I could also decide to make the rear plate longer to further increase the height if I want.

Components that I plan to use:
- Aluminium tooling plates
- HIWIN linear rails: 20mm
- Delta 400W servo motor, with brake - Was JMC 180W servo motor.
- Ballscrew: 1605 - Was 1605 or 2005.
- HTD 5M pulleys with a 2:1 ratio Was 3:1
- 3kW Auto tool change spindle

X-axis design & components
The gantry will L-shaped and be made with 120x80 or 160x80 aluminium profiles. Aluminium keys will be inserted into the t-slots and then the profiles are bolted together from the bottom. Side plates will be used to mount the profiles together from the side. I am still debating if I will go for high gantry sides combined with aluminium profiles similar to JAZZCNC’s design, or with raised gantry. Linear rails with be top/front mounted to get maximum gantry clearance and maximum spacing between the guide carriages. It will also allow for Z-axis vertical tram. The ballscrew will be placed on the rear side of the gantry to stay clear of chips.

Components that I plan to use:
- Aluminium tooling plates
- Item equivalent aluminium profiles
- HIWIN linear rails: 20mm
- Delta 400W servo motor, without brake - Was JMC 180W servo motor.
- Ballscrew: 2010 - Was 1610 or 2010.
- HTD 5M pulleys with a 2:1 ratio

Y-axis design & components
I’ve learned that it is recommended to make a welded steel frame and use epoxy levelling to get flat surfaces for the linear rails. It is also cheaper. I consider myself a good welder and I have the possibility to make a welded steel frame. Does the welded frame have to be stress relieved or is that not required? I am not sure where I would find any local company that could do that for me. If I go for the welded steel frame, what are the recommended box section dimensions?

The machine will have be moved to my own garage at some point in the future (2-3 years from now), since I am building it in my parents garage. This is the main reason why I am still debating on using aluminium profiles for the base frame. With aluminium profiles I would be able to disassemble it and assemble it again once I move to my own garage. If I decide to use aluminium profiles I will probably use 120x80 or 160x80 aluminium profiles.

To allow for easier alignment of the linear rails I am planning to mount them to the top of the base frame. The guide carriages will be mounted together with separate aluminium plates, which will allow for some adjustment of the X-axis gantry to get the machine square.

Components that I plan to use:
- Aluminium tooling plates
- Item equivalent aluminium profiles
- HIWIN linear rails: 20mm
- Double Delta 400W servo motor, without brake - Was double JMC 180W servo motor.
- Double ballscrew: 2010 - Was 1610 or 2010.
- HTD 5M pulleys with a 2:1 ratio

Continuing forward
I will start posting updates of the design as it progresses. Any advice is very welcome.

[jarjar]

Hi,

Some comments in no particular order, assuming you want to build a decent machine and not a toy:

- 1400 x 1000 x 200 mm is a _huge_ working area. I would expect a machine that size to weighs at least 1500 kg.

About servos:
- those 180W JMC servos are toys and seriously underpowered if you want decent speeds. If you want to go servos, use some proper ones. Delta servos are cheap and good. You need to size the motors depending on the max feed/rapid rates and accelerations you seek. 400W is probably the bare minimum.
- Personally I expect a router this size to be fast, I would target 30m/min (1000 ipm) and maybe 0.5g accel.

About ballscrews:
- 10mm pitch ballscrew would be the minimum I'd use, without any reduction. With servos direct-drive is the best and the cheapest. Good pulleys are expensive, and add unwanted inertia to the system. Servos are usually rated at 3000rpm, with 10mm pitch this gives you a max speed of 30m/min.
- 5mm pitch ballscrew is acceptable for the Z axis. Again without reduction. If you really want pulleys for space issues, use 1:1.
- Ballscrew diameter: you need to calculate it depending on the desired max speed. A ballscrew 1400mm long that don't whip at 3000rpm with fixed-floating ends is around 60mm diameter... If this sounds too big you can go with rotating ballnut (complex) or larger pitch. 20mm pitch would be ideal and you would only need 32mm ballscrews.

About rails:
- don't bother with anything less than 25mm. For the simple reason that 20 rails use M5 bolts and require tighter drilling positioning tolerances.

[joe.ninety]

Hi,

- 1400 x 1000 x 200 mm is a _huge_ footprint. I would expect a machine that size to weighs at least 1500 kg.



Those measurements are all in MM you know? Do you consider that to be huge? That just seems a little bigger than AndyUKs and a couple of others I've seen on here.

EDIT: Are you saying it's huge for a MILL?

The reason I'm asking is because I was looking to build a machine that had a cutting area that could accommodate 1.2m X 0.8m and the proportions in Nordic's build are about the same as I was calculating.

Cheers

[jarjar]

It all depends on your requirements.
But if you're after speed, good tolerances and surface finish in aluminium, it is, indeed, huge.

[NordicCnc]

Hi,

Hi and thanks for your comments, very much appreciated! I will add my replies below.


Some comments in no particular order, assuming you want to build a decent machine and not a toy:

- 1400 x 1000 x 200 mm is a _huge_ working area. I would expect a machine that size to weighs at least 1500 kg.

Yes it is a huge working area! About the weight, I've quickly added all components to my build with assigned material weights etc. and it sums up to about 200kg right now. I excpect it to be around the 300kg mark once everything is mounted (with aluminium profiles as base frame). Not much is missing so I am not sure if you thought that the whole machine would be built out of steel? Then the 1500kg would make sense.


About servos:
- those 180W JMC servos are toys and seriously underpowered if you want decent speeds. If you want to go servos, use some proper ones. Delta servos are cheap and good. You need to size the motors depending on the max feed/rapid rates and accelerations you seek. 400W is probably the bare minimum.
- Personally I expect a router this size to be fast, I would target 30m/min (1000 ipm) and maybe 0.5g accel.

Yep, I realized this later (NEVER BUY PARTS BEFORE DESIGN IS READY) that I would've been better of getting either 400W JMC servo motors or 400W delta servo motors (seems to be pretty much the same price and specs!) or even stepper motors.

Targeted speed is 15m/min and about 0.25-0.5G acceleration. This I plan to achieve with a 2:1 ratio with 3000rpm and 10mm pitched ballscrew. I will also achieve some increased torque with the 2:1 ratio, which will be helpful for these small beasts of servos.

Here are some references where some people are testing these specific servo motors. According to description section he is running at 12m/min at 0.25G at a 2:1 ratio with a 10mm pitch.
https://www.youtube.com/watch?v=xyVr35pgx7s
https://www.youtube.com/watch?v=nd19IBllD2I&t=32s

Here is another one with some amusing testing. Same video in both links but at different time stamps.
https://youtu.be/ZMxzRn9GrKY?t=330
https://youtu.be/ZMxzRn9GrKY?t=420


About ballscrews:
- 10mm pitch ballscrew would be the minimum I'd use, without any reduction. With servos direct-drive is the best and the cheapest. Good pulleys are expensive, and add unwanted inertia to the system. Servos are usually rated at 3000rpm, with 10mm pitch this gives you a max speed of 30m/min.
- 5mm pitch ballscrew is acceptable for the Z axis. Again without reduction. If you really want pulleys for space issues, use 1:1.
- Ballscrew diameter: you need to calculate it depending on the desired max speed. A ballscrew 1400mm long that don't whip at 3000rpm with fixed-floating ends is around 60mm diameter... If this sounds too big you can go with rotating ballnut (complex) or larger pitch. 20mm pitch would be ideal and you would only need 32mm ballscrews.

I wanted to use belt drive to be able to experiment with ratios, mounting simplicity and to hide the motors so that they are not protruding. Using direct drive with 3000rpm and 10mm would be too much as you said for 16mm ball screws. This is another reason why I intend to use a 2:1 ratio, which would result in 1500rpm.

What calculator are you using to calculate those diameters (link or fomula)? I have not seen anyone in here that size of ball screws! I keep seeing the recommended sizes being 1610 and 2010. I will be happy with 15m/min anyway.


About rails:
- don't bother with anything less than 25mm. For the simple reason that 20 rails use M5 bolts and require tighter drilling positioning tolerances.

If I go for the aluminium profiles on the Y-axis (base frame), then I will only have to drill and tap the M5 bolts onto the rear Z-axis plate. This plate size fits well in my shitty CNC router that I built 10 years ago. It can handle the requires tolerances needed for those holes, I am pretty sure.

Anyway, do you think 25mm would be to prefer over 20mm rails for the sake of stiffness?

[AndyUK]

Some comments in no particular order, assuming you want to build a decent machine and not a toy:


Well that really depends on the definition of decent and a toy. Arn't all home machines practically toys when compared to industrial VMCs? Your line seems very arbitrary. Nordic has stated he wants a 'good' surface finish in Aluminium, which really he should be using a mill style machine for - so I see where you're coming from there, but again, define 'good'.



- 1400 x 1000 x 200 mm is a _huge_ working area. I would expect a machine that size to weighs at least 1500 kg.


... Why? This size is about standard for most of the build logs. Huge for a mill, still quite a small router. 1500kg? I think mines about 500kg... Sure added weight helps dampen vibration and makes sense, but there isn't any reason it must weigh that much. And hey - if its on rollers who cares? Lift and move with jacks.



About rails:
- don't bother with anything less than 25mm. For the simple reason that 20 rails use M5 bolts and require tighter drilling positioning tolerances.

That seems like a terrible reason to spend more money. A standard clearance hole for M5 gives you 0.5mm of wiggle. By marking out carefully, setting up the rails with clamps and a dial guage, then using a transfer punch, this is more than enough.

[NordicCnc]

Well that really depends on the definition of decent and a toy. Arn't all home machines practically toys when compared to industrial VMCs? Your line seems very arbitrary. Nordic has stated he wants a 'good' surface finish in Aluminium, which really he should be using a mill style machine for - so I see where you're coming from there, but again, define 'good'.

Hi Andy! Thanks for pointing this out. I will update the starting post with what my definition of good means. With good I mean that after my finishing pass, I want a smooth surface on which you should not/rarely be able to see vibration marks if you don't look extremely closely. The tolerances that you can achieve with a machine also very much depends on the machine operator. You cannot expect to get a good tolerance with roughing passes, even with industrial machines without using some tool compensation or adjusting the G-codes.

[jarjar]

Nordic has stated he wants a 'good' surface finish in Aluminium, which really he should be using a mill style machine for - so I see where you're coming from there, but again, define 'good'.
A "mill style" machine doesn't mean anything. The style of structure (C-frame, fixed gantry, moving gantry, ...) of a machine has nothing to do with its capabilities. Some are just easier to make stiff than others.




... Why? This size is about standard for most of the build logs. Huge for a mill, still quite a small router. 1500kg? I think mines about 500kg... Sure added weight helps dampen vibration and makes sense, but there isn't any reason it must weigh that much. And hey - if its on rollers who cares? Lift and move with jacks.

Just some number I have in mind when someone tells me he wants to build a 8000€ router this size to do aluminium work. A Datron with similar working area is 2.5t for comparison. Stiffness comes with weight when using standard material (metal).




That seems like a terrible reason to spend more money. A standard clearance hole for M5 gives you 0.5mm of wiggle. By marking out carefully, setting up the rails with clamps and a dial guage, then using a transfer punch, this is more than enough.
Indeed I was wrong about that. The clearance is the same for M6 bolts. Still the price difference is low and bigger is always better for rails... It helps for stiffness, ballscrew clearance, and the rails are better supported when mounted on T-slots profiles.

@NordicCNC
You still need a 32mm ballscrew at 1500rpm and 10mm pitch for 1400mm travel. Calculator for critical speed (http://www.nookindustries.com/EngineeringCalculator/MetricCriticalSpeed).

[NordicCnc]

Just some number I have in mind when someone tells me he wants to build a 8000€ router this size to do aluminium work. A Datron with similar working area is 2.5t for comparison. Stiffness comes with weight when using standard material (metal).

I've added some clarification to what I mean with good finish. I am not looking to be competitive with a Datron machine, luckily! Maybe an even better description of what I mean by good finish is that the finished product should not have a terrible surface.The surface finish achieved in this video, I would be more than happy with: https://www.youtube.com/watch?v=xyVr35pgx7s



Indeed I was wrong about that. The clearance is the same for M6 bolts. Still the price difference is low and bigger is always better for rails... It helps for stiffness, ballscrew clearance, and the rails are better supported when mounted on T-slots profiles.

Yes, price difference seems to be only about 15-20% for either 25mm or 30mm rails. If it is bringing me added value and it isn't overkill, I think I am ready to go for the 30mm rails.



@NordicCNC
You still need a 32mm ballscrew at 1500rpm and 10mm pitch for 1400mm travel. Calculator for critical speed (http://www.nookindustries.com/EngineeringCalculator/MetricCriticalSpeed).

Many thanks for the link! Seems like if I would use fixed bearings in both ends, I would barely make it!

[jarjar]

1. You need to add a safety factor.
2. The root diameter of a 16mm ballscrew is not 16, but 14-15 (outer diameter minus ball grooves).
3. Don't use fixed-fixed unless you know exactly what you're doing. The ballscrew needs to be pre-tensioned properly to account for thermal expansion or you will introduce backlash.

[NordicCnc]

1. You need to add a safety factor.
2. The root diameter of a 16mm ballscrew is not 16, but 14-15 (outer diameter minus ball grooves).
3. Don't use fixed-fixed unless you know exactly what you're doing. The ballscrew needs to be pre-tensioned properly to account for thermal expansion or you will introduce backlash.

Good points. A 2:1 ratio resulting in the ballscrew rotating with 1500rpm, I will have to go for either 2010 or 2510 ballscrews, and decrease the ballscrew length to about 1100-1200mm.. This will give me a safety factor of around 1.2 with fixed-floating bearings.

Seems ridiculous to have to use any larger diameter ballscrews. I have not seen anyone else using that here on the forum. How does everyone else achieve 15-20m/min?

[AndyUK]

Yes, price difference seems to be only about 15-20% for either 25mm or 30mm rails. If it is bringing me added value and it isn't overkill, I think I am ready to go for the 30mm rails.

Don't forget to include the carriages in the cost estimate.

You'd be just about okay with a 1200mm 2010 screw @ 1500rpm.

Those of us using steppers rather than servos don't tend to go above 10m/min because the steppers will really lose torque above 1k rpm. My machine can just about handle 15m/min, but I'm leaving it at 10.

You'll have to find build logs for comparable routers using servos - I don't have any experience with them so won't advise.

[NordicCnc]

Don't forget to include the carriages in the cost estimate.

I quickly compared the 1100mm HGR20 and HGR30 sets offered from BST Motion on Aliexpress (I have seen them recommended a lot here, Fred was the seller) and the prices came down to 210€ vs 265€, if I recall the numbers correctly. Both sets included 2 rails and 4 carriages. Seems like a very good deal to me. Even with VAT added is only like 260-320€.


You'd be just about okay with a 1200mm 2010 screw @ 1500rpm.

Those of us using steppers rather than servos don't tend to go above 10m/min because the steppers will really lose torque above 1k rpm. My machine can just about handle 15m/min, but I'm leaving it at 10.
Sounds good. I will have to see if I go for 2510 or 2010.

What diameter and pitch, ratio and stepper RPM are you running on your machine to achieve the 10m/min?

You'll have to find build logs for comparable routers using servos - I don't have any experience with them so won't advise.
Good advice, thanks. I have looked around here and I will also browse through CNCZONE to see if I can find something similar.



Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

The surface finish achieved in this video, I would be more than happy with:

Thought you said wanted a good surface finish.!!. . . . Any decently built machine will give you that finish.

Also:
If you go with 30mm rails and 25 or 32mm ball-screws then you can forget 180w or 400w servos. The extra inertia will freak a 400w servo motor when trying to stop 25mm screws from 3000rpm that are attached to a reasonably heavy Gantry which is sat on heavy bearings with a heavy ATC spindle hanging off it.

The best machines are those that have the right balance of power and weight which means getting the design and components matched. The choices you are looking at now will give you a very unbalanced machine.!

20mm screws and rails are more than enough for a machine this size. Regards the screw size and whip etc then forget whip calculators because they can't and don't account for the whole machine.
If you want proof if 20mm won't whip at 15mtr/min then just ask anyone who's using them on similar sized machine.! . . . . I've built dozens and trust me they don't if properly aligned and with correct end bearings. However, I've told you how to eliminate any chance of whip and still get the speed you require.

