Am I heading in the right direction?....

[Lloyd]

Dear all

thanks for taking the time to read my post...

I am about to embark on a CNC router build.

id like to know if I'm thinking along the right track for each aspect of the machine - I know it's hard to picture it without a picture... I'm in the process of designing it and will create a build log in the meantime...

The size I'd like to be able to cut (minimum) is 1200/800/300 -x/y/z respectively

so I'm thinking I'll need approx 1600/950/550 rails (don't ask where I got those numbers from :) they're a guess loosely based on a rough guess of space between bearings)

i want to be able to cut aluminium, but the primary function is to cut PU foam and wood, like MDF or similar...

for the base frame I'm thinking 80x80x3 box-section - similar to designs already posted... I'm used to building heavy duty tables that need to be level, and understand these machines clearly need to be level, namely the long X axis rails need to be spot on parallel in both horizontal and vertical planes. - I heard about using epoxy - didn't really get it if I'm honest - when fabricating the base I would try employ all the usual tricks to ensure accuracy - but any further explanation would be greatly appreciated

the X and Y rails I'd go for profiled rails(square) and for the Z the SBR20 - all with 1x 20mm ball screw (X possibly upgraded later to 2x ball screw) what are the best aspect ratios to go for when setting the bearing blocks under the gantry sides? I've set them 150mm apart (distance between them)

the sides of the Y/Z gantry will be alu plate 20mm over the rails, with series 6 profile (60x120) (rails located top and bottom) spanning the two X-rails.

the Z will be alu plate with the rails on the spindle side of things...

going based on what existing members have already discussed, for the electronics, I'm still not so sure..

what are the best steppers to use? - size and power I can work out I guess , and I'm thinking in the ball park of 3Nm... Readily available, but would need more info with regard to spec-ing the best and what to watch out for... Digital/resonance/racking are all words I've heard... What is racking?

what drivers are going to be the most reliable? Which power source the best...

I did a forum search on this, but I didn't find anything really conclusive - that said, I'm new to the forum game, so it's probably an oversight on my part...

have I missed anything????? what other rules of thumb are there that you guys adopt when building a machine for the purpose of cutting those materials?

thanks in advance

[EddyCurrent]

Lloyd,

I see you have been doing some reading up already and you have a good idea about what you need.
Starting from the top of your post;

Length of rails will be best determined once you have a drawing together but they look to be in the right area.

80x80x3 box section is fine but if you are bolting your rails to this then the wall thickness is insufficient. I used 50x50x3 for my 900x600 machine but used 6mm flat bar inside the box section where my rails bolted to so as to increase the thickness to 9mm, much more meat to tap holes into. Some people make a bed that can be moved up and down so you might want to consider that, it might mean you could reduce the stroke of your Z axis.

The epoxy method involves creating a moat around the top surfaces where the rails bolt to, the moat is then filled with low viscosity epoxy resin and like water will find it's own level thanks to gravity. Because the two X rails are mounted on two separate pieces of box section, an additional 'bridge' has to be created to join them so that the epoxy can flow between them like an aqueduct thus ensuring all top surfaces of the epoxy are at the same level.

It's a pity you mention SRB20 rails for the Z axis, you would be far better using profiled rails all round such as the Hiwin type.

Go with 2 ball screws on X right from the start, there are reported issues when using a central one. When cutting at each extreme of the Y axis there is a tendency for the gantry to twist or rack, pivoting on the central ball screw.

From what I've read I believe you would be better with 16mm ball screws, 10mm pitch for X and Y and 5mm pitch for Z

This is a good diagram I think to explain positioning of parts Do it yourself CNC router: Design Considerations, the Gantry
plus there is this calculator thanks to 'routercnc' http://www.mycncuk.com/forums/linear...html#post15791

Make a drawing of your gantry and post it for comment before starting, 20mm plate is the norm. Most people use the free Sketchup software.

3Nm sounds fine, I used these Nema23 Stepper Motor 4Nm

Digital drivers as supposed to be the best, it used to be Leadshine AM882's but I think the EM806 is taking over Stepper Motors & Drivers - Zapp Automation Ltd
Of course you could use servo's instead or a closed loop system.

The best power source is one you build yourself, it's very easy and all the info will be available here when you need it. The design will consist of, toroidal transformer, bridge rectifier, smoothing capacitor(s), fuses.

You also need to think about which spindle you are going for, most common is the 2.2kW water cooled ones from China that come with a variable frequency drive (vfd) e.g. WATER-COOLE MOTOR SPINDLE 2.2KW AND INVERTER VFD good | eBay

what other rules of thumb are there that you guys adopt when building a machine for the purpose of cutting those materials?

If you keep posting and refining your design according to forum input I'm sure you will end up with a suitable machine, I did for one.

[Lloyd]

Those links look great! Thank you...

on the design I'm working on, I've now change all rails to square profile type RE your suggestion...

