New build - Steel - 3 Axis initially then conversion to 4 axis.

Hi guys,

Dave here. I'm designing a new machine for the office in the hopes of carving precision rifle stocks (4 axis) and whatever else is required from hardwood and aluminium. :)

The build is already underway and frame currently ready to be painted.

Details of the build
==============

Linear bearings: SBR20 fully supported rails - X:1300mm - Y:1000mm - Z:450mm
Ball screws: RM2005 Chinese - X:1350mm - Y:1050mm - Z:500mm
Steppers: Nema23 - 23HS45-4204S
Drivers: M542H
PSU: 48V 600w DC

Frame/Chassis: 65x65x3mm box steel
X axis chassis: 75x75x4mm box steel
Gantry cross beam: 200x100x5mm box steel
Gantry supports: 300x125x10mm steel angle
Z axis supports: 200x480x10mm steel plate

Contrary to most builds on here I do not have a CAD diagram to show as the build is a running design at present. I have very little CAD skills unlike my wife, who regularly makes designs in Solidworks.

Here's some pics of the current build:

original design
Attachment 18672

Adjustible feet
Attachment 18671

Front on view - all level within ~1mm
Attachment 18673

Side view
Attachment 18674

Z axis setup.
Attachment 18675

Gantry side on.
Attachment 18676

Anyway, thought i'd share guys. let me know what you think. :)

- Dave.

Looking good so far Dave.
Any reason for the 2005 ballscrews? The 1605/1610 are more popular because they have lower inertia which helps with the acceleration and deceleration, especially on the modest 48V you are using.

Quote:

---

Originally Posted by routercnc

Looking good so far Dave.
Any reason for the 2005 ballscrews? The 1605/1610 are more popular because they have lower inertia which helps with the acceleration and deceleration, especially on the modest 48V you are using.

---

I havent purchased the PSU yet. so may go larger. the motors claim to have a max voltage of 48v. i can get a 60V no troubles but am not sure of the impact that would have on the steppers. As far as the RM2005's goes, i was aiming for a rigid setup for cutting aluminium and didnt want any wobble from ball screws. Feed rate isnt that critical for me.

Quote:

---

Originally Posted by Slixxor

I havent purchased the PSU yet. so may go larger. the motors claim to have a max voltage of 48v. i can get a 60V no troubles but am not sure of the impact that would have on the steppers. As far as the RM2005's goes, i was aiming for a rigid setup for cutting aluminium and didnt want any wobble from ball screws. Feed rate isnt that critical for me.

---

I wouldn't be worried about the steppers but the drives in your spec you have said that you are using these M542H If I am not mistaken the max voltage they will handle is 50V so 48V would be too high for them. AM882 drives would be a better choice running at about 68-70 V.

If you want a decent machine you have to have decent electronics.

Quote:

---

Originally Posted by Clive S

I wouldn't be worried about the steppers but the drives in your spec you have said that you are using these M542H If I am not mistaken the max voltage they will handle is 50V so 48V would be too high for them. AM882 drives would be a better choice running at about 68-70 V.

If you want a decent machine you have to have decent electronics.

---

Ok, it's now being considered. :) havent bought the drivers yet either. you are right indeed, my voltage restriction was the M542H. price is only AU $20 more per driver.

Quote:

---

Originally Posted by Slixxor

Ok, it's now being considered. :) havent bought the drivers yet either. you are right indeed, my voltage restriction was the M542H. price is only AU $20 more per driver.

---

Ok then if you have not bought the motors either then try and find 8 wire so they can be connected in parallel and get the lowest inductance motors you can find these are proved and tested by many over here :- http://cnc4you.co.uk/resources/Stepp...1B%203.1Nm.pdf.

Good luck with the build.

Quote:

---

Originally Posted by Clive S

Ok then if you have not bought the motors either then try and find 8 wire so they can be connected in parallel and get the lowest inductance motors you can find these are proved and tested by many over here :- http://cnc4you.co.uk/resources/Stepp...1B%203.1Nm.pdf.

Good luck with the build.