Also on that note, you have to factor in the Servos and the extra resolution and power they offer over steppers. This means you can use higher pitch screws and still end up with good resolution compared to steppers. Use this along with applying a ratio and the screw speed can be lowered greatly.

Regards the weight then I think you must have got something wrong because a machine this size will easily weigh more than 200kg, even 300Kg by time it's all finished.

[AndyUK]

Regards the weight then I think you must have got something wrong because a machine this size will easily weigh more than 200kg, even 300Kg by time it's all finished.

Whats the maximum weight you'd expect though? I think a 1,500 Kg minimum as suggested in post #2 is a little overkill.

[NordicCnc]

Thought you said wanted a good surface finish.!!. . . . Any decently built machine will give you that finish.

Hehe, seems like we have different standards! I think that finish is quite good and I would be happy with that. Of course, I will try to achieve the best finish I can with my DIY build. I will not settle with decent, if I can get good.


Also:
If you go with 30mm rails and 25 or 32mm ball-screws then you can forget 180w or 400w servos. The extra inertia will freak a 400w servo motor when trying to stop 25mm screws from 3000rpm that are attached to a reasonably heavy Gantry which is sat on heavy bearings with a heavy ATC spindle hanging off it.

The best machines are those that have the right balance of power and weight which means getting the design and components matched. The choices you are looking at now will give you a very unbalanced machine.!

Yes, I got confused with the advice from jarjar. It seemed quite contradicting to the general advice I've seen on the forums. He mentioned that a 400W would probably be the bare minimum. Anyway I don't plan to rotate any ballscrews with 3000rpm. I will use at least a 2:1 ratio, resulting in max. 1500rpm.


20mm screws and rails are more than enough for a machine this size. Regards the screw size and whip etc then forget whip calculators because they can't and don't account for the whole machine.

As I planned from then beginning, to use 20mm rails. Thanks for confirming!


If you want proof if 20mm won't whip at 15mtr/min then just ask anyone who's using them on similar sized machine.! . . . .

I've seen those builds, which is why I was skeptical and questioned jarjar's advice about the huge ballscrews.


I've built dozens and trust me they don't if properly aligned and with correct end bearings. However, I've told you how to eliminate any chance of whip and still get the speed you require.

Correct. About preloading the ballscrew with tension, using double fixed bearings. Is there any guide how to do this? I guess the applied force by tightening the fixed bearing nuts will be critical?


Regards the weight then I think you must have got something wrong because a machine this size will easily weigh more than 200kg, even 300Kg by time it's all finished.

That is exactly what I said, that I expect it to be around the 300kg mark once finished. Probably above! :)


Also on that note, you have to factor in the Servos and the extra resolution and power they offer over steppers. This means you can use higher pitch screws and still end up with good resolution compared to steppers. Use this along with applying a ratio and the screw speed can be lowered greatly.

My plan is now the following:

- I will build the machine with motor brackets than can be replaced. This will allow me to at least first test the 180W JMC servo motors. If they are not powerful enough even experimenting with ratios, I will have to buy new motors - probably some 400W servo motors or get stepper motors. This will be an expensive learning lesson, but that is life! I will most likely never buy any parts before design is ready again, lol.
- 20mm HIWIN linear rails on all axis's.
- 2010 ball screws on the Y-axis (base frame).
- 1610 or 2010 ball screw on the X-axis (gantry).
- 1605 ball screw on the Z-axis.

[NordicCnc]

Alright first actual design post with some pictures.

Z-axis design

- 3kW ATC spindle & Ø100mm spindle bracket. The spindle alone weights about 19kg according to the supplier (RattmMotors).

27501 27502

- 20x200x500mm aluminium rear plate, onto which spacers, BK12/BF12 bearings and motor bracket are mounted.
- 15x50x500mm aluminium spacers to make room for BK12/BF12 bearings and 1605 ballscrew. Linear rails are mounted to the spacers.
- 20x200x171mm aluminium top plate for HIWIN narrow guide carriages that goes on top of the gantry. Gantry ballscrew nut bracket will be mounted to the same plate.
- 20x200x240mm aluminium front plate, onto which the HIWIN narrow guide carriages and ballscrew nut bracket DSG16H are mounted.
27490 27491 27494 27493 27492 27495

- About 200mm gantry clearance
- 146mm overhang
- Spindle can be raised about 60mm above gantry
- 1605 ballscrew, 400m long with 25mm F-length end machining for BK12 fixed bearing to allow for pulley. Floating bearing in the other end.
- 20T HTD 5M pulley on the servomotor and 40T HTD 5M pulley on the ballscrew.
- 180W JMC servo motor.
- Motor bracket with oval holes to allow belt tensioning.

27496 27497 27498 27499 27500 27505

- Moving weight in Z-axis direction, about 26kg - still missing cables, water pipes, proximity switches etc
- Moving weight in X-axis (gantry) direction: 40kg - still missing cables, water pipes, proximity switches etc.

27503 27504

I am pretty happy with the total weight of the Z-axis. I don't think it'll go over 45-50kg with everything added.

I could still trim down the X-axis direction moving weight, by removing the linear rail spacers and adding machining to the front and rear plates. But I am thinking that I will need that stiffness which the 20mm plates gives me. Would it weaken the structure a lot by adding machining of 7x60mm along the middle of the plates, to allow room for the BK/BF bearings and ballscrew nut bracket?

[JAZZCNC]

Whats the maximum weight you'd expect though? I think a 1,500 Kg minimum as suggested in post #2 is a little overkill.

Think JarJar is talking about a different strength machine, esp when he's comparing to a Datron which is a big lump of epoxy Granite. Thou that's like comparing apples with pears when a Datron cost's like £100K.

I'd expect a machine this size to be around 400Kg when finished using the materials and components suggested. It's amazing just how much weight fasteners etc add.

[JAZZCNC]

- 3kW ATC spindle & Ø100mm spindle bracket. The spindle alone weights about 19kg according to the supplier (RattmMotors).

The spindle with Mount + BT30 Collet chuck weighs 19.5Kg. Don't forget if you are using dust extraction then you will need to retract the dust hood to allow access for tool change. This will mean adding pneumatic actuators and solenoid valves along with brackets etc and the dust hood its self.!... The weight creeps up fast.

Your estimate of 45-50Kg is about right. You will need a brake on the Z-axis motor to stop it dropping when not under power and i'm pretty sure your 180w motor won't handle the weight of ATC spindle.

[pippin88]

Add a second spindle bracket up high on the spindle.

Add ribs on the edge of the Z axis plates. This will add a huge amount to stiffness. (Essentially make them into C channels).
Nowhere on the machine should you have a simple flat plate - everything should be box sections of at all possible, otherwise should have stiffening ribs.

[NordicCnc]

The spindle with Mount + BT30 Collet chuck weighs 19.5Kg. Don't forget if you are using dust extraction then you will need to retract the dust hood to allow access for tool change. This will mean adding pneumatic actuators and solenoid valves along with brackets etc and the dust hood its self.!... The weight creeps up fast.

I had not thought about retractable dust shoe. I need to investigate if I will need one!


Your estimate of 45-50Kg is about right. You will need a brake on the Z-axis motor to stop it dropping when not under power and i'm pretty sure your 180w motor won't handle the weight of ATC spindle.

The moving weight in the Z-axis direction is 25kg and the complete Z-axis will be around 50kg. 25kg will for sure drop when the Z-motor is not under power. Could I use stepper motor brakes on these servo motors?

In the one of the YouTube videos I linked to earlier, one guy is using the same servos and he told me that his Z-axis is around 20kg (I asked). This is only about 5kg lighter than mine and he said it works very well! His axis drops when not under power too.

Anyway as I mentioned earlier, I won't let the servos I have dictate the final design. I will buy new stronger ones if it turns out that they are too weak!

[NordicCnc]

Add a second spindle bracket up high on the spindle.

According to the spindle specification, the clamping area is defined here:

27511

I am afraid that clamping higher up (I am at the highest allowed point already) could potentially damage the spindle. Please correct me if I am wrong? The reason why the spindle is clamped like this is for gantry clearance.


Add ribs on the edge of the Z axis plates. This will add a huge amount to stiffness. (Essentially make them into C channels).
Nowhere on the machine should you have a simple flat plate - everything should be box sections of at all possible, otherwise should have stiffening ribs.

Good point! While I am trying to keep the Z-axis weight low, I think adding 15mm aluminium stiffening ribs wouldn't be too harmful (only around 1.2kg added). If I understood you correctly, you means something similar to this:

27512 27513 27514

Additionally I could also move the Z-axis linear rail spacers to the front plate. This would increase the moving weight by another 1kg. I am not sure fi I should keep them on the rear plate or not. I could move them to the front plate and add stiffening ribs to the back of the rear plate. The ribs would be bolted to the rear plate, top bearing plate and motor bracket, effectively forming a rigid box. Is this just adding dead weight?

27516

[pippin88]

Regarding spindle, I didn't realise it was an ATC spindle, which are a lot longer. Clamping high up may not be appropriate (though with a decent tolerance clamp I doubt you'd be crushing the spindle). The point is the longer the clamping area (or the further apart the clamping points) the better.

Yes, that's what I meant about the ribs. You don't need 15mm thick plate, even 5mm would be fine. The stiffness comes from the dimension in the direction of the force.

You should also add stiffening ribs to the carriage .

[AndyUK]

The ribs would be bolted to the rear plate, top bearing plate and motor bracket, effectively forming a rigid box. Is this just adding dead weight?

Machines only as strong as its weakest point. Currently what is keeping the back of your Z axis still? Its those carriages, which are attached via a 90 degree but-joint. So I'd say those two or so screws are the weakest point! That's why I created the side plates on my Z axis to give some support to the joints.

[NordicCnc]

Regarding spindle, I didn't realise it was an ATC spindle, which are a lot longer. Clamping high up may not be appropriate (though with a decent tolerance clamp I doubt you'd be crushing the spindle). The point is the longer the clamping area (or the further apart the clamping points) the better.

Yes, that's what I meant about the ribs. You don't need 15mm thick plate, even 5mm would be fine. The stiffness comes from the dimension in the direction of the force.

You should also add stiffening ribs to the carriage .


Machines only as strong as its weakest point. Currently what is keeping the back of your Z axis still? Its those carriages, which are attached via a 90 degree but-joint. So I'd say those two or so screws are the weakest point! That's why I created the side plates on my Z axis to give some support to the joints.

Thanks both of you pippin and Andy! I took your advice and now I've added 10mm aluminium ribs bolted to the front plate. I've also added 10mm side plates that are bolted to the rear plate, motor bracket and top bearing plate.

27517 27518

New total weights did not change much, while I think I have added quite a bit of stiffness!

Z-axis moving weight: 26.8kg -> 26.4kg
Z-axis total weight: 39.7kg -> 42.4kg

27519 27520

I am now debating, should I do something of the following things:

1. Remove the spacer plates used for the linear rails completely and add machining (on both the rear and front plate)? This would reduce weight (maybe 3-4kg?) and fewer parts needed. More machining though.

27521

2. Move the spacers from the rear plate to the front plate. Linear rails would then go directly on the rear plate. Shorter spacers would be added behind the guide carriages and the front plate. This would increase the Z-axis moving weight by like 1kg, but reduce total weight by 1kg. I would gain some stiffness on the front plate, but lose some on the rear plate. Not sure where the weak point is and where the spacers are needed the most.

27522

[JAZZCNC]

Ok well first, the ribs on the front plate add virtually nothing but weight because the plate is fastened directly to the bearings with no overhang, it's also short and strong so very little vibrations come from that area.

Regards the plates on the sides then these would help, however, I have a better suggestion. If you look at the pic and at the other machines I've built I always put a cover over the Z-axis. This isn't just to protect from debris it's also structural as it does the same job those plates. You don't need a 10mm plate, I use 3mm for the covers and it's more than strong enough.

The spindle mounting area is limited to the lower area (see the pic, it's the silver area, the black is drawbar cylinders.) However, they are steel and would easily handle another clamp.

Regards putting the spacer plates on the front plate to save weight then I wouldn't do that either because of your shifting weight to the wrong area.
To be honest I wouldn't use spacer plates as it lowers precision and makes it complicated.

Regards Motor brake then I can't answer that because I don't know those motors, but servos are not like steppers where they have a shaft sticking out the back, mostly because this is often used for the encoder.

How the motor handles the weight will depend on usage, but if you plan on doing 3D type work with lots of short moves with high acceleration then I think they 180W motors will struggle with the inertia and you'll get some following errors. Because again servos are unlike steppers where if they lose the position you visually see the lost steps, instead the encoders along with the drives close the loop and catch back up so will always return to exact same place when stopped provided they don't go outside of parameters set in the drives.

However, what you get is a following error which if it stays within following error parameters can trick people into thinking they don't have a problem because if they do a type of work which is mostly 2D the Z-axis isn't whizzing up & down so it gets missed.
27530 27531 27532

[NordicCnc]

Ok well first, the ribs on the front plate add virtually nothing but weight because the plate is fastened directly to the bearings with no overhang, it's also short and strong so very little vibrations come from that area.

Regards the plates on the sides then these would help, however, I have a better suggestion. If you look at the pic and at the other machines I've built I always put a cover over the Z-axis. This isn't just to protect from debris it's also structural as it does the same job those plates. You don't need a 10mm plate, I use 3mm for the covers and it's more than strong enough.

The spindle mounting area is limited to the lower area (see the pic, it's the silver area, the black is drawbar cylinders.) However, they are steel and would easily handle another clamp.

Regards putting the spacer plates on the front plate to save weight then I wouldn't do that either because of your shifting weight to the wrong area.
To be honest I wouldn't use spacer plates as it lowers precision and makes it complicated.

Regards Motor brake then I can't answer that because I don't know those motors, but servos are not like steppers where they have a shaft sticking out the back, mostly because this is often used for the encoder.

How the motor handles the weight will depend on usage, but if you plan on doing 3D type work with lots of short moves with high acceleration then I think they 180W motors will struggle with the inertia and you'll get some following errors. Because again servos are unlike steppers where if they lose the position you visually see the lost steps, instead the encoders along with the drives close the loop and catch back up so will always return to exact same place when stopped provided they don't go outside of parameters set in the drives.

However, what you get is a following error which if it stays within following error parameters can trick people into thinking they don't have a problem because if they do a type of work which is mostly 2D the Z-axis isn't whizzing up & down so it gets missed.

Thanks for your advice. I think you forgot to add a picture though?

I will reply back once you have added, as I think things will make more sense to me then.

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

Thanks for your advice. I think you forgot to add a picture though?

I will reply back once you have added, as I think things will make more sense to me then.

Skickat från min SM-A530F via Tapatalk

No didn't forget the forum is playing silly buggers again and won't let me upload the pic. I'll try again later.

[NordicCnc]

No didn't forget the forum is playing silly buggers again and won't let me upload the pic. I'll try again later.Thanks man!

Skickat från min SM-A530F via Tapatalk

[phill05]

Here's where I throw a spanner in the works.

This is from my own way of working I think you are going way overboard with the construction of the Z axis adding more and more to it.

I made a Z axis 510mm high x 220mm wide using 12mm backing plate with cut outs to ease the weight on the Y axis with 20mm Hywin rails direct on the plate no packers and carriages direct on the spindle plate, I made a rotating ball nut and fixed 1605 screw which helps to have the very minimum of overhang and adds to the strength of the design.
27525

It carries a German 2.2kw water cooled spindle, (I know not as heavy as yours) I do not get any drop of the Z when powered off weather this is down to the 2 to 1 reduction to the nut I am not sure but it works very well, it has cut wood, plastics, aluminium, steel plate and stone with no problems whatsoever.

Recently I setup a 4th axis to cut 2 1/2D designs around the edges of 150mm disks again works really well.
I am currently designing another machine with extended Y axis to be able to have a 4th axis running down the length of the X axis, and be able to cut around 300mm dia disks and have no problem using the same kind of design in fact I am going to use a rotating ball nut on the X & Y axis as well it is more compact and so smooth compared to turning screws.
27526

Phill

[NordicCnc]

Here's where I throw a spanner in the works.