I see what you mean about the epoxy method...

a design I'm trying now (based on a cnc I've seen) is to use 90x60 aluminium profile, heavy duty, which will hold the rail and will be linked to the steel sub-frame at 5 plates along its length - each plate will have 4/6x M10 stubby bolts to allow me to make fine adjustments to the level.

Ive be attached a pic of the design in progress with this system - probably pretty confusing - I'm not using any regular design program, only what I'm used to, illustrator... But with some inspection would give a rough idea of where I'm heading... Anyway, it's certainly not finished, once I'm finalised on the equipment, I will modify the design to suit...

[JAZZCNC]

Eddy is correct with dropping the SBR rails it will let the machine down badly. The Z axis is THE most important part of the machine so if weak it doesn't matter how good the rest of machine is the cut finish will be poor.
Also 300mm Z travel is quite a lot so if your not having a adjustable bed then you will struggle to cut anything hard and thin(<100mm) unless you raise the material to the cutter and reduce Z extension. If not you'll experience bad chatter and poor finish on hard materials and you can forget Aluminium at full 300mm extension.

Regards ballscrews then I wouldn't use 16mm diameter for the long axis because in my experience it's just a bit too long. Now you can get 10mm pitch in 20mm Diameter then your much better using them over 16mm at this length. The difference in performance is neglible and your far less likely to get whip so often can go faster than 16mm screws that whip.!
Use 16mm Dia 5mm pitch for Z axis.
Also I recommend you connect steppers to screw with timing belts and pulleys to help reduce resonance and add some flexabilty with ratios if required.

The gantry you have drawn will need some bracing on the bracket holding the ballnut or much thicker bracket.
To be honest I'm not a fan of the place you have put the ballscrew as it's in the firing line of chips and is fulnerable to damage being high up and exposed.

Electrics then I urge you to go with good digital drives, Esp if building from steel as they handle resonance so much better than anologue drives. Also if planning on using Mach3 then invest in a motion control card. Both these will give you the best machine possible and least hassle along with best performance.

3Nm motors run around 65-70Vdc will be fine but check out the indutance of the motors first, if buying from Zapp or CNC4You then you'll be fine but if from Ebay then they could be High inductance motors which will underperform.
The drives you pick will need to be capable of a higher voltage than 65-70Vdc you'll run motors at with a little safety margin so 80Vdc are the norm.

Toroidal PSU like eddy mentions is best for CNC and you'll need to size so outputs correct Dc voltage running drives at with Amps roughly 70% of total motor amp ratings IE: 4 x 4.2A motors = 16.8A / 70% = 11.76A so output any where around 11-12A will be fine. It's not critical as you'll never get anywhere near pulling the full 11A at one time.

GOLDEN RULE Don't buy anything unless your 110% sure it's correct for your needs.!! . . . . If not then ask on here first.!

[Lloyd]

Thanks Jazz,

Regards ballscrews then I wouldn't use 16mm diameter for the long axis because in my experience it's just a bit too long. Now you can get 10mm pitch in 20mm Diameter then your much better using them over 16mm at this length. The difference in performance is neglible and your far less likely to get whip so often can go faster than 16mm screws that whip.!Use 16mm Dia 5mm pitch for Z axis.
Also I recommend you connect steppers to screw with timing belts and pulleys to help reduce resonance and add some flexabilty with ratios if required.

i was going to use 25mm ball screws for all of them, but I see it would be better with 16mm on the Z... Pulleys idea - I hadn't thought of...

Regarding the gantry sides, I am using 20mm alu as the bracket, I thought this would be plenty rigid enough based on other designs I saw - how would you size it different?

the location of the ball screw - I accept your opinion on this... My thought was because at the max depth I'll be cutting just PU foam, and this is far below the ballscrew it should be ok, and in a simple location build-wise...

Where would you try set it otherwise?

thank you for the heads up on the motors from Zapp

i have heard that Mach 3 can be unreliable - what's the consensus experience with it?

anyway, regarding the hardwear, I'm definitely going for profile rails, circa 25mm, with ballscrews on X and Y of 25mm, and 16 on Z, each with 10/10/5mm pitch respectively.

motors - definitely digital sounds better... Power likely as advised

[JAZZCNC]

Regarding the gantry sides, I am using 20mm alu as the bracket, I thought this would be plenty rigid enough based on other designs I saw - how would you size it different?

20mm is ok-ish but I'd brace it more. Don't under estimate the power of steppers connected to ballscrews. If the cutter jams for any reason but doesn't snap straight away, which does happen with thicker cutters, then something will bend and those plates are directly connected to the screws so are first on the list.! Closely followed by the Z axis front plate( I've seen my front plate bend many times)

the location of the ball screw - I accept your opinion on this... My thought was because at the max depth I'll be cutting just PU foam, and this is far below the ballscrew it should be ok, and in a simple location build-wise...