---

I have the motors already and all mounting hardware. saving the electronics for last. The AM882's are in the price range. I am running dual ballscrews on the X Axis. so additonal cost. but oh well :)

You can run both X axis screws with one motor and belt connecting them,it's common practice with the guys on the forum.
Good luck
Mike

Quote:

---

Originally Posted by mekanik

You can run both X axis screws with one motor and belt connecting them,it's common practice with the guys on the forum.
Good luck
Mike

---

I have noticed that. but whats the go with custom belts? Obviously each machine will differ in length required. Do you simply ask for a particular length?

You can get custom belts but they are more expensive, but you have a lot of leeway length wise as you will have adjustment on the motor and your tension/routing belt layout. there was a post on the forum a while ago showing a belt routing setup to address this issue.
Regards
Mike

Ok. Z AXIS design time. What is the acceptable leverage ratio of a sturdy Z?

My length of linear rail is 450mm. I was hoping for 300mm
Of cutting depth but it seems like there would be excessive difficulty keeping my Z cutting straight with a 150mm (1/3 ratio) binding area for the spindle on a 450mm run. What do you all suggest?

Also, would be wiser to mount the linesr bearings on the spindle plate or on the gantry plate?

Rails on the spindle plate. To make a variable lever. Bearing blocks spaced at 50% of z travel and as wide as possible.

I'll let the resident experts correct me if I am wrong. I'm still drawing up my design. However, this is my understanding.

X and Y axis bolted and mounted. Was a very tight fit on the Y axis bearing blocks but It fits like a glove :)

Attachment 18733

Attachment 18734

Few more updates. a redesign of the Z axis and some major work to keep it all straight. moves very smoothly now. Z axis is 6mm steel. working on the ball screws next.

Here's a quick vid of the movement: https://youtu.be/DRLe3EPQ01M

Here's some photos.
Attachment 18805Attachment 18806Attachment 18807Attachment 18808Attachment 18809Attachment 18810Attachment 18811Attachment 18812Attachment 18813

OK, so progress has been slow but i've finalized and purchased the electronics.

PSU's: 70v 5.7A Switchmode PSU.

Drivers: 2MA860H.

Motors:

Manufacturer Part Number	23HS45-4204S
Step Angle	1.8°
Holding Torque	3.0Nm(425oz.in)
Rated Current/phase	4.2A
Phase Resistance	0.9ohms
Recommended Voltage	24-48V
Inductance	3.8mH±20%(1KHz)
Weight	1.8kg

Thanks to CliveS and the others on suggesting not buying a kit.

More to come shortly. there have been some more changes :)

Z Axis finally fitted and relatively smooth with motion (No racking), Some filler welds will be done in the coming days and the wiring should be done shortly. :) I'm finally starting to see it all come together. I wasn't really sure if my Z design would work, but it seems extremely rigid. Although I've lost 100mm of travel on the Y axis I've maintained a good center of gravity on the Z axis and increased the X axis work area.

Hopefully cutting in January!

Attachment 20049Attachment 20050Attachment 20051

Nice concept of Z axis, only problem is that doesn't allow for any adjustment other than shims. How smooth it travels up and down?
Asking as I machined similar stuff on chinese mill and wasn't parallel.
Same thing for entire machine - you are mounting rails directly to profiles - no machined base, no epoxy for leveling (parallel, squareness?)
I do not criticize you - just wander if you aim for accuracy in your machine.

Quote:

---

Originally Posted by Tom J

Nice concept of Z axis, only problem is that doesn't allow for any adjustment other than shims. How smooth it travels up and down?
Asking as I machined similar stuff on chinese mill and wasn't parallel.
Same thing for entire machine - you are mounting rails directly to profiles - no machined base, no epoxy for leveling (parallel, squareness?)
I do not criticize you - just wander if you aim for accuracy in your machine.

---

Pretty smoothly. I do have a video of it turning by hand..

https://youtu.be/lEZzJ3c4UJA

I think the reason for the success in motion is the heating of various sections during the fabrication process.

I like working with steel because it warps and moves when heated.

When designing the Z most of the welding occured on Job with clamps. As the steel heats and warps it smoothly contours around its pressure points. I.e: the fasteners on the bearing blocks and the tension applied by the SBR25 rail.

Not to mention meticulous time spent hammering the fudge out of it at times :) the outer walls were hardest. They have a slight warp in the 6mm and needed lots of bashing and even driving the car over it.