This is from my own way of working I think you are going way overboard with the construction of the Z axis adding more and more to it.

I made a Z axis 510mm high x 220mm wide using 12mm backing plate with cut outs to ease the weight on the Y axis with 20mm Hywin rails direct on the plate no packers and carriages direct on the spindle plate, I made a rotating ball nut and fixed 1605 screw which helps to have the very minimum of overhang and adds to the strength of the design.
27525

It carries a German 2.2kw water cooled spindle, (I know not as heavy as yours) I do not get any drop of the Z when powered off weather this is down to the 2 to 1 reduction to the nut I am not sure but it works very well, it has cut wood, plastics, aluminium, steel plate and stone with no problems whatsoever.

Recently I setup a 4th axis to cut 2 1/2D designs around the edges of 150mm disks again works really well.
I am currently designing another machine with extended Y axis to be able to have a 4th axis running down the length of the X axis, and be able to cut around 300mm dia disks and have no problem using the same kind of design in fact I am going to use a rotating ball nut on the X & Y axis as well it is more compact and so smooth compared to turning screws.
27526

Phill

Hello, looks good! Interesting with the rotating ball nut. I understand it so that you have placed the fixed bearing on the spindle plate and then the other ball screw end is lose. Does it stick up in the air or what does it look like when the Z-axis is fully retracted?

Anyway I have to say that your advice is quite contradicting compared to the advice I have received from others. I am not saying you are wrong but I try to stay open minded!

[phill05]

Hello, looks good! Interesting with the rotating ball nut. I understand it so that you have placed the fixed bearing on the spindle plate and then the other ball screw end is lose. Does it stick up in the air or what does it look like when the Z-axis is fully retracted?

Anyway I have to say that your advice is quite contradicting compared to the advice I have received from others. I am not saying you are wrong but I try to stay open minded!

I did say I was throwing a spanner in the works, I know it's not a conventional way of making this and yes contradicting to what has gone before but it works and works really well it is compact, less weight, and the machine cuts very well, just wanted to give you an idea of what works without going to far.

The screw is bolted tight to bottom of spindle plate (no bearing) and screw goes up with plate, attached is an image when I first put it together.
27528

Phill

[NordicCnc]

I did say I was throwing a spanner in the works, I know it's not a conventional way of making this and yes contradicting to what has gone before but it works and works really well it is compact, less weight, and the machine cuts very well, just wanted to give you an idea of what works without going to far.

The screw is bolted tight to bottom of spindle plate (no bearing) and screw goes up with plate, attached is an image when I first put it together.
27528

Phill

Oh my! You have quite some Z-axis travel height there. I can see that the screw will be sticking out in the air when Z-axis is fully retracted (at the top end position). I think you rarely use that Z-height or are you doing some custom jobs?

Looks like the gantry clearance is around 100-120mm, am I right? It would also be interesting seeing a video of its performance in aluminium/steel!

You have definitely given me something to consider, thanks for that.

[JAZZCNC]

Here's where I throw a spanner in the works.

This is from my own way of working I think you are going way overboard with the construction of the Z-axis adding more and more to it.

You can't be serious.? You are using a rotating ball-nut on the Z-axis and you think he's going overboard with things.!

That design with the screw sticking up 400-500mm above the z-axis when at the top is just a crazy, pointless design and overly complicated for its application. Rotating ball-nuts are great on longer axis but for a Z-axis they will make no difference because the screw is so short that whipping or inertia don't factor into the equation which is the only reason to use a rotating nut, even with servo's which spin much faster than steppers.

Also regards your intended design with the L-shape Gantry that is all mixed up as well. Your design wastes Gantry clearance height because using the middle slot on the profile and misses the whole point of the "L" design. The point being the L shape is so most of the cutting forces are directed into the horizontal piece placed at the bottom where forces are highest.

[NordicCnc]

Ok well first, the ribs on the front plate add virtually nothing but weight because the plate is fastened directly to the bearings with no overhang, it's also short and strong so very little vibrations come from that area.

Regards the plates on the sides then these would help, however, I have a better suggestion. If you look at the pic and at the other machines I've built I always put a cover over the Z-axis. This isn't just to protect from debris it's also structural as it does the same job those plates. You don't need a 10mm plate, I use 3mm for the covers and it's more than strong enough.

The spindle mounting area is limited to the lower area (see the pic, it's the silver area, the black is drawbar cylinders.) However, they are steel and would easily handle another clamp.

Regards putting the spacer plates on the front plate to save weight then I wouldn't do that either because of your shifting weight to the wrong area.
To be honest I wouldn't use spacer plates as it lowers precision and makes it complicated.

Regards Motor brake then I can't answer that because I don't know those motors, but servos are not like steppers where they have a shaft sticking out the back, mostly because this is often used for the encoder.

How the motor handles the weight will depend on usage, but if you plan on doing 3D type work with lots of short moves with high acceleration then I think they 180W motors will struggle with the inertia and you'll get some following errors. Because again servos are unlike steppers where if they lose the position you visually see the lost steps, instead the encoders along with the drives close the loop and catch back up so will always return to exact same place when stopped provided they don't go outside of parameters set in the drives.

However, what you get is a following error which if it stays within following error parameters can trick people into thinking they don't have a problem because if they do a type of work which is mostly 2D the Z-axis isn't whizzing up & down so it gets missed.
27530 27531 27532

JAZZ, I think the way you have installed the pneumatic actuators and the Z-axis cover design is great! The whole machine design is phenomenal.. I hope you don't me taking ideas and copying some features from it?

I've removed my spacer plates now and added a machining to both the rear and the front plate to make room for the bearing and ball screw nut housing. This reduced the overhang by 15mm! It also reduced the total weight a few kg's. Thanks for the tip!

[BobTSkutter]

Here's where I throw a spanner in the works.

This is from my own way of working I think you are going way overboard with the construction of the Z axis adding more and more to it.

I made a Z axis 510mm high x 220mm wide using 12mm backing plate with cut outs to ease the weight on the Y axis with 20mm Hywin rails direct on the plate no packers and carriages direct on the spindle plate, I made a rotating ball nut and fixed 1605 screw which helps to have the very minimum of overhang and adds to the strength of the design.
27525

It carries a German 2.2kw water cooled spindle, (I know not as heavy as yours) I do not get any drop of the Z when powered off weather this is down to the 2 to 1 reduction to the nut I am not sure but it works very well, it has cut wood, plastics, aluminium, steel plate and stone with no problems whatsoever.

Recently I setup a 4th axis to cut 2 1/2D designs around the edges of 150mm disks again works really well.
I am currently designing another machine with extended Y axis to be able to have a 4th axis running down the length of the X axis, and be able to cut around 300mm dia disks and have no problem using the same kind of design in fact I am going to use a rotating ball nut on the X & Y axis as well it is more compact and so smooth compared to turning screws.
27526

Phill

hello, i would really like to know how you made that rotating ball nut. It looks like a flanged ball nut sandwiched in a pillow block bearing housing with a pulley holding it together. please can you share some pictures.

thanks
bob

[JAZZCNC]

JAZZ, I think the way you have installed the pneumatic actuators and the Z-axis cover design is great! The whole machine design is phenomenal.. I hope you don't me taking ideas and copying some features from it?

No don't mind at all , go for it.! . . . Thou bare in mind some of these pictures are not of the finished article, the covers, for instance, are still missing some fastenings and slot covers, etc but hopefully show what I've been saying.

[NordicCnc]

No don't mind at all , go for it.! . . . Thou bare in mind some of these pictures are not of the finished article, the covers, for instance, are still missing some fastenings and slot covers, etc but hopefully show what I've been saying.

Great and you have explained it very well!

By the way I have now decided that I will not use the 180W JMC servo motors that I bought in advance. I've come to realize that they are holding me back by limiting the weight of he machine. It was an expensive learning lesson but now I know that I should always wait with buying any components until the design is completed.

I've contacted the supplier and they might be willing to accept a return and I would pay a little more to get the 400W JMC servo motors in change (https://www.aliexpress.com/item/4000067490526.html). In case they don't accept the return for some reason then I have already found another use case for them, so that problem is solved - no money lost. If I then still buy the 400W JMC's, I don't know, I will have to check what you and the others here think. I saw that delta servo's were recommended as well and it was also mentioned that larger than 400W would probably be good?

Anyway the motor power issue is now out of the way and I have the money to spend on new and way more powerful motors!

So now having solved the motor power issue, do I still keep the Z-axis weight as low as possible or should I start trying to beef it up? The covers I will design to be structural, taking ideas from your machines JAZZ. The Z-axis front plate I am not so sure if I can improve much since it is already very stiff. Perhaps the Z-axis design is good as light weight and I can just stop worrying if the motor power will be enough to handle the heavy ATC spindle! :peaceful:

[NordicCnc]

Short update on the Z-axis design! You might notice that the design is heavily influenced by JAZZCNC's build's (e.g. the cover and L-shape gantry), although still very different and definitely no where near his quality!

Modifications:

Based on feedback I have now removed the stiffening ribs from the front plate. The reason is that since the front plate is directly mounted to the X-axis bearings with literally no overhang, the stiffening ribs would add nothing but weight. Instead a 3mm aluminium sheet is bent and welded, forming a cover that is structural while also protecting the components and cables from chips.
Removed the linear rail spacer plates and added machining to both the rear and front plate, to make room for the BK12/BF12 bearings and the 1605 ball screw.
180W servo motors are no longer a constraint. Now I am thinking about getting Delta 400W servo motors. On the Z-axis I plan to use one with brake: https://www.aliexpress.com/item/32671356577.html?spm=2114.12010612.8148356.2.23d15 2a8ZJcwMZ. On the X- and Y-axis I plan to use ones without brakes: https://www.aliexpress.com/item/32671356516.html?spm. All axis with still be geared 2:1, to limit the ball screw maximum speed to 1500rpm.
Modified the top bearing plate to account for BK15/BK15 bearings and 2010 ball screw on the X-axis (in design process).
Preliminary positioning of proximity switches. The X-axis will only need one, but for the Z-axis I might need 2 if I can't find a location which I can use for both homing/top limit switch as well as bottom limit switch.
New Z-axis weights are 45kg (total) and 25kg (moving).


27538 27539 27540 27541 27542 27543 27544

Here is the Z-axis mounted on the draft X-axis design, made with 120x80 Item profiles (I will post a more detailed X-axis design description later):
27545 27546 27547

[Clive S]

Preliminary positioning of proximity switches. The X-axis will only need one, but for the Z-axis I might need 2 if I can't find a location which I can use for both homing/top limit switch as well as bottom limit switch.

Its not easy or practical to put a bottom limit switch on the Z axis as the tool lengths will all vary .

[NordicCnc]

Its not easy or practical to put a bottom limit switch on the Z axis as the tool lengths will all vary .

True but my intention of using one is not to protect the tool itself, but to protect the machine from crashing into the BF12 floating bearing for any unintended reason. On my current machine I've been completely fine without any limit switches, but that machine is made with weaker stepper motors and is only running 4000mm/min at maximum feed rate.

[Clive S]

True but my intention of using one is not to protect the tool itself, but to protect the machine from crashing into the BF12 floating bearing for any unintended reason. On my current machine I've been completely fine without any limit switches, but that machine is made with weaker stepper motors and is only running 4000mm/min at maximum feed rate.

Have you considered soft limits?

[NordicCnc]

Have you considered soft limits?

No I have not, but that should work perfectly with position feedback from the servos. Thanks for the tip!

[NordicCnc]

Regards Motor brake then I can't answer that because I don't know those motors, but servos are not like steppers where they have a shaft sticking out the back, mostly because this is often used for the encoder.
27530 27531 27532

JAZZ, what motors are you using to drive the ball screws and is the Z-axis motor with a brake? The Z-axis for my machine will be quite the same weight.

If you read one of my last posts in the build log http://www.mycncuk.com/threads/13448-NordicCNC-s-build-log?p=114823#post114823, I wrote that I will probably use Delta 400W servo motors and with a brake on the Z-axis. Well now I've got an offer for a full return from the supplier of those JMC 180W with integrated drivers, if I swap them for JMC 400W servo motors with integrated drivers. The problem here is that those JMC 400W servo motors does not come with a brake option.

[jarjar]

A 400W servo (rated at 1.2Nm @ 3000rpm) is plenty enough. 200W is also probably enough.
I will be using a Yaskawa 400W servo with brake for a 60kg Z axis (moving part only), driven by a 5mm pitch ballscrew @ 1g accel and 20m/min.
The servo has a peak torque of 4Nm, and moving this axis will only require about 2Nm peak.

[NordicCnc]

A 400W servo (rated at 1.2Nm @ 3000rpm) is plenty enough. 200W is also probably enough.
I will be using a Yaskawa 400W servo with brake for a 60kg Z axis (moving part only), driven by a 5mm pitch ballscrew @ 1g accel and 20m/min.
The servo has a peak torque of 4Nm, and moving this axis will only require about 2Nm peak.

In post #2 http://www.mycncuk.com/threads/13448-NordicCNC-s-build-log?p=114672#post114672 you said that these 180W servos that I had bought were only toys and that 400W would be the bare minimum. I think you were talking about a little bit different design in that case so I can just ignore it, correct? Anyway the price difference between a Delta 200W servo and a Delta 400W servo is minimal.

Will you be using direct drive for that Z-axis? I checked that those servos are rated at 3000rpm, so direct drive with 5mm pitch would result in 15m/min. Also what diameter will that ball screw be and is it possible to spin it that fast?

Your moving weight sounds extremely heavy! Are you designing it in steel? How did you calculate how much Nm will be needed to move the axis?

Wow, this turned into a lot of questions, lol!

[jarjar]

A JMC 180W might be enough power wise, but industrial servos are much more sophisticated. Encoder resolution, motion smoothing (much needed for high accel when the controller doesn't handle S-curve motion profiles), vibration suppression, ...
To be honest I also had a different machine in mind from what you're planning when you stated your requirements :)

For calculating the required motor, I simply plugged the numbers in Yaskawa's calculator (SigmaSelect).
It's a 2005 ballscrew coupled with 1:1 pulleys (I would have liked to go direct-drive, but didn't have enough space to inline the motor). The axis can reach 30m/min because servo max speed is 6000 rpm. Ballscrew critical speed is not a concern since it's short (400mm) and could in theory handle 13'000rpms.

Yes the axis is all steel, but the spindle is only 2.2kW ISO20 30krpm. The goal is high-speed aluminium milling.

[JAZZCNC]

A JMC 180W might be enough power wise, but industrial servos are much more sophisticated. Encoder resolution, motion smoothing (much needed for high accel when the controller doesn't handle S-curve motion profiles), vibration suppression, ...
To be honest I also had a different machine in mind from what you're planning when you stated your requirements :)

For calculating the required motor, I simply plugged the numbers in Yaskawa's calculator (SigmaSelect).
It's a 2005 ballscrew coupled with 1:1 pulleys (I would have liked to go direct-drive, but didn't have enough space to inline the motor). The axis can reach 30m/min because servo max speed is 6000 rpm. Ballscrew critical speed is not a concern since it's short (400mm) and could in theory handle 13'000rpms.

Yes the axis is all steel, but the spindle is only 2.2kW ISO20 30krpm. The goal is high-speed aluminium milling.

Nordic be careful here because it's like comparing apples with oranges when comparing Yaskawa motors with Cheaper Chinese motors. The price difference is big and so is the quality. I'd also guess to get those speeds the Yaskawa will use an absolute encoder system with high count encoders requiring a high-frequency controller.

When you move up into these levels then everything else attached to them needs to move up in quality with it, else no point fitting expensive high-quality servos.
Your 8K budget will soon get eaten up with motors and ball-screws.

[jarjar]

Delta is Taiwanese and they make proper cheap servos. Also 400W Yaskawa servos can be found rather cheap (~$400 motor + drive + cables). Most similar sized servos have the same ratings. 6000 rpm is nothing special, same for absolute encoders.

I asked a quote for the latest ASD-A3 drives from Delta with a 750W servo, 24-bit incremental/absolute encoder. Price was $350.

But Jazz is right to warn you. Servos are more complex than steppers. I went Yaskawa just because their manual is more comprehensive than the Delta one.