It's a personal thing and while it's in simple location it's in a vunerable location for getting leant on or against while leaning into machine, crap hitting it etc.

Where would you try set it otherwise?

Tucked in at the sides lower down with some form of cover.

i have heard that Mach 3 can be unreliable - what's the consensus experience with it?

No not at all, yes it as some faults just like others do, inc Linux CNC, but in the main it's IMO the best control for begginers thru to advanced users. There's more support from a massive user base so if you do have any problems there will be someone to help.
The main or most problems with Mach comes from the PC parallel port and people over tuning there motors not the program.
This is why I recommend you invest in a motion control card if using Mach, they allow much smoother and faster performance along with much better reliabilty due to the very stable pulse streams they provide, unlike the parallel port.

anyway, regarding the hardwear, I'm definitely going for profile rails, circa 25mm, with ballscrews on X and Y of 25mm, and 16 on Z, each with 10/10/5mm pitch respectively.

DONT use 25mm ballscrews this is a BIG mistake.? The inertia from 25mm screws is too much for nema 23/4 motors and you'll end up with a very slow under performing machine. Like wise if you go with nema34 Motors then unless you use expensive high voltage drives you'll have underperforming machine due to the slower speed of the motors.

Also you won't need 25mm rails 20mm will be fine, even 15mm would be more than good enough. The main reason why 20mm is prefered is due to bearing size being better suited. Plus the cost differance is nothing. Where as 25mm bearings are just a little too big and then to a lesser degree you have the negative inertia affect and the weight while accelerating/de-accelerating. Loadings wise even 15mm profiled linear bearings can handle the loads you'll need with some to spare.

The combination of 3.1 or 4Nm Nema 23 motors run at 70Vdc connected to 20mm dia 10mm for axis over 1300-1400mm long is well proven if you move away from it you'll regret it when you see a machine that as it. The machine running 25mm screws on 25mm rails with nema 23 motors will struggle perform half as well.!!
At the length your building your even on the border of being able to use 16mm ballscrews for the long axis but 20mm are safer bet in my experience.
I would recommend 20mm for X axis if screw longer than 1400mm, 16mm for Y & Z. 20mm rails all round.

[EddyCurrent]

Regards ballscrews then I wouldn't use 16mm diameter for the long axis because in my experience it's just a bit too long. Now you can get 10mm pitch in 20mm Diameter then your much better using them over 16mm at this length. The difference in performance is neglible and your far less likely to get whip so often can go faster than 16mm screws that whip.!

Yes, sorry about that, I overlooked the length, I was still thinking about my own build.

[JAZZCNC]

Yes, sorry about that, I overlooked the length, I was still thinking about my own build.

Wasn't having ago at you Eddy because he is on the edge of being able to use 16mm just IME 20mm are better at this length and the higher inertia doesn't have a big negative affect compared to if they whip.!

[Lloyd]

DONT use 25mm ballscrews this is a BIG mistake.? The inertia from 25mm screws is too much for nema 23/4 motors and you'll end up with a very slow under performing machine. Like wise if you go with nema34 Motors then unless you use expensive high voltage drives you'll have underperforming machine due to the slower speed of the motors.

I'm considering investing quite some money into the machine - as I'll be using it for some work I do... So I don't mind kitting it out with good quality motors...

that said, it would be good to see anyone's idea of what percentage of your budget you' should allocate to which part of the machine... e.g. 15% on base structure, 35% xyz hard wear 50% on power, control, motors etc....

So you get the best out for your buck..

any takers?

anyway, I'll take your advice regarding the brackets, positioning, and screws, - but this could change should I decide to invest in more expensive digital drivers?

thanks again for the comments

[Ger21]

that said, it would be good to see anyone's idea of what percentage of your budget you' should allocate to which part of the machine... e.g. 15% on base structure, 35% xyz hard wear 50% on power, control, motors etc....

So you get the best out for your buck..

any takers?

I've never seen anyone try to break it down that way. Probably because it doesn't work. You can't really cut corners anywhere without sacrificing the quality of the machine.

Good linear guides are the foundation of a good machine. Cutting corners there directly impacts the quality of the machine.
Good linear guides require a solid, rigid frame. The straightness and rigidity of the linear guides is dependent on the mounting surfaces, so you can't cut corners there.

Good motors and bad motors are all about the same price. There aren't really good and bad motors, but rather correct and incorrect for your application.
You can save a bit of money on drives. However, most people that cut corners on drives tend to upgrade to the better drives to improve performance, and end up spending more than if they had purchased the good drives in the first place.

The cheapest way to get a good machine is to spend more money. Any places you try to save money will usually result in spending more money later to replace the cheap components. Or you live with an inferior machine that still cost 80% of what a good machine would have cost.