Cutting a square for the ballscrew bracket was the best idea. As it allowed me to drill it then weld it back in with just the tension of the ballscrew to warp to.

There really was no other way though. Accessing the correct drill points would have been almost impossible to get right as it was completely blind.

Sent from my SM-G920I using Tapatalk

Quote:

---

Originally Posted by Slixxor

Pretty smoothly. I do have a video of it turning by hand..

https://youtu.be/lEZzJ3c4UJA

I think the reason for the success in motion is the heating of various sections during the fabrication process.

I like working with steel because it warps and moves when heated.

When designing the Z most of the welding occured on Job with clamps. As the steel heats and warps it smoothly contours around its pressure points. I.e: the fasteners on the bearing blocks and the tension applied by the SBR25 rail.

Not to mention meticulous time spent hammering the fudge out of it at times :) the outer walls were hardest. They have a slight warp in the 6mm and needed lots of bashing and even driving the car over it.

Cutting a square for the ballscrew bracket was the best idea. As it allowed me to drill it then weld it back in with just the tension of the ballscrew to warp to.

There really was no other way though. Accessing the correct drill points would have been almost impossible to get right as it was completely blind.

Sent from my SM-G920I using Tapatalk

---

My first bed was ok till I weld those flat bars 8mm thick - it warp badly in the middle
Attachment 20096Attachment 20097

Quote:

---

Originally Posted by Tom J

My first bed was ok till I weld those flat bars 8mm thick - it warp badly in the middle
Attachment 20096Attachment 20097

---

Looking goid mate! Tack welds are the way to go. And lots of clamping. I would have got a piece of say 35x70mm box tube and clamped it in several places to the flat bar piece you were working on. Then going from the outsides in to the middle as your box frame will also slightly warp from the welds. Tack everything first then repeat the clamping process all over again with your filler welds.

It's hard with long runs like that. Partly the reason bojan S uses C channel for his bed. Although I find welding much easier, drilling and bolting would have given a better result.

Whats the link for your thread?

Sent from my SM-G920I using Tapatalk

Good job your using round rails because wouldn't get away with rough work like this if using Profiled linear rails.!!

Tom is correct your level of accuracy will be low or will require lot of shimming and tweaking. How much will only come to light when finished and being used at which point you'll wished you'd took little more care when building.:thumbdown:

Quote:

---

Originally Posted by Slixxor

Looking goid mate! Tack welds are the way to go. And lots of clamping. I would have got a piece of say 35x70mm box tube and clamped it in several places to the flat bar piece you were working on. Then going from the outsides in to the middle as your box frame will also slightly warp from the welds. Tack everything first then repeat the clamping process all over again with your filler welds.

It's hard with long runs like that. Partly the reason bojan S uses C channel for his bed. Although I find welding much easier, drilling and bolting would have given a better result.

Whats the link for your thread?

Sent from my SM-G920I using Tapatalk

---

Here is my latest build
http://www.mycncuk.com/threads/9073-...-router-3-axis
Small work area but was aiming for accuracy. Box 100x100 end up with ~0.07mm that is why i build again fixed gantry one, link:
http://www.mycncuk.com/threads/10459...teel-frame-MK2
Boyan is master in precision fallow him - he gets 0.02mm wich is more than my chinese mill

Quote:

---

Originally Posted by Tom J

Boyan is master in precision fallow him - he gets 0.02mm wich is more than my chinese mill

---

Ye and I'm father Xmas.!!

Quote:

---

Originally Posted by Slixxor

Looking goid mate! Tack welds are the way to go. And lots of clamping. I would have got a piece of say 35x70mm box tube and clamped it in several places to the flat bar piece you were working on. Then going from the outsides in to the middle as your box frame will also slightly warp from the welds. Tack everything first then repeat the clamping process all over again with your filler welds.

It's hard with long runs like that. Partly the reason bojan S uses C channel for his bed. Although I find welding much easier, drilling and bolting would have given a better result.

Whats the link for your thread?