[NordicCnc]

Nordic be careful here because it's like comparing apples with oranges when comparing Yaskawa motors with Cheaper Chinese motors. The price difference is big and so is the quality. I'd also guess to get those speeds the Yaskawa will use an absolute encoder system with high count encoders requiring a high-frequency controller.

When you move up into these levels then everything else attached to them needs to move up in quality with it, else no point fitting expensive high-quality servos.
Your 8K budget will soon get eaten up with motors and ball-screws.

Thanks for the heads up JAZZ. I am set on using chinese servos anyway because of the price and their good reputation.

Regarding the controller pulse output then I don't think I need to worry. The ESS SmoothStepper can output up to 4MHz according to its specs.

Would you also have time to check my questions in post #44? I would really appreciate it man! http://www.mycncuk.com/showthread.php?p=114834

Skickat från min SM-A530F via Tapatalk

[NordicCnc]

A JMC 180W might be enough power wise, but industrial servos are much more sophisticated. Encoder resolution, motion smoothing (much needed for high accel when the controller doesn't handle S-curve motion profiles), vibration suppression, ...
To be honest I also had a different machine in mind from what you're planning when you stated your requirements :)

For calculating the required motor, I simply plugged the numbers in Yaskawa's calculator (SigmaSelect).
It's a 2005 ballscrew coupled with 1:1 pulleys (I would have liked to go direct-drive, but didn't have enough space to inline the motor). The axis can reach 30m/min because servo max speed is 6000 rpm. Ballscrew critical speed is not a concern since it's short (400mm) and could in theory handle 13'000rpms.

Yes the axis is all steel, but the spindle is only 2.2kW ISO20 30krpm. The goal is high-speed aluminium milling.

Thanks for the link to the calculation software, this will be very useful to me.

Myself I think I am going to consider a 2005 ballscrew for the Z-axis. My ball screw length is also 400mm long and who would not appreciate a 15m/min rapid feed rate! Originally I planned for a 1605 with a 2:1 reduction, since I worried about whip (ball screw critical speed) and increased inertia (180W servos). These things are no longer an issue!

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

Delta is Taiwanese and they make proper cheap servos. Also 400W Yaskawa servos can be found rather cheap (~$400 motor + drive + cables). Most similar sized servos have the same ratings. 6000 rpm is nothing special, same for absolute encoders.

I asked a quote for the latest ASD-A3 drives from Delta with a 750W servo, 24-bit incremental/absolute encoder. Price was $350.

But Jazz is right to warn you. Servos are more complex than steppers. I went Yaskawa just because their manual is more comprehensive than the Delta one.

I wasn't referring to Delta when I said Chinese servo's it was a general point between Cheap and Expensive Servo's. Regards the Yaskawa motors -drive-cable package then can you post a link to where your buying because that's a very good price. What model is that.?

Regards the speed I don't see many that are rated at 6k rpm in the lower price ranges, 3K rpm is a typical rating with incremental encoders. Yes, they can be run up to 5-6K rpm but they are not rated to run at that speed constant.!

[jarjar]

I wasn't referring to Delta when I said Chinese servo's it was a general point between Cheap and Expensive Servo's. Regards the Yaskawa motors -drive-cable package then can you post a link to where your buying because that's a very good price. What model is that.?

Regards the speed I don't see many that are rated at 6k rpm in the lower price ranges, 3K rpm is a typical rating with incremental encoders. Yes, they can be run up to 5-6K rpm but they are not rated to run at that speed constant.!
There are currently tons of offers on Ebay for SGM7J-04AF + SGD7S-2R8A00 at $450 new. I bought one like this a while back for a little less. I just asked for cables and they included them with no extra charges.
You can even buy the model with absolute encoder separately for less :rolleyes: (SGM7J-04A7 $160 + SGD7S-2R8A00 $260)

I didn't say 6000rpm is the nominal rating, but it's still useful for short rapid moves.

[JAZZCNC]

JAZZ, what motors are you using to drive the ball screws and is the Z-axis motor with a brake? The Z-axis for my machine will be quite the same weight.

If you read one of my last posts in the build log http://www.mycncuk.com/threads/13448-NordicCNC-s-build-log?p=114823#post114823, I wrote that I will probably use Delta 400W servo motors and with a brake on the Z-axis. Well now I've got an offer for a full return from the supplier of those JMC 180W with integrated drivers, if I swap them for JMC 400W servo motors with integrated drivers. The problem here is that those JMC 400W servo motors does not come with a brake option.

The motors are 8Nm Lichuan closed loop steppers using 230Vac mains drives. It doesn't use a break because the gantry is vertical so Z-axis isn't fighting gravity, plus 8Nm is more than enough for this Z-axis even with heavy ATC spindle on it.

Regards the JMC then I've no experience using these motors or drives so I can't comment on how good or bad they are and I haven't looked at the specs, but if I'm honest I wouldn't use motors with integrated drives on a CNC machine. The heat and vibrations from the motors along with cutting debris, esp from aluminum coolant, etc, can't help with life expectancy but for me what's worse is the fact your running long signal wires back to the controller leaving you wide open to issues with interference from noise.

[NordicCnc]

The motors are 8Nm Lichuan closed loop steppers using 230Vac mains drives. It doesn't use a break because the gantry is vertical so Z-axis isn't fighting gravity, plus 8Nm is more than enough for this Z-axis even with heavy ATC spindle on it.
I totally missed that.. It is of course the X-axis in your case that could perhaps need a brake, but seems like its fine without.

Would there be any chance that a 400W Delta servo with brake couldn't handle my Z-axis of around 25-30kg, with a 2:1 reduction or even 1:1? I could go for 600W if that would be better.

Regards the JMC then I've no experience using these motors or drives so I can't comment on how good or bad they are and I haven't looked at the specs, but if I'm honest I wouldn't use motors with integrated drives on a CNC machine. The heat and vibrations from the motors along with cutting debris, esp from aluminum coolant, etc, can't help with life expectancy but for me what's worse is the fact your running long signal wires back to the controller leaving you wide open to issues with interference from noise.
Got it, thanks again. I will now completely forget about these servos with integrated drives.



Skickat från min SM-A530F via Tapatalk

[jarjar]

Would there be any chance that a 400W Delta servo with brake couldn't handle my Z-axis of around 25-30kg, with a 2:1 reduction or even 1:1? I could go for 600W if that would be better.

I believe I already answered that in details...

[NordicCnc]

I believe I already answered that in details...Gah you are right! In post #44 you did answer..

JAZZ, please ignore my repetitive question, and thanks jarjar for pointing out!

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

There are currently tons of offers on Ebay for SGM7J-04AF + SGD7S-2R8A00 at $450 new. I bought one like this a while back for a little less. I just asked for cables and they included them with no extra charges.
You can even buy the model with absolute encoder separately for less :rolleyes: (SGM7J-04A7 $160 + SGD7S-2R8A00 $260)

I didn't say 6000rpm is the nominal rating, but it's still useful for short rapid moves.

I'm surprised how much these have dropped in price, last time I looked, which must admit is quite some time ago, Yaskawa kits were like $900-$1100 range for 400W. I've never used them for this reason so I may look at these when a suitable job comes up.! . . . How easy or hard are they to setup/tune.? Do Yaskawa supply the Sigma tuning software for free.?

I'm surprised the absolute encoders are cheaper as well, not as many controllers about that can accept Absolute positioning so I would have expected them to be more. Thou absolute systems and Bus/Ethercat drives/controllers are coming down in price and more popular so maybe this is the reason.?

[JAZZCNC]

Nordic: Forgot to say in the previous post when mentioned about higher-level servos requiring higher quality components. That this is also true to lower degree for cheaper servos over a Stepper system.

For instance, with motor coupling used on Stepper, you can get away with more springy coupler and it won't affect the motor in any way. Whereas a springy coupler on even a cheap servo system can cause big issues with tuning the motor. The same goes if using belts and pulleys, buy good quality belts and pulleys.

The stiffer you can make the coupling between motor and screws the better this goes right down to motor mounting. Flimsy motor mounts etc will cause resonance which you'll have to tune out of the system. This isn't always so easy with cheap servo's so the more you can do to limit resonance the better.

[NordicCnc]

Nordic: Forgot to say in the previous post when mentioned about higher-level servos requiring higher quality components. That this is also true to lower degree for cheaper servos over a Stepper system.

For instance, with motor coupling used on Stepper, you can get away with more springy coupler and it won't affect the motor in any way. Whereas a springy coupler on even a cheap servo system can cause big issues with tuning the motor. The same goes if using belts and pulleys, buy good quality belts and pulleys.

The stiffer you can make the coupling between motor and screws the better this goes right down to motor mounting. Flimsy motor mounts etc will cause resonance which you'll have to tune out of the system. This isn't always so easy with cheap servo's so the more you can do to limit resonance the better.

I have never used servos before so I am not familiar with how the tuning works. All I know is that the Delta servos and most of the other servos comes with some auto tuning software. How good auto tuning works I have no idea.

Seems like tuning manually is very difficult and requires a lot of experience to get it right. If you don't know how to do it, the servo system would end up being even worse than a stepper system. I am starting to wonder if I would be better of with a good stepper system instead of diving into the realm of servos.

To me it seems like you favor steppers at for your own builds, but I may have got this wrong. If I am right, I guess you favor steppers for the simplicity and the above reasons?

By the way what would be a good supplier for belts and pulleys? What supplier are you using?

Thanks again for your great insight!


Skickat från min SM-A530F via Tapatalk

[jarjar]

I'm surprised how much these have dropped in price, last time I looked, which must admit is quite some time ago, Yaskawa kits were like $900-$1100 range for 400W. I've never used them for this reason so I may look at these when a suitable job comes up.! . . . How easy or hard are they to setup/tune.? Do Yaskawa supply the Sigma tuning software for free.?

I'm surprised the absolute encoders are cheaper as well, not as many controllers about that can accept Absolute positioning so I would have expected them to be more. Thou absolute systems and Bus/Ethercat drives/controllers are coming down in price and more popular so maybe this is the reason.?

Yes of course software is free. I'm not yet to the point of tuning but having read the manual multiple times I don't expect many issues. You start with the auto-tuning and fine-tune from there.

IMO rotary absolute encoders are not worth the hassle. Their only use in a router/mill context is to spare you the homing after power-up. And obviously the controller must be able to interpret the special Yaskawa protocol and I think no hobby controller can at this point. It could be done with some knowledge with LinuxCNC+Mesa.

[JAZZCNC]

Yes of course software is free.

No, it's not obvious really, it's not uncommon for high-end Servo manufacturers to charge for specialist tuning software if not using their own controllers just like it's not uncommon for PLC providers to charge for programming software.



I'm not yet to the point of tuning but having read the manual multiple times I don't expect many issues. You start with the auto-tuning and fine-tune from there.

I'd be interested in hearing how you get on with tuning. What controller and which method of control +/-10v or Step/Dir .?



IMO rotary absolute encoders are not worth the hassle.

What hassle.? Far less wiring and setting up far as I can tell, esp if using Bus type.! . . . I've not used any yet but soon will be doing so I'll let you know.

[JAZZCNC]

To me it seems like you favor steppers at for your own builds, but I may have got this wrong. If I am right, I guess you favor steppers for the simplicity and the above reasons?

Yes exactly for that reason! The machines I build mostly go to MEN in SHEDS type users or small business where the speed or high resolutions etc servos offer isn't needed. Reliability and simplicity are more desirable than speed for these users. Also from my point of view it also means less work and happy customers.

A good properly sized and tuned closed loop system is more than good enough for most builds and far simpler.


By the way what would be a good supplier for belts and pulleys? What supplier are you using?

I use a local supplier.

[jarjar]

I'd be interested in hearing how you get on with tuning. What controller and which method of control +/-10v or Step/Dir .?
Step/Dir is the most reliable and flexible. With analog +10V you're forced to control the servo in velocity mode.
With Step/Dir you can control it in position or velocity mode.
I'm also convinced position mode (letting the drive close all the loops internally) is the best way since the drive can do it much faster than the controller.




What hassle.? Far less wiring and setting up far as I can tell, esp if using Bus type.! . . . I've not used any yet but soon will be doing so I'll let you know.

The interface of the servo drive is a whole other topic and has nothing to do with absolute encoders. For absolute encoders you need to connect a battery and 2 extra wires to request the position data when using the standard analog/pulse-train drives. And as I said above, the controller needs to be able to read the position data for the specific servo manufacturer.

Now if you use EtherCAT drives, indeed you just need a RJ-45 cable between the drive and controller (but still need the battery). But you have 2 more problems:
- you need to actually find one of those rare EtherCat drives. Yaskawa for example doesn't sell them in the Chinese market. So you would need to go through an official reseller and pay $$$, or buy another brand: Omron / Panasonic / Mitsubishi / Delta / Estun...
- be prepared to spend weeks and write some C to make the controller talk to the drive (LinuxCNC in this case, the only one supporting EtherCAT).

[NordicCnc]

Yes exactly for that reason! The machines I build mostly go to MEN in SHEDS type users or small business where the speed or high resolutions etc servos offer isn't needed. Reliability and simplicity are more desirable than speed for these users. Also from my point of view it also means less work and happy customers.

A good properly sized and tuned closed loop system is more than good enough for most builds and far simpler.

Glad to hear this actually. Of course it is nice with speed and the non-existent noise! Who would not want 20m/min if they could have it. Anyway 10-15m/min will probably be enough for most small businesses and hobbyist (like me) as you said. I will let this thought about steppers sink for a while before I make any decision. The design is still not finalized anyway so no need to rush things.

You said you are using Lichuan 8Nm 230Vac for the build with the ATC spindle and that you use Lichuan closed loop system nowadays for your build. I checked aliexpress and they are all either 2 phase or 3 phase. So they work with a 3 phase output but uses only 230V? I did not find any 8Nm steppers, but I found this one: https://www.aliexpress.com/item/32799187522.html?spm=2114.12010612.8148356.16.5260 6833P16XMw


I use a local supplier.

Okay, I will research area more. I read somewhere that GT2 is supposed to be really good quality compared to the commonly used HTD 5.

[JAZZCNC]

Step/Dir is the most reliable and flexible. With analog +10V you're forced to control the servo in velocity mode.
With Step/Dir you can control it in position or velocity mode.
I'm also convinced position mode (letting the drive close all the loops internally) is the best way since the drive can do it much faster than the controller.

I agree about step/dir being simpler when drives handle position but not so sure that it's easier when your closing the loop back to the controller. Analog gives a little more control is my experience. But that said, my experience with servo's isn't massive and mostly with full closed-loop Analog controllers.



The interface of the servo drive is a whole other topic and has nothing to do with absolute encoders. For absolute encoders you need to connect a battery and 2 extra wires to request the position data when using the standard analog/pulse-train drives. And as I said above, the controller needs to be able to read the position data for the specific servo manufacturer.

Now if you use EtherCAT drives, indeed you just need a RJ-45 cable between the drive and controller (but still need the battery). But you have 2 more problems:
- you need to actually find one of those rare EtherCat drives. Yaskawa for example doesn't sell them in the Chinese market. So you would need to go through an official reseller and pay $$$, or buy another brand: Omron / Panasonic / Mitsubishi / Delta / Estun...
- be prepared to spend weeks and write some C to make the controller talk to the drive (LinuxCNC in this case, the only one supporting EtherCAT).

To be honest I think we are getting crossed wires a little here.! I know exactly what's involved with interfacing drives and wiring Absolute encoders etc.
My point was there's no more hassle using Absolute than incremental and there's actually less wiring required because you're not fitting home switches etc. Also if your using EtherCat or Bus type then there's even less wiring involved and less tuning involved because the controller and drives work closely together.

Also, I wasn't talking about Yaskawa Ethercat or trying to make Ethercat work with Linux or any other DIY controller. I'll be using an industrial Bus type controller with matched drives so i won't be needing to write any code etc.! . . . How long it takes to get up and running however will remain to be seen... .Lol

[JAZZCNC]

Glad to hear this actually. Of course it is nice with speed and the non-existent noise! Who would not want 20m/min if they could have it. Anyway 10-15m/min will probably be enough for most small businesses and hobbyist (like me) as you said. I will let this thought about steppers sink for a while before I make any decision. The design is still not finalized anyway so no need to rush things.

You can have 20mtr/min speeds with steppers if you choose the right drives/motors along with linear components, the 10x5 vertical machine I've been showing will happily whizz the gantry around at 30mtr/min.


You said you are using Lichuan 8Nm 230Vac for the build with the ATC spindle and that you use Lichuan closed loop system nowadays for your build. I checked aliexpress and they are all either 2 phase or 3 phase. So they work with a 3 phase output but uses only 230V? I did not find any 8Nm steppers, but I found this one: https://www.aliexpress.com/item/32799187522.html?spm=2114.12010612.8148356.16.5260 6833P16XMw

Yes, those are exactly what I'm using for Z-axis with 12Nm on X&Y. I think the heading is an error, if you look further down at the spec you'll see it says 8Nm.
This size would be way overkill for your little machine thou. The secret to the speed with steppers, in general, is the relationship between inductance and voltage, but even more so with large motors which is why Mains powered drives work best.




Okay, I will research area more. I read somewhere that GT2 is supposed to be really good quality compared to the commonly used HTD 5.

It's Got nothing to do with quality, that's down to manufacturer and spec. HTD, GT2, etc is the tooth profile. GT2 is actually a modified HTD tooth profile which supposedly offers better handling of torque and speed than HTD.

[jarjar]

My point was there's no more hassle using Absolute than incremental and there's actually less wiring required because you're not fitting home switches etc.
You still need a homing switch to setup the absolute encoder reference. The only difference is that you don't need to home on subsequent power ups.

[JAZZCNC]

You still need a homing switch to setup the absolute encoder reference. The only difference is that you don't need to home on subsequent power ups.

Been told doesn't need a physical switch the system just needs to be calibrated with encoder offset for the home position which can be obtained through software..! . . . I will find out soon enough.

[NordicCnc]

Some design progress has been made.

X-axis design

27591 27592 27593 27594

- L-shape gantry with 120x80 (heavy) Item profiles 1200mm long, with approximaterly 950mm travel length . The profiles for the L-shape are bolted together from the bottom. I plan to use aluminium plates as guides in the T-slots. Should these guides be fastened somehow or is it enough to insert them into the slots and shim away the clearance?
- High side gantry profiles with 120x80 (heavy) Item profiles, 160mm long. These are bolted to the bottom profile of the L-section, from the top of the bottom profile.
- 20mm aluminium gantry side plates, bolted to both the L-section and the high side gantry profiles. It is also bolted to the bottom plates, that will be mounted to the Y-axis bearing plates. The purpose of the bottom plates being separated from the Y-axis bearing plates, is to allow some adjustment later when squaring the gantry.
- 20mm HIWIN linear rails.
- 2010 ballscrew, 1150mm long.
- Delta 400W servo motor with a 2:1 reduction, to limit ballscrew RPM.

27595 27596 27597 27600 27598 27599 27601

Y-axis design will continue from here!

[JAZZCNC]

I would use HGW bearings rather than HGH, they offer more support.

[NordicCnc]

I would use HGW bearings rather than HGH, they offer more support.

Got it, that is easy to change. Does this apply for the Z-axis as well?

[JAZZCNC]

Got it, that is easy to change. Does this apply for the Z-axis as well?

Yes and no, Yes if you have room but I only use them on X & Y as it can make the Z a little wide.

[JAZZCNC]

You don't need or want 6 counterbored holes in the lower gantry just use the 3 furthest away from the gantry side and the center front one, you can miss the out the center bolt through the gantry side or put it in but with short threads. Reason being the bolts going through gantry side clash with bolts and it's more important you get good fastening into sides so longer bolts needed. Between them, the gantry is going now were trust me.!

Also with the ball screw spacer plates pay careful attention to fastening bolts and that you can actually access them with BK bearing and screw in place because you'll need to adjust these on the machine when setting alignment etc.

[NordicCnc]

You don't need or want 6 counterbored holes in the lower gantry just use the 3 furthest away from the gantry side and the center front one, you can miss the out the center bolt through the gantry side or put it in but with short threads. Reason being the bolts going through gantry side clash with bolts and it's more important you get good fastening into sides so longer bolts needed. Between them, the gantry is going now were trust me.!

Good notice! I have corrected that now and I will use shorter threaded bolts for the center ones (red arrow).

27604 27605


Also with the ball screw spacer plates pay careful attention to fastening bolts and that you can actually access them with BK bearing and screw in place because you'll need to adjust these on the machine when setting alignment etc.

Yes, I should be able to access those at anytime with current design.

27606 27607

I have also replaced the guide carriages with the wider version, as you suggested.

27608 27609

Thanks for your feedback, I appreciate it!

[JAZZCNC]

I just noticed the size of the pulley and I'm assuming it's to scale.? Try to keep the pulley size small as possible to keep inertia low but at the same time making sure you have enough teeth engaged so don't get jumping teeth or premature wear. If you must use a large pulley then try to use an aluminium one, again to keep inertia low.

[NordicCnc]

I just noticed the size of the pulley and I'm assuming it's to scale.? Try to keep the pulley size small as possible to keep inertia low but at the same time making sure you have enough teeth engaged so don't get jumping teeth or premature wear. If you must use a large pulley then try to use an aluminium one, again to keep inertia low.

Yes, the pulley sizes in the pictures are 20T and 40T. How many teeth need to be engaged at bare minimum? I have not yet found any supplier (I am trying to find those pulley's with taper lock) so for now during the design phase, I went with some sizes I could find on Ebay.

Skickat från min SM-A530F via Tapatalk

[jarjar]

Yes, the pulley sizes in the pictures are 20T and 40T. How many teeth need to be engaged at bare minimum? I have not yet found any supplier (I am trying to find those pulley's with taper lock) so for now during the design phase, I went with some sizes I could find on Ebay.

Skickat från min SM-A530F via Tapatalk
20T and 40T is good, it's exactly what I use on one axis. Another handy calculator (https://www.bbman.com/belt-length-calculator/).

I had to deal with the same issue: finding clamping type aluminium pulleys, either HTD or GT3. From my researches those are rare beasts.
HTD taper-lock steel pulleys are relatively common, but add way too much inertia to be considered.

I found 2 options:
- CMT pulleys (https://www.cmtco.com/product-catalogs/stock-timing-pulleys). The best but expensive, about $60 piece.
- Misumi, with separate taper-lock (https://uk.misumi-ec.com/vona2/detail/221005492623/) and pulley. Slightly less expensive, about $40.

[JAZZCNC]

Yes, the pulley sizes in the pictures are 20T and 40T. How many teeth need to be engaged at bare minimum? I have not yet found any supplier (I am trying to find those pulley's with taper lock) so for now during the design phase, I went with some sizes I could find on Ebay.

Skickat från min SM-A530F via Tapatalk

Different manufacturers quote different numbers but I think it's roughly 6 teeth before you have to start lowering the HP rating they can handle.

However, your problem will be the smaller pulley size will limit the number of teeth based on it's Boss size and Motor shaft relationship. 20T / 40T sounds about right with a 14mm motor shaft, if it's 19mm shaft then you may have problems.!

[NordicCnc]

Y-axis design

- 1600mm long, 160x80 (heavy) Item profiles will be used for the sides, onto which the linear rails are mounted. Travel length is approximately 1400mm.
- 1065mm long, 160x80 (heavy) Item profile is used for the front.
- 1065mm long, 80x80 (heavy) Item profiles will be used between the 160x80 side profiles. The base frame profiles are mounted together with front/end plates and 80x80 angles.

27634 27635 27636 27637 27639

- The plates onto which the guide carriages and the ball screw nut housing are mounted are separated from the gantry. The purpose is to allow some adjustment later when squaring the gantry. It will also separate the aligning process of the Y-axis ball screw and linear rail from the gantry squaring process. This will make the aligning easier.
- 20mm HIWIN linear rails with wide guide carriages.
- 2010 ball screws, 1500mm long, with BK15/BF15 bearings.
- Double Delta 400W servo motors with a 2:1 reduction, to limit ball screw RPM. 20T and 40T timing pulley's will be used. Preferably with taper-lock mounting.
- Same servo motor brackets, BK/BF bearing spacers and ball screw nut housing concept are used as in the X-axis. Commonality and re-usability is the goal here.

27638 27640 27641
27643

- The spindle will have a 100mm overhang at when the Y-axis is at the end of the table. The purpose is to be able to mount work pieces vertically, for dovetail machining.

27644 27642

- The table surface I am not sure about yet. Either I will install a HDPE sheet or and MDF spoil board.
- The complete machine will later be mounted to a sturdy metal bench/frame. Adjustable feet will be used.

27645 27646

Originally I intended to have the Y-axis linear rails side mounted, to make the footprint as small as possible and to protect the rails from dust and chips. I was informed that alignment is significally harder with side mounted rails, which is why the rails are now top mounted.

[JAZZCNC]

Ermm to me by arranging the profiles at the top of the 160 side pieces your wasting 80mm of height that could be used to lower the gantry height and increase stiffness or keep as it is and gain clearance.
The gain in having the motors under the bench isn't worth wasting stiffness, also moving the motors to the outside will shorten the belt length.

Also, the slightly elevated sides put the rails up higher and stop debris from hitting them directly and building up against the rail.

[NordicCnc]

Ermm to me by arranging the profiles at the top of the 160 side pieces your wasting 80mm of height that could be used to lower the gantry height and increase stiffness or keep as it is and gain clearance.
The gain in having the motors under the bench isn't worth wasting stiffness, also moving the motors to the outside will shorten the belt length.

Also, the slightly elevated sides put the rails up higher and stop debris from hitting them directly and building up against the rail.

Those are some extremely good points..

By moving the motors to the outside of 160x80 profiles, the belt lengths will be shorter and I will be able to move the 80x80 profiles down 80mm. This in turn will decrease the needed height of my gantry sides by 80mm while also protecting the Y-axis rails.

There is no doubt, I will make those changes. Maybe I could even change the sides to 200x80 and be able to make the gantry sides even lower! Now it starts to turn into a raised gantry design. The drawback will be that it becomes more difficult so access the machine table from the side, when e.g. fastening the workpiece.

Thanks JAZZ!

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

The drawback will be that it becomes more difficult so access the machine table from the side, when e.g. fastening the workpiece.


Yep there's always going to be some trade off with design like this, you just have to decide which is most important to you.!

[NordicCnc]

Yep there's always going to be some trade off with design like this, you just have to decide which is most important to you.!Yes.

If I think about it, side loading would not be too complicated with only 200mm raised gantry sides. You can still reach for all the clamps quite well and large wooden sheets can be loaded from the front.

The purpose of the high gantry sides was exactly for that reason, to simplify side loading. But now when analysing the problem closer, this will probably not have any impact to me at least.

Time to modify the design... again!

Skickat från min SM-A530F via Tapatalk

[BobTSkutter]

Hello, I like the way you're showing how your design is evolving, it's great for other people to learn from.

I notice your servo mounts for the x axis (long axis) are separate pieces. Could they be combined into the aluminium end plate you're using to hold the profiles into a 90degree corner?

Regards
Bob

[NordicCnc]

Hello, I like the way you're showing how your design is evolving, it's great for other people to learn from.

I notice your servo mounts for the x axis (long axis) are separate pieces. Could they be combined into the aluminium end plate you're using to hold the profiles into a 90degree corner?

Regards
Bob

Hello,

Thanks for your feedback! I am glad that you appreciate it.

About the servo motor mounts (the long axis is the Y-axis in my case). The servo motor mounts and the BK15/BF15 bearing spacers are exactly the same as the ones used in the X-axis (gantry). The goal here is to have as few different parts as possible (commonality and re-usability).

Yes, it would surely be possible to have the motor mounts integrated into the end plates. The design of the BK15 bearing plate would perhaps need to look a bit different and. The servo motor axis length and the end plate thickness are the constraints. Machining is needed in the servo motor mounts in order for the motor axis to align with the ball screw (the F-length can be machined longer but the motor axis length can't be changed) and also allow for timing pulley mounting. Too thick end plates or too shallow motor mount machining, will make the motor axis length to short for mounting any kind of pulley. I also like to have the motor mount bracket separated from the machine. This way I can design new brackets later if I want to change the servo motors for some reason, without tearing down the machine itself and then having to realign the axis/ball screw.

Good point anyway and I like your idea. I will have to check the feasibility of this!

[JAZZCNC]

If you are bolting into the 80x80 ends through the 160 sides then I wouldn't waste the aluminum on those endplates because they offer very little to the stiffness, There are no forces involved in those areas so they are more cosmetic than structural.!

[NordicCnc]

Quick update based last feedback collection. Many thanks to everyone contributing with their opinions!

- Raised gantry design with 200x80 item profiles. This allows for significantly lower gantry sides and will increase stiffness. It will also protect the Y-axis rails when they are further away from debris.
- 120x80 item profiles as table bracing, bolted through the 200x80 using counter bored holes. Alternatively I will use some item equivalent connectors, but I think bolting through is better in terms of rigidity.
- Motors on the outside, to allow for shorter belts and also lowering the table bracing to the bottom plane.
- End plates replaced with end caps from Item. These are only for aesthetics.

27649 27650 27651 27652 27653

Next up will be figuring out cable routing and placement of proximity sensors. Is there any significant difference in terms of tolerances when comparing mechanical and optical sensors?

[Sterob]

Its coming along very nicely Nordic.
What software are you using?

[NordicCnc]

Its coming along very nicely Nordic.
What software are you using?Thanks Sterob! I am using Siemens NX for CAD. I would probably use Fusion but I am lucky to have a work computer with a NX license. In the future I plan to move to Fusion, to ensure that whatever happens with work, I will still have CAD access, lol.

Skickat från min SM-A530F via Tapatalk

[NordicCnc]

Home & limit switches

Z-axis home switch at the top end of the travel. Soft limit will be used for the lower end of the travel.
27667

X-axis home switch & limit switch on the moving part. Adjustable bolts in the t-slot in the lower right and upper left of the picture. The lower right will be used for home & limit switch and the upper left will be used for limit switch.
27668

Y-axis home and limit switch on the moving part. Adjustable bolts in the t-slot on the left side of the base frame. Upper left one is used for limit switch and lower right one is used for home & limit switch. On the opposite side (right side) is the second home switch. The purpose with 2 home switches is to be auto to auto square the gantry when homing the axis.
27669 27670

Alternatively I could have all the proximity sensors fixed and not on the moving parts. Then I could use soft limits for all axis's in the opposite end of the homing position, same way as with the Z-axis. Would that be a bad idea? It would be less cable routing of the sensors in the cable drag chains,

[AndyUK]

I'd recommend having the homing positions on the dual screw axis finely adjustable - like how your Z position can be adjusted by turning that bolt. At the moment you'll have to loosen that bolt and t-nut and just hope you can push it accurately back and forth. It'll help when trying to square the gantry, where you need (ideally) sub-mm changes to switch triggering position and I wish I'd thought of that earlier. I've had to do quite a bit of back and forth trying to get the gantry squared.

You can work around it in software (I can't remember what controller software combo you're planning to use - but generally speaking its possible), but easier done once in hardware.

[Clive S]

Z-axis home switch at the top end of the travel. Soft limit will be used for the lower end of the travel.

I personally would have the switch fixed and have the target moving

[NordicCnc]

I personally would have the switch fixed and have the target moving

Yes, that is how the Z-axis switch is planned in the picture.

[NordicCnc]

I'd recommend having the homing positions on the dual screw axis finely adjustable - like how your Z position can be adjusted by turning that bolt. At the moment you'll have to loosen that bolt and t-nut and just hope you can push it accurately back and forth. It'll help when trying to square the gantry, where you need (ideally) sub-mm changes to switch triggering position and I wish I'd thought of that earlier. I've had to do quite a bit of back and forth trying to get the gantry squared.

You can work around it in software (I can't remember what controller software combo you're planning to use - but generally speaking its possible), but easier done once in hardware.

Extremely good point, thanks for letting me know. I will make some changes to the design to allow for fine adjustment.

[BobTSkutter]

The proximity sensor at the top of the Z-Axis appears to be very close to the grease nipple on the bearing block. Could the grease nipple cause interference with the sensor?
Adjustable bolts looks like a clever way to adjust the trigger point but are the bolt heads big enough for the gantry limit switches to "see". The sensor is round and I think you're using set screws which appear to be the same size as the sensor with the hex indent in the middle. It looks to me as though the hex indent is exactly where you want the sensor to trigger. Would using a conventional hex head bolt instead of a set screw be better?
Regards
bob

[Clive S]

Yes, that is how the Z-axis switch is planned in the picture.

Yes sorry I was looking at it wrong.

[NordicCnc]

The proximity sensor at the top of the Z-Axis appears to be very close to the grease nipple on the bearing block. Could the grease nipple cause interference with the sensor?
Adjustable bolts looks like a clever way to adjust the trigger point but are the bolt heads big enough for the gantry limit switches to "see". The sensor is round and I think you've set screws which appear to be the same size as the sensor with the hex indent in the middle. It looks to me as though the hex indent is exactly where you want the sensor to trigger. Would using a conventional hex head bolt instead of a set screw be better?
Regards
bob

Good catch, but this is just a 3D error that I didn't correct yet. I plan to have all grease nipples facing inwards and with a common lubricating hose and nipple externally mounted.

As for the screw heads, I am not sure really. This is something that I can easily fix later if it turns out it is not working as planned. I have access to both lathe and mill so manufacturing a better target, e.g. with a flat surface for the sensor will be possible.

[JAZZCNC]

Extremely good point, thanks for letting me know. I will make some changes to the design to allow for fine adjustment.

Don't waste your time, the way you have it now will work fine, that's exactly as I do them. It's simple and works. If I was to make a suggestion it would be to move the switch onto the side, fastened to the ballnut bracket and use the top slot for targets. How you have it now leaves the switch wide open to debris when cutting and if you cut steel or aluminium then could get false triggers when blasted with chips. Also the crap from cutting aluminium with lube can build up on the sensor.



As for the screw heads, I am not sure really. This is something that I can easily fix later if it turns out it is not working as planned. I have access to both lathe and mill so manufacturing a better target, e.g. with a flat surface for the sensor will be possible.

Again exactly as I do them and they work fine, the round head doesn't cause any issues. The only suggestion I'd make is to stiffen up the bracket. The one I use are 3D printed, You can just see it between the mass of pipes and wires in this pic.

27671

[Kitwn]

For users of LiunuxCNC with the option to auto-square the gantry it is essential that the switches cannot overshoot their targets during deceleration and reset themselves before the whole sequence is finished. In this case you may need a larger target, especially with a large and heavy gantry.

[NordicCnc]

Don't waste your time, the way you have it now will work fine, that's exactly as I do them. It's simple and works. If I was to make a suggestion it would be to move the switch onto the side, fastened to the ballnut bracket and use the top slot for targets. How you have it now leaves the switch wide open to debris when cutting and if you cut steel or aluminium then could get false triggers when blasted with chips. Also the crap from cutting aluminium with lube can build up on the sensor.

Thanks for the suggestion. I will check the feasibility of moving the sensor to the ballnut bracket! I was also thinking that for fine adjustment with this simple setup, one could use a dial indicator when adjusting the target.


Again exactly as I do them and they work fine, the round head doesn't cause any issues. The only suggestion I'd make is to stiffen up the bracket. The one I use are 3D printed, You can just see it between the mass of pipes and wires in this pic.

27671

Looks good.I think I could make a stiffer bracket out of aluminium without any problems. That X-axis ballnut bracket looks interesting. If I am not mistaken I would think you are using a rotating ballnut. Any chance I could get a closer look at that? :eagerness:

[NordicCnc]

For users of LiunuxCNC with the option to auto-square the gantry it is essential that the switches cannot overshoot their targets during deceleration and reset themselves before the whole sequence is finished. In this case you may need a larger target, especially with a large and heavy gantry.

I see, I didn't know that! Luckily I will be using Mach3! If I still need a bigger target for some reason then I can lathe a new target.

[Kitwn]

I see, I didn't know that! Luckily I will be using Mach3! If I still need a bigger target for some reason then I can lathe a new target.

The devil is always in the detail.

[JAZZCNC]

Thanks for the suggestion. I will check the feasibility of moving the sensor to the ballnut bracket! I was also thinking that for fine adjustment with this simple setup, one could use a dial indicator when adjusting the target.

You won't need to adjust it to that degree. I don't use Linux CNC but believe it just works like any other gantry squaring routine in that it sees the switch and you tell it in software how far from the switch you want it to move so it squares the gantry. The switch is just a trigger, the closer you get to square the better because your not loading the bearings, etc, but you don't need it to be to the micron.!

The size of the target shouldn't make a difference if the switch distance is set correctly. What a larger target does allow is more room for clearance, basically, it allows more room over shoot, which you don't want anyway because of the twisting gantry.
Also if you are using Servo's then it's all immaterial because you'd use Home-to-Index, so again you just using the switch like a trigger, then the servo will Index to the Encoder.




Looks good.I think I could make a stiffer bracket out of aluminium without any problems. That X-axis ballnut bracket looks interesting. If I am not mistaken I would think you are using a rotating ballnut. Any chance I could get a closer look at that? :eagerness:

Yes I used to make them out of aluminium but I prefer to 3D print them now because if anything hits them the bracket snaps off and saves the switch. The bracket is strong enough so it doesn't flex causing false triggers etc but yet weak enough to snap if thumped with somthing heavy.

Regards the Rotating Ballnuts then yes your right. All Axis except Z-axis are using rotating nuts but I cannot take any claim for the design, Jonathan designed them and kindly allowed me to use.( I actually bought some of the parts from him and made the others) If you search the forum you'll find a thread about them.

27673

[joe.ninety]

@NordicCnc

Can I ask you a couple of questions regarding your CAD drawings that you're posting here...

1. Were they done using Fusion360?

2. Have you modeled all the elements yourself, like the carriages and rails etc, or have you just found existing DXF files and imported them?

3. If imports then could you share the source?

I'm asking because I'm using F360 and am having real problems with the dimensions of my imported DXFs. I've been using DXFs from the HiWin website, but whenever I import them into a design the dimensions are all out. The imported DXF are never the same dimensions as they are on the 2D drawings, so aren't really useful.

All I'm doing at the present is just using solid shapes to represent the items I want. So a HGH20CA carriage for example is a block 77.5 x 44 x 25.4 with a channel running through it to represent where the rail goes.

Looking at your design drawings, you seem to have everything fine?

Cheers

[JAZZCNC]

I don't use F360 but would imagine it works just like Solid works where it will allow you to import Step or Stl models which if Hi-win don't supply a direct F360 format like they do for SW you could use. DXF is mostly a 2D format and you want a 3D model.

Also for Component models I use Traceparts
https://www.traceparts.com/en

[NordicCnc]

@NordicCnc

Can I ask you a couple of questions regarding your CAD drawings that you're posting here...

1. Were they done using Fusion360?

2. Have you modeled all the elements yourself, like the carriages and rails etc, or have you just found existing DXF files and imported them?

3. If imports then could you share the source?

I'm asking because I'm using F360 and am having real problems with the dimensions of my imported DXFs. I've been using DXFs from the HiWin website, but whenever I import them into a design the dimensions are all out. The imported DXF are never the same dimensions as they are on the 2D drawings, so aren't really useful.

All I'm doing at the present is just using solid shapes to represent the items I want. So a HGH20CA carriage for example is a block 77.5 x 44 x 25.4 with a channel running through it to represent where the rail goes.

Looking at your design drawings, you seem to have everything fine?

Cheers

I use Siemens NX which I am lucky enough to have a work related license for. If I wouldn't have that I would probably be using Fusion 360.

The linear rails and blocks are downloaded as STEP files from the hiwin configurator: https://motioncontrolsystems.hiwin.com/category/linear-guideways

The ballscrews I have modeled myself without any threads. This is because I want to model the ballscrew F-lengths myself, and I have no need for the thread visualization anyway

I also use traceparts and grabcad for other models like servo motors. I like the STEP format best, no issues so far!

Skickat från min SM-A530F via Tapatalk

[joe.ninety]

@NordicCnc

Thanks for that tip re the Step files. I went back in and re-downloaded all the files as 'Step' and my problems went away.

The only problem with the rails is that you need to specify a length before the website will create the 3D file for you. Think I'll knock up my own version and just keep that as a 2D drawing that can be extruded to the length required. Although it will obviously lack the actual positioning of the mounting holes.

Cheers

[joe.ninety]

Also for Component models I use Traceparts
https://www.traceparts.com/en

Thanks for that tip. Looks like an interesting site.

Cheers

[JAZZCNC]

@NordicCnc

Thanks for that tip re the Step files. I went back in and re-downloaded all the files as 'Step' and my problems went away.

The only problem with the rails is that you need to specify a length before the website will create the 3D file for you.

Just get a few different lengths and lay them on top of each other, aligning with each bolt hole to keep the spacing. It's always better to work accurately and mate the bearings to the surfaces of the actualy rail they would run on, this way you don't get accumulative errors building up as the model grows.

[NordicCnc]

Short update:

No progress for the last 2 weeks on the router design. I have been really busy with other projects and commitments. I had already bough 180W servos but those I already returned for some 400W servos with integrated drives. Those servos are now fitted on a Hitachi Seiki lathe (huge lathe with 400mm 4-jaw chuck) with a 1:1 ratio on the X-axis and a 2:1 ratio on the Z-axis. Ballscrews are 2005. WOW, I can now run the lathe at 10m/min without any problems. Ballscrew whip is no problem. I am very happy with that upgrade.

Anyway back to the router. Since I will be using servo motors with a 3000rpm rated speed, I thought I could take full advantage of that. I plan to do the following:

Swap the 2010 ballscrews to 2020 on the Y-axis and X-axis.
Swap the 1605 ballsrew to 1610 on the Z-axis
Change gearing ratio from 2:1 to 3:1 on all axis. This will give me more torque, with lower ballsrew rpm while increasing the ballscrew pitch.
Add belt tensioners to get more teeth engaged because of the 3:1 gearing.

This will limit the ballscrew speed to 1000rpm, and increase the Y- and X-axis maximum feed rate to 20m/min. Z-axis feed rate will stay at 10m/min. I feel like this is a better idea than trying to spin the ballscrews at 1500rpm with the 2:1 ratio. I want to take full advantage of the servo motors RPM.

On another note I was planning to use 400W (maybe even 750W) Delta servo motors, with dual motors on the Y-axis. I need to do some calculations if it would be a good idea to go for 750W. Anyway I saw on YouTube that there is a auto tuning function available in the software but I am not sure how to tune the dual motors on the Y-axis.


What do you guys think about the increased ballscrew pitch, 3:1 gearing and belt tensioners?
Is it possible to auto tune 2 motors at the same time?
Perhaps I could change the design and use only 1 servo motor on the Y-axis with belt drive connected to both sides. This in turn will not allow for squaring the gantry with motion controller but I think i can get it square any way.
How wide can the gantry be without having to use dual slaved Y-axis motors?

[NordicCnc]

So machine design is now what I consider 99% completed so that I can start ordering parts. Since none replied on the ballscrew change I've decided to do the following:

- Keep the 2:1 ratio
- Change Y- and X-axis ballscrews to 2020
- Keep the Z-axis ballscrews as 1605
- No belt tensioners
- Double Y-axis motors will be used


I will put together a shopping list for all components so that I can get as much as possible in one order without missing anything. The final piece of design was to find a good location for the Y-axis home/limit switch. Cable chains I will fit later and wont add those to the 3D at all. Please ignore that the screw is not in the center of the limit switch. I will lathe and mill some custom screws but I could not care to design those yet. :victorious:

27849 27850 27851

Final assembly!

27852 27853 27854 27855 27856

Some things still bugging me is the 120x80 for the gantry. I hope that those will not be too weak for machining aluminium with the ATC spindle. I am still consider changing to 160x80. JAZZ, if you read this, I remember that you wrote somewhere that for larger machines you use the 160x80 for the L-shape gantry. Since my gantry is 1200mm wide and I am using an ATC spindle, would it be wise to go for 160x80 anyway?

[joe.ninety]

Can I just ask what the benefit is of using blocks of profile to sit the gantry profile on, as indicated in the picture?

27859

It seems because of this you were forced to place the bottom linear rail on the front, thereby pushing the spindle further away from the gantry. I'd always though that the idea is to try and keep the centre of gravity of the spindle as close as possible to the bearing blocks the gantry rides on.

This is not in anyway meant to be a criticism, but just me trying to understand if this design means there are benefits to it that outweigh the spindle being pushed further outward.

Cheers

[mekanik]

As a rule you ideally want the spindle IN the same envelope as your X axis bearings, i would make the distance between these bearings further apart.

[NordicCnc]

Can I just ask what the benefit is of using blocks of profile to sit the gantry profile on, as indicated in the picture?

27859

It seems because of this you were forced to place the bottom linear rail on the front, thereby pushing the spindle further away from the gantry. I'd always though that the idea is to try and keep the centre of gravity of the spindle as close as possible to the bearing blocks the gantry rides on.

This is not in anyway meant to be a criticism, but just me trying to understand if this design means there are benefits to it that outweigh the spindle being pushed further outward.

CheersThose block are used to increase the gantry clearance. I want to build machine out of aluminium profiles and the 200x80 profile for the Y-axis sides are not high enough. Also by placing the linear rail on the front of the X-axis will allow me to use the work area fully, without increasing the footprint. If I were to put the linear rail on the bottom of the gantry profile I would lose approximately 160mm in the X-axis travel.

You are right about the overhang, but I would claim that it is not that critical. As long as the Z-axis and gantry is stiff enough, it should not be any problem at all. If you check other builds, the overhang is similar!

Thanks for your comment.

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

Ok well I'll answer this question seen as he's mostly following my design.!

It always makes me smile when I see comments like these because while the physics shows that what you are saying is the ideal location the reality is that it makes no difference to the machine and how it works.

To prove this point I'm going to give you a challenge.! . . . Industry demands the highest cut quality and performance, so logic dictates that they would follow the physics and optimum location very closely.?

So go find me a machine from the major manufacturers that place the spindle smack between the bearings.! . . . If you find one then I guarantee you'll have seen ten before it that don't.!! . .. In fact, you'll be lucky if you find any with spindle inside the bearings.!

You all need to stop worrying about the physics and virtual world so much and get building so you can realize just how little if at all, these affect a real-world machine.!

Exhibit: A
27860





Can I just ask what the benefit is of using blocks of profile to sit the gantry profile on, as indicated in the picture?

It seems because of this you were forced to place the bottom linear rail on the front, thereby pushing the spindle further away from the gantry. I'd always though that the idea is to try and keep the centre of gravity of the spindle as close as possible to the bearing blocks the gantry rides on.

This is not in anyway meant to be a criticism, but just me trying to understand if this design means there are benefits to it that outweigh the spindle being pushed further outward.

Cheers

To answer you directly Joe, then to place the gantry further back to bring spindle into line with bearings would actually weaken the machine not make it stronger. To do what you suggest without getting into complex gantry side designs means using plates for gantry sides that can flex side to side and introduce vibrations at the tool.
I can tell you from building many different designs of router that the design he's using is the best possibly way to build a router without getting silly about.
Mike is correct in that having longer distance between the bearings is good but thats a trade off between travel and foot print of the machine. In the grand scheme again it makes very little difference.


As a rule you ideally want the spindle IN the same envelope as your X axis bearings, i would make the distance between these bearings further apart.

[NordicCnc]

As a rule you ideally want the spindle IN the same envelope as your X axis bearings, i would make the distance between these bearings further apart.Can you explain what you mean? I am not sure I follow.

Edit: I think you meant the spacing between the Y-axis bearings? Doing this would also decrease the Y-axis travel. I also want to be able to machine dovetails at the end of the table, so overtravel is a must!

Skickat från min SM-A530F via Tapatalk

[mekanik]

I was working on the assumption X was your Base and Y your Gantry, Dean is the go to man for advice and if he says it's OK just crack on.
Regards
Mike

[NordicCnc]

I was working on the assumption X was your Base and Y your Gantry, Dean is the go to man for advice and if he says it's OK just crack on.
Regards
Mike

I see! My design is defined with Y-axis on the base frame, X-axis on the gantry and then Z-axis on the spindle. Anyway thanks for checking in!

[NordicCnc]

I have started thinking about precision tools that I will need for the build. I have a post already on the forum where JAZZ explained the tolerances very well: http://www.mycncuk.com/threads/13435-Straight-edge-and-machinists-square

I already own a good dial indicator so now I need a set of straight edge and/or machinists square.


Do I need both a straight edge and a machinists square, or would just a machinists square be sufficient?
What straight edge length is required? Linear rails are 1500mm (Y, base frame), 1200mm (X, gantry) and 500mm (Z).
What machinists square dimension (height x width) is required? Linear rails are 1500mm (Y, base frame), 1200mm (X, gantry) and 500mm (Z).


Helios Preisser straight edges:

Steel square straight edge, cross section 40x8mm, length 1000mm, DIN874/2, @44,40€ https://online.helios-preisser.com/procat/Catalog.jsp;jsessionid=F5D0C341B8D502C02F5F1897E62 703AB?mode=details&groupId=(%5B@IMID:Node:group5676770086736%5D)&productId=(%5B@IMID:Object:product40041050432013%5 D)
Stainless steel square straight edge,cross section 40x8, 1000mm, DIN874/2, @98,00€ https://online.helios-preisser.com/procat/Catalog.jsp;jsessionid=F5D0C341B8D502C02F5F1897E62 703AB?mode=details&groupId=(%5B@IMID:Node:group5940547402332%5D)&productId=(%5B@IMID:Object:product4004973167720%5D )


Helios Preisser machinist squares:

Steel square with base, cross section 30x5mm, length of beams 1000x500mm, DIN unknown?, @57,00€ https://online.helios-preisser.com/procat/Catalog.jsp;jsessionid=3145E7C3183578521A07D9D96C8 CA012?mode=details&groupId=(%5B@IMID:Node:group31164210657298%5D)&productId=(%5B@IMID:Object:product39801027812493%5 D)
Steel square with base, cross section 50x10mm, length of beams 1000x500mm, DIN 875/2, @165,00€ https://online.helios-preisser.com/procat/Catalog.jsp;jsessionid=3145E7C3183578521A07D9D96C8 CA012?mode=details&groupId=(%5B@IMID:Node:group107921060942152%5D)&productId=(%5B@IMID:Object:product4004525824975%5D )
Steel square without base, cross section 30x5mm, length of beams 1000x500mm, DIN unknown?, @24,50€ https://online.helios-preisser.com/procat/Catalog.jsp;jsessionid=3145E7C3183578521A07D9D96C8 CA012?mode=details&groupId=(%5B@IMID:Node:group298721037008617%5D)&productId=(%5B@IMID:Object:product3980938670859%5D )
Steel square without base, cross section 50x10mm, length of beams 1000x500mm, DIN 875/2, @135,00€ https://online.helios-preisser.com/procat/Catalog.jsp;jsessionid=3145E7C3183578521A07D9D96C8 CA012?mode=details&groupId=(%5B@IMID:Node:group1079239842683%5D)&productId=(%5B@IMID:Object:product400456342882%5D)


So many options, but I think that the DIN874/2 and DIN875/2 would be sufficient. Then it is only about with or without base and the length?

[joe.ninety]

Those block are used to increase the gantry clearance. I want to build machine out of aluminium profiles and the 200x80 profile for the Y-axis sides are not high enough.

Ok, in that context I understand. It's just in most designs I've seen there is simply a gap there... The desired height achieved by where on the gantry sides you bolt the cross-section (the main bit of profile that forms the gantry).
So in your design, the gantry cross section is supported by being both bolted to the side plates and also by sitting on other pieces of extrusion? Are the blocks underneath connected to the cross-section or does the cross-section simply sit atop them?

Cheers

[joe.ninety]

Ok well I'll answer this question seen as he's mostly following my design.!

It always makes me smile when I see comments like these because while the physics shows that what you are saying is the ideal location the reality is that it makes no difference to the machine and how it works.

To prove this point I'm going to give you a challenge.! . . . Industry demands the highest cut quality and performance, so logic dictates that they would follow the physics and optimum location very closely.?

So go find me a machine from the major manufacturers that place the spindle smack between the bearings.! . . . If you find one then I guarantee you'll have seen ten before it that don't.!! . .. In fact, you'll be lucky if you find any with spindle inside the bearings.!

You all need to stop worrying about the physics and virtual world so much and get building so you can realize just how little if at all, these affect a real-world machine.!


To answer you directly Joe, then to place the gantry further back to bring spindle into line with bearings would actually weaken the machine not make it stronger. To do what you suggest without getting into complex gantry side designs means using plates for gantry sides that can flex side to side and introduce vibrations at the tool.
I can tell you from building many different designs of router that the design he's using is the best possibly way to build a router without getting silly about.
Mike is correct in that having longer distance between the bearings is good but thats a trade off between travel and foot print of the machine. In the grand scheme again it makes very little difference.

Is it any wonder that us mere novices get totally confused by all this, as there seems to be loads of conflicting advice floating around in the ether and I'm not talking about just on here.

So there I was working on some supposed designs... Linear rails on back of spindle mount, carriages on z axis, and trying to have a design that keeps the spindle as close to the bearings as possible... Now I find out all that is unnecessary, or not as necessary as I'd been led to believe. But as you've given this design the seal of approval then I guess I'll just try and make mine more like this one, only smaller.

Cheers

[NordicCnc]

Ok, in that context I understand. It's just in most designs I've seen there is simply a gap there... The desired height achieved by where on the gantry sides you bolt the cross-section (the main bit of profile that forms the gantry).
So in your design, the gantry cross section is supported by being both bolted to the side plates and also by sitting on other pieces of extrusion? Are the blocks underneath connected to the cross-section or does the cross-section simply sit atop them?

Cheers

Yes, L-shape bolted together to the gantry sides. Screw clearance holes in the lower profile, threading into t-nuts in the upper profile. I also intend to use aluminium plates and shim them in the t-slot, to act as guides/stopper so that the gantry profiles doesn't move. The blocks underneath the L-shape also have holes and mounts to the lower profile.

Actually I am extremely glad that you commented on the design. It made me realize that I could further increase the stiffness while not compromising gantry clearance significantly!

I've now removed the 60mm high profiles from the gantry sides. Instead I've replaced the 20mm thick blocks underneath, with 40mm blocks. These are going to be machined so that the Z-axis can pass above the blocks. This effectively decreased height the gantry sides by 40mm, less parts, increase simplicity & accuracy while making the gantry stiffer.

27883 27884 27885

However this means that I need to modify the table to compensate for the lost 40mm clearance.

In the current base frame, I've used 120x80 profiles bolted into the sides, consisting of 200x80 profiles. I have 2 alternatives I think:
1. Replace the 120x80 with 120x60 to gain 20mm clearance. Now I've only lost 20mm clearance.
2. Move the 120x80 underneath the 200x80 profiles (+40mm clearance) and bolt from the bottom. Add 20mm item profiles (wide version) as a "T-slot bed" (+/-0mm clearance).

I think I will go with option 2, which will make the bed stiffer also! The budget allows for it anyway.

[AndyUK]

Might be worth considering how you'll access some of these bolts.

Take for instance the front bearings on the gantry - does the rail being longer than needed foul adjusting those bolts? Does the servo on the back prevent access to the rear ones?

If you're trying to adjust the machine for gantry squareness, and you have to remove the servo between adjustments, that is going to get old fast.

I'm sure someone will be along shortly to tell you not to worry and to ignore this advice it'll all be fine... but hey, worth a thought.

[NordicCnc]

Might be worth considering how you'll access some of these bolts.

Take for instance the front bearings on the gantry - does the rail being longer than needed foul adjusting those bolts?

The rails are not longer than needed. The X-axis can travel all the way to the end actually. The clearance between the rail and the block is 30mm, so I am quite sure there will be enough room. I need to double check this.


Does the servo on the back prevent access to the rear ones?

If you're trying to adjust the machine for gantry squareness, and you have to remove the servo between adjustments, that is going to get old fast.

I will have to look into this as well, hehe.

Thank you for the heads up!

[Scustom]

Hi nordic
Why dont you consider doing bed insted of 200x80 similar what robocnc on youtube did?

[NordicCnc]

Hi nordic
Why dont you consider doing bed insted of 200x80 similar what robocnc on youtube did?Hello! I am using 200x80 because it is very stiff. I could of course do it as you describe but I would be sacrificing stiffness. The base frame stiffness is key to a good machine! I also want to keep the part count as low as possible.

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

Is it any wonder that us mere novices get totally confused by all this, as there seems to be loads of conflicting advice floating around in the ether and I'm not talking about just on here.

So there I was working on some supposed designs... Linear rails on back of spindle mount, carriages on z axis, and trying to have a design that keeps the spindle as close to the bearings as possible... Now I find out all that is unnecessary, or not as necessary as I'd been led to believe. But as you've given this design the seal of approval then I guess I'll just try and make mine more like this one, only smaller.

Cheers

Joe It's not really fair to hi-jack Nordics thread so let's do it on yours or start another asking this question if you want some guidance and I'll gladly come along and explain the differences etc.

[joe.ninety]

Joe It's not really fair to hi-jack Nordics thread so let's do it on yours or start another asking this question if you want some guidance and I'll gladly come along and explain the differences etc.

Sorry, that was never my intention. But in my defense, that was more of a statement of a bit of frustration than a question relating to my own design.

All my questions on this thread have, I think, been aimed at Nordic and his design. I will however be more mindful in the future.

Cheers

[NordicCnc]

Sorry, that was never my intention. But in my defense, that was more of a statement of a bit of frustration than a question relating to my own design.

All my questions on this thread have, I think, been aimed at Nordic and his design. I will however be more mindful in the future.

Cheers

Hey Joe,

Please don't worry about that. You are free to continue asking any questions you want in my build log. JAZZ was only being polite, since I have not mentioned anything about it. But now I have, so feel free to ask anything!

Also thanks to your questions about the bearing blocks and the standing profiles, I've come up with a much better design I think. I will post more about that later.

[NordicCnc]

Okay I now have a dilemma.

The price difference between 400W and 750W Delta servo motors are only about 100€ each, so total of 400€ price difference. This means that I would get 2 times the torque if deciding to go for 750W servos. Anyway are there any possible drawbacks with these huge servos other than more space requirement and somewhat more expensive? Larger pulley diameters would be needed due to 19mm shafts instead of 14mm shafts.

Also would 400W servos have enough torque to spin 2020 ballscrew with a 2:1 ratio? That would give me a speed of 30m/min. Same case with 750W servos, but those would for sure have enough if not too much available torque.

Also do I need that kind of speed for anything? Machine working area is approx 1200x1000mm anyway. Perhaps 15m/min (achieved with 2010 ballscrews) would be more than enough.

I need help. Please advise! :witless::witless:

Some info that might be useful:

Z-axis weight: 27kg (1605 ballscrew, 2:1 ratio)
X-axis weight: 60kg (2020 or 2010 ballscrew, 2:1 ratio)
Y-axis weight (gantry): 120kg (dual motor drive, 2020 or 2010 ballscrew, 2:1 ratio)

[JAZZCNC]

400W with a 2:1 on 10mm pitch will be fine. The ratio will double the torque and Like you say you won't need more than 15mtr/min, in fact, you could go 3:1 or 2.5:1 and still have plenty of speed.

Save the money and put the 400 into better ball screws.!

[NordicCnc]

400W with a 2:1 on 10mm pitch will be fine. The ratio will double the torque and Like you say you won't need more than 15mtr/min, in fact, you could go 3:1 or 2.5:1 and still have plenty of speed.

Save the money and put the 400 into better ball screws.!Thanks JAZZ. I am now waiting for a quote from DMM Technology for 400W servo motors. Their servo motors looks really good and the manual/software as well!

The ballscrews that I was planning to get arw C7 rolled ballscrews from BSTmotion. I guess the next step would be either:
1. C7 rolled, with double ballnut
2. C5 ground, with single ballnut
3. C5 ground, with double ballnut

I think that Fred doesn't offer C5 ballscrews so I am mot sure where to buy those from!

Edit: I looked into it and I will probable have a go at double ballnuts. I may need to increase the ballscrew lengths to maintain the same working area, but it will be well worth!

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

Thanks JAZZ. I am now waiting for a quote from DMM Technology for 400W servo motors. Their servo motors looks really good and the manual/software as well!

The ballscrews that I was planning to get arw C7 rolled ballscrews from BSTmotion. I guess the next step would be either:
1. C7 rolled, with double ballnut
2. C5 ground, with single ballnut
3. C5 ground, with double ballnut

I think that Fred doesn't offer C5 ballscrews so I am mot sure where to buy those from!

Edit: I looked into it and I will probable have a go at double ballnuts. I may need to increase the ballscrew lengths to maintain the same working area, but it will be well worth!

Skickat från min SM-A530F via Tapatalk

Never used DMM but I've heard good and bad things so who knows, you just have to try then make up your own mind.!

Ask fred about C5 because I've had C5 from him before and I think he even has or can get Ground screws in shorter lengths. Unless they have stopped supplying them.?

You have to stop at some point because you can go on for ever choosing better quality or performance parts, each upgrade usually affects some other area which then needs upgrading to match else it's pointless. The better machine isn't the one with a few high spec parts, the one that uses lower spec parts that are evenly matched will be the more balanced machine and overall usually work best. . . . Remember machine is only as good as it's weakest link, upgrade one area the rest becomes weak.!!

[NordicCnc]

Never used DMM but I've heard good and bad things so who knows, you just have to try then make up your own mind.!

Ask fred about C5 because I've had C5 from him before and I think he even has or can get Ground screws in shorter lengths. Unless they have stopped supplying them.?

You have to stop at some point because you can go on for ever choosing better quality or performance parts, each upgrade usually affects some other area which then needs upgrading to match else it's pointless. The better machine isn't the one with a few high spec parts, the one that uses lower spec parts that are evenly matched will be the more balanced machine and overall usually work best. . . . Remember machine is only as good as it's weakest link, upgrade one area the rest becomes weak.!!

Can you specify some of the bad things? But, yes I will most lilely go for these anyway. Many have used them with great success. The manual and software is really good (not chinglish!) There must be bad things about the Delta servos also.

Fred can supply C5 ballscrews but it will be TBI brand and expensive (about 2.5x more expensive)!

This I have noticed, haha. I have saved every iteration of the machine design. I am up to v11 now..



Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

Can you specify some of the bad things? But, yes I will most lilely go for these anyway. Many have used them with great success. The manual and software is really good (not chinglish!) There must be bad things about the Delta servos also.

It's been a while but it mostly related to the encoder's output from the drives when using with a controller that can fully close the loop. There were some timing issues that caused problems between controller and drive. But that was some time ago so they will have probably resolved that issue by now but I've seen all kinds of little things and conflicts on various forums etc.
But like you say Delta and other manufacturers do have there own issues as well and it's how they are resolved that matters most, DMM are very good in this respect I believe,

[NordicCnc]

Does anyone know what size cable drag chain is preferred? I am thinking about 25x77mm: https://www.aliexpress.com/item/530003796.html?spm=2114.12010612.8148356.1.264066e c24hVi7

[AndyUK]

Does anyone know what size cable drag chain is preferred? I am thinking about 25x77mm: https://www.aliexpress.com/item/530003796.html?spm=2114.12010612.8148356.1.264066e c24hVi7

Depends on the amount and size of your cables! I measured the diameter of each, approximated them as square boxes, and then calculated the internal area needed. Then you want to fill to about 60%.

[NordicCnc]

Short update to show the final design. NO MORE CHANGES! :nevreness: Started ordering parts today, I am excited.

27991 27992

Not sure if the frame and/or table needs more bracing. Please give me some advice!

[NordicCnc]

I know I said no more changes but I was wrong. I modified the frame and bed quite a bit. All parts are now ordered and the first parts will be arriving in the beginning of next week. The bed will be adjustable in height to allow for higher work parts if ever needed. This is a safety measure so that I don't regret the design. Table will be adjusted with scissor lifts and should be quite easy to realign. I will also add some triangle bracing to the frame, which are not shown in the pictures. These will be made of steel.

Oh and by the way, budget of 8000€ was not enough, which I expected anyway. It will be closer to 10000€. I am still happy!

Also JAZZ was right in the beginning of the build log about the total weight. It will be over 400kg once everything is assembled.

https://uploads.tapatalk-cdn.com/20200502/06bc0f0e598cfe9516cb1adebfa80a76.jpghttps://uploads.tapatalk-cdn.com/20200502/287ff69376dd3f437a4943f3f42b55a8.jpghttps://uploads.tapatalk-cdn.com/20200502/e7908eaa3599b4d643e7cec64a88e1f6.jpg

Skickat från min SM-A530F via Tapatalk

[NordicCnc]

What kind of signal wires are recommended between the controller and AC servo drives? Would ethernet cables be good? Also do they need to be shielded?

Cat5 or Cat6?

Skickat från min SM-A530F via Tapatalk

[pippin88]

Most AC servos have manufacturer made cables between the motor and drive.

Are you taking about within the control box? Or outside?

[NordicCnc]

Most AC servos have manufacturer made cables between the motor and drive.

Are you taking about within the control box? Or outside?Yes, within the control box between the ESS breakout board and the servo drives.

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

What kind of signal wires are recommended between the controller and AC servo drives? Would ethernet cables be good? Also do they need to be shielded?

Cat5 or Cat6?

Skickat från min SM-A530F via Tapatalk

Cat5 or 6 will work provided it's shielded and you ground one end of the shield at the star point.

[Gustave]

Hi,

from #111 "Perhaps I could change the design and use only 1 servo motor on the Y-axis with belt drive connected to both sides. This in turn will not allow for squaring the gantry with motion controller but I think i can get it square any way."

I am face to the same question and first thought is doubling motor + driver + home switch is a costly option compare to 1 motor + belt drive... Have you develop the track of 1 motor for the both Y with belt drive ? Could you advice buildLog of thread that discuss this point ?
Regards

[NordicCnc]

Hi,

from #111 "Perhaps I could change the design and use only 1 servo motor on the Y-axis with belt drive connected to both sides. This in turn will not allow for squaring the gantry with motion controller but I think i can get it square any way."

I am face to the same question and first thought is doubling motor + driver + home switch is a costly option compare to 1 motor + belt drive... Have you develop the track of 1 motor for the both Y with belt drive ? Could you advice buildLog of thread that discuss this point ?
RegardsNo I didn't go for that. I did alrady buy the components fpr dual drive Y-axis and they are being shipped currently. I never looked into the design either. I bet it could work, but perhaps it is not ideal in my case for a 1200mm gantry!

Skickat från min SM-A530F via Tapatalk

[NordicCnc]

Status update.

I have ordered all parts and have now received linear rails and ballscrews from BST automation. Stuff looks high quality, but to my unpleasant surprise I ahbe received 2005 ballscrews instead of the requested and quoted 2010 ballscrews... Looks like I will only be getting 7.5m/min rapids now...

I will try to sort this out with Fred, but my expectations are not high. I am not sure how it works with returning the parts. I already paid extremely expensive shipping (due to current world situation) and rip off custom clearance.. Feelings are not the greaterst right now, downgrading from 15m/min to 7.5m/min.

As mentioned earlier in the build log, I will be using servos rated at 3000rpm, with a 2:1 ratio to limit ballscrew rpm to 1500. As I said, I have also already ordered belts and pulleys for that gearing. Fuck. Rant over.

Skickat från min SM-A530F via Tapatalk

[AndyUK]

Are you sure there aren't just two threads on the screws? I initially thought the same thing, then realised if I follow one helix it goes 10mm in one rotation.

[NordicCnc]

Are you sure there aren't just two threads on the screws? I initially thought the same thing, then realised if I follow one helix it goes 10mm in one rotation.Not sure but I quickly measured from one high point on the thread to the next. I neez to check again tomorrow since I already left the garage. I hope you are right though!

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

No they will be correct, they are twin start screws. If you look carefully at the end you will see multiple threads, but it's easy to check just rotate the ballnut and measure how far it moves.

It's a common mistake.

The reason they use this method is it gives higher torque and lessens the depth required and keeps the screw stronger.

[NordicCnc]

No they will be correct, they are twin start screws. If you look carefully at the end you will see multiple threads, but it's easy to check just rotate the ballnut and measure how far it moves.

It's a common mistake.

The reason they use this method is it gives higher torque and lessens the depth required and keeps the screw stronger.Omg I will be so happy when confirming this. I believe you and Andy. It really sucked out all my energy when measuring today. Now I have hope again!

I have also received following parts:
- 4x DMM 400w servo motors and drives. One is with a brake.
- ATC spindle from Jianken, 3.2kW with BT30 interface, sunfar E550 VFD, Waterpump and 100mm spindle bracket.
- Aluminium tooling plates from Germany.
- Ethernet SmoothStepper combined with MB3 breakout board.
- 1000mm straight edge DIN 874/2 and 300x200 square DIN875/1 from Helios Preisser.
- 20mm linear rails, 1605 and 2010 ballscrews, cable dragchains from BSTmotion.

I am still waiting for aluminium profiles and HTD pulleys.

I will post some pictures soon to show you guys all the hardware.

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

Omg I will be so happy when confirming this. I believe you and Andy. It really sucked out all my energy when measuring today. Now I have hope again

99.9999% Sure they will be correct. I did exactly the same first time I came across them so yes I know that feeling you felt.!:dejection:

[Kitwn]

Can someone of yous enlighten me as to why nobody uses them

Derek

You'll see some diagonals in welded steel frames but I suspect with aluminium it's the practicality as you sugest. If the profile manufacturers made the fittings for it and the builders bought the saws for it then you'd see nothing else.

[NordicCnc]

Any recommendations on building your electronics enclosure yourself versus buying one? I am struggling to find one that I think will be large enough for any sensible price. I want to fit everything in the sam enclose, which includes:

(3 phase 400V mains)
- AC line reactor
- Mains filter
- E550 Sunfar VFD
- 4x Dyn4 drives
- 1A schaffner filters for each drive
- MB3 breakout board
- 1 or 2 24VDC power supplys
- Mains circuit breaker
- Contactor for E-stop circuit
- Relay for E-stop circuit
- Circuit breakers for each independent phase (insteaf of regular fuses)
- Terminals

It starts to seem like I would nees atleast a cabinet sized 800x600x250mm. I figure welding my out of sheet metal would be cheapest.

Skickat från min SM-A530F via Tapatalk

[jarjar]

I can assure you that 800x600 is too small :)
This is what I have and everything is cramped inside.

I bought mine from rs-online, price seemed reasonable.

[Sterob]

My next build I'm going to look at using small extrusion for a enclsoure frame, then one can make it any size one wants.

[AndyUK]

I dithered for ages about this, feels like they should be cheaper.

In the end I went for a Tempa Pano EKO, which enabled me to have a glass front. From memory, the non-glass version, 800x600x300 was about £115, with glass circa £130.

http://www.tempapano.com/en/eko/

Is it overkill? Yeah. Is it nice? Uh-huh.

Space wise, I think its about right... theres plenty of room left in mine, and allowed lots of space for airflow and future expansion. But, honestly, I'm not using most of the depth and its all about how you want to package it. I could probably have fitted the same components in 500x500 or less. Have a look on my log for a reminder of how full up my cabinet is, page 9 post #87 has a photo.

Alternatively, I also looked at used server racks, which seemed like a viable option.

[JAZZCNC]

Any recommendations on building your electronics enclosure yourself versus buying one? I am struggling to find one that I think will be large enough for any sensible price. I want to fit everything in the sam enclose, which includes:

(3 phase 400V mains)
- AC line reactor
- Mains filter
- E550 Sunfar VFD
- 4x Dyn4 drives
- 1A schaffner filters for each drive
- MB3 breakout board
- 1 or 2 24VDC power supplys
- Mains circuit breaker
- Contactor for E-stop circuit
- Relay for E-stop circuit
- Circuit breakers for each independent phase (insteaf of regular fuses)
- Terminals

It starts to seem like I would nees atleast a cabinet sized 800x600x250mm. I figure welding my out of sheet metal would be cheapest.

Skickat från min SM-A530F via Tapatalk

Don't buy anything until you have everything electrical and lay it out on the bench. Servo drives require a lot more space than steppers for cooling purposes and just physically larger anyway. Also think about fitting fans and the Air flow when placing drives etc.

Given that your mostly using extrusion then I'd use aluminium panels and some small extrusion and build the control box into the lower frame if possible.

[JAZZCNC]

Talk about keeping us in suspenders,??? . . . . Was the pitch wrong.?

[NordicCnc]

Talk about keeping us in suspenders,??? . . . . Was the pitch wrong.?Sorry, the pitch was of course correct! Thanks for having me double check, I dont think I've realized unless until the build was finished!

About building the enclosure into lower frame with aluminium plates is a good idea. I have a large 2mm aluminium sheet lying around that I could use. For the next step I will lay out the electrical components in 3D and make a wiring diagram to ensure I have accounted for everything.

I've seen that atleast both DMM and Sunfar recommends either mains filter or both mains filter and line reactor for the mains power supply. I have seen people here using line filters but I haven't come across anyone using line reactors yet. Is that really as necessary as the manufacturers are claiming it to be?

Also mounting the electrical enclosure directly to the frame (made of aluminium) would it be a problem since it would basically be in direct contact with the machine frame? I mean that would be a direct earth connection for the enclosure. I am not too into electronics (yet)!

Thanks everyone for your good and helpful advice!

Skickat från min SM-A530F via Tapatalk

[Clive S]

Also mounting the electrical enclosure directly to the frame (made of aluminium) would it be a problem since it would basically be in direct contact with the machine frame? I mean that would be a direct earth connection for the enclosure. I am not too into electronics (yet)!

No. That will be fine just make sure that all grounds go to one star point. If the frame is painted I would make a direct connection from the box to the frame.

[JAZZCNC]

I've seen that atleast both DMM and Sunfar recommends either mains filter or both mains filter and line reactor for the mains power supply. I have seen people here using line filters but I haven't come across anyone using line reactors yet. Is that really as necessary as the manufacturers are claiming it to be?

Also mounting the electrical enclosure directly to the frame (made of aluminium) would it be a problem since it would basically be in direct contact with the machine frame? I mean that would be a direct earth connection for the enclosure. I am not too into electronics (yet)!


I would at least fit a mains filter because of them being Servo's, however, I have seen that some people have had issues with DMM servo's because they didn't use line reactors so you might want to fit them to be safe, or just go with mains filter and if you get any issues you can always slip a line reactor in between.!
It mostly depends on your incoming supply quality I think and I seem to remember those who have had issues where US based which works on different Freq and voltage to EU.

Regards the frame then yes it's fine and you would want the earth to the frame anyway for safety just in case a wire shorted to the frame.

[NordicCnc]

I would at least fit a mains filter because of them being Servo's, however, I have seen that some people have had issues with DMM servo's because they didn't use line reactors so you might want to fit them to be safe, or just go with mains filter and if you get any issues you can always slip a line reactor in between.!
It mostly depends on your incoming supply quality I think and I seem to remember those who have had issues where US based which works on different Freq and voltage to EU.

Regards the frame then yes it's fine and you would want the earth to the frame anyway for safety just in case a wire shorted to the frame.


No. That will be fine just make sure that all grounds go to one star point. If the frame is painted I would make a direct connection from the box to the frame.

Thanks both of you! This helps a ton. A line reactor costs about 100€, so not a big deal. I will fit one just to be sure!

Skickat från min SM-A530F via Tapatalk

[NordicCnc]

JAZZCNC, if you read this I have a question for you. How do you go about assembling the L-shape gantry together? I have received my aluminium profiles and to my surprise I noticed that they are not flat in the middle. It is probably hard to see in the picture but it is about 0.5mm.

28260

I was thinking that I could sand it down, since the only critical thing would be that it is flat on the sides so that it wont wiggle when assembled.

[JAZZCNC]

JAZZCNC, if you read this I have a question for you. How do you go about assembling the L-shape gantry together? I have received my aluminium profiles and to my surprise I noticed that they are not flat in the middle. It is probably hard to see in the picture but it is about 0.5mm.

I was thinking that I could sand it down, since the only critical thing would be that it is flat on the sides so that it wont wiggle when assembled.

I've never seen profile made like that.? If that picture is showing it correctly it looks like a flattened pyramid.!! . . . The profile I use is always flat but if there are any slight highs or lows I just shim the rail to suit.

[NordicCnc]

I've never seen profile made like that.? If that picture is showing it correctly it looks like a flattened pyramid.!! . . . The profile I use is always flat but if there are any slight highs or lows I just shim the rail to suit.Yes exactly. I got them from a local supplier, since I was not able to find any 160x80 heavy version profiles anywhere. Only the local supplier could offer them.

Ill probably face them in my crappy cnc router. It will be flat enough that way. I can probably get within 0.2mm over the 1200mm lenght. I wll then just shim the top raik.

Skickat från min SM-A530F via Tapatalk

[NordicCnc]

So I have installed 2 linear rail carriages just to test how they perform. Once again to my surprise, both of the tested carriages are running extremely bad. Same as in Joe's video. I am not sure if this is normal. I didn't apply any grease yet, I will try with grease applied tomorrow also. But I figure they should at least not bind/stagger every 10mm even without grease??

https://youtu.be/b_DDT43lBIchttps://uploads.tapatalk-cdn.com/20200601/f42953b103f2088fccd019ce64708adf.jpg

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

Well first have you got the bearing reference edge and rail reference edge-aligned.?

Also, they can feel stiff at first, esp when unloaded, just put some weight on them and feel how they slide. I think you'll find the slide easily with weight on them and greased up.

[NordicCnc]

Well first have you got the bearing reference edge and rail reference edge-aligned.?

Also, they can feel stiff at first, esp when unloaded, just put some weight on them and feel how they slide. I think you'll find the slide easily with weight on them and greased up.Yes I have those aligned. I reached out to Free and he said that it is most likely the plastic seal that is not aligned correctlt and thay it should be easy to fix. I will report back how it goes!

Skickat från min SM-A530F via Tapatalk

[JAZZCNC]

Yes I have those aligned. I reached out to Free and he said that it is most likely the plastic seal that is not aligned correctlt and thay it should be easy to fix. I will report back how it goes!

Skickat från min SM-A530F via Tapatalk

No I very much doubt it. The plastic seals don't grip that much and not something that misalign because of how the bearings are put together.
Normally I don't get trouble with Hi-win bearings needing cleaning from new but if it feels like just one of the bearings is doing it then it may be worth taking apart and checking there's no debris in it.
They are very easy to take apart and rebuild, just do it inside a tray because the balls go every where.

Personally I think they will be ok and your feeling the normal Sticky clicky notchy feeling they can give when dry and new. Obviously very difficult to now without feeling them.

[NordicCnc]

No I very much doubt it. The plastic seals don't grip that much and not something that misalign because of how the bearings are put together.
Normally I don't get trouble with Hi-win bearings needing cleaning from new but if it feels like just one of the bearings is doing it then it may be worth taking apart and checking there's no debris in it.
They are very easy to take apart and rebuild, just do it inside a tray because the balls go every where.

Personally I think they will be ok and your feeling the normal Sticky clicky notchy feeling they can give when dry and new. Obviously very difficult to now without feeling them.

You were right. After pushing them around for a while and greasing them, they now run much smoother. Problem solved!

[JAZZCNC]

You were right. After pushing them around for a while and greasing them, they now run much smoother. Problem solved!

Good, It's common for people not used to profiled rails to think they would glide as if on air when they are actually quite stiff. Preloaded ones are even worse, they can feel as if they have sized up.!!

[NordicCnc]

Sorry everyone for keeping you in the dark for such a long time. I have had a really busy couple of months, but I've managed to almost finalize the mechanical side and I am now working on the electronics. First some test was done for the X-axis yesterday. 15m/min rapid easily as expected, with auto tuned DMM 400w servo!

Anyway there was some vibration on the fixed bearing mounting plate, which I believe is the because of the high mounting. I plan to fix this by adding a 20mm aluminium plate on top of the X-axis ballscrew nut housing, to get the bearing plates lower down on the aluminium profile.

And by the way, lots and lots of shimming and dialing has been done for the rails..!

I didnt take any pictures or videos so here are some old pictures when I was laying out the electronics. More pictures will follow later this week..
https://uploads.tapatalk-cdn.com/20201231/25e42ffbdb9e5b37222b41d104bfd2e8.jpghttps://uploads.tapatalk-cdn.com/20201231/97b43f36177a9436d6889d5e9a28f391.jpg

Skickat från min SM-A530F via Tapatalk