Sent from my SM-G920I using Tapatalk

---

My frame was not more than 1mm warp as I tack it everything and weld one bead at the time, cool down and opposite side and so on. I have to admit that I broke my rule and did not let to cool enough (firstly was warp corners down than after cool opposite) - £100+ waste on steel and gas, lesson learnt. I bit it with sledge hammer, driven by the car and finally filled down, so my 8mm bars became 4mm in places:)
Now I use Tig just to see if I can do better
Attachment 20098Attachment 20099

Quote:

---

Originally Posted by JAZZCNC

Good job your using round rails because wouldn't get away with rough work like this if using Profiled linear rails.!!

Tom is correct your level of accuracy will be low or will require lot of shimming and tweaking. How much will only come to light when finished and being used at which point you'll wished you'd took little more care when building.[emoji107]

---

Haha savage feedback. Thanks Jazzy, time will tell.

Sent from my SM-G920I using Tapatalk

Quote:

---

Originally Posted by Slixxor

Haha savage feedback. Thanks Jazzy, time will tell.

Sent from my SM-G920I using Tapatalk

---

I could blow smoke up your arse but still won't make it any better when cutting time arrives.:encouragement:

Quote:

---

Originally Posted by JAZZCNC

Ye and I'm father Xmas.!!

---

do not be jealous Dean, results speaks itself. And on the end of the he takes your advises like we all do:)

Quote:

---

Originally Posted by Tom J

do not be jealous Dean, results speaks itself. And on the end of the he takes your advises like we all do:)

---

I'm not into jealousy but I am realist and know there's no way he's measuring 0.02mm over 3000mm like he's claimed in the past or even over shorter distance.!! . . . Ask me why.?

Ok I'll tell you.!! . . . . . Would be bloody impossible because he's using C7 ballscrews From China which at best will only give 0.05mm over 300mm. Then throw in few belts/pulleys etc and this goes down further.

So who's fooling who.?

Quote:

---

Originally Posted by JAZZCNC

I could blow smoke up your arse but still won't make it any better when cutting time arrives.:encouragement:

---

I just like the way you assumed I needed to be more careful. What exactly gives you the impression I haven't been?

Hence the reason I will see how it goes when it's done. I've endeavoured to make it as square as possible on all axes and feel comfortable that I've managed to achieve that within the best of my ability.

I'll be sure to let you know if you're right or wrong once it's done and I tweak everything :)

Sent from my SM-G920I using Tapatalk

Quote:

---

Originally Posted by JAZZCNC

I'm not into jealousy but I am realist and know there's no way he's measuring 0.02mm over 3000mm like he's claimed in the past or even over shorter distance.!! . . . Ask me why.?

Ok I'll tell you.!! . . . . . Would be bloody impossible because he's using C7 ballscrews From China which at best will only give 0.05mm over 300mm. Then throw in few belts/pulleys etc and this goes down further.

So who's fooling who.?

---

When did I make that claim? This is my first build and I don't see that claim made :/? I think you grossly underestimate my realism.. [emoji14]

Sent from my SM-G920I using Tapatalk

Quote:

---

Originally Posted by Tom J

My frame was not more than 1mm warp as I tack it everything and weld one bead at the time, cool down and opposite side and so on. I have to admit that I broke my rule and did not let to cool enough (firstly was warp corners down than after cool opposite) - £100+ waste on steel and gas, lesson learnt. I bit it with sledge hammer, driven by the car and finally filled down, so my 8mm bars became 4mm in places:)
Now I use Tig just to see if I can do better
Attachment 20098Attachment 20099

---

TIG will give a better result but will still warp. Either way it can be a pickle when working with steel. I still haven't decided what bed to go with on mine. I will probably do MDF to start with.

Sent from my SM-G920I using Tapatalk

Quote:

---

Originally Posted by Slixxor

When did I make that claim? This is my first build and I don't see that claim made :/? I think you grossly underestimate my realism.. [emoji14]

Sent from my SM-G920I using Tapatalk

---

Crossed purposes wasn't talking about you with that statement.

Quote:

---

Originally Posted by Slixxor

I just like the way you assumed I needed to be more careful. What exactly gives you the impression I haven't been?

---

Because I'm not blind and can see there's been no machine work done on any of those surfaces or any thought regards rails being on same plane etc.
I've never yet seen piece of angle iron that is perfectly flat or 90deg and you are bolting straight to the face of one.

Then we have the heat put into the metal which can clearly be seen. You can Stitch it with pigeon Shit and beat the shite out of it all you like but it will still be warped or twisted in some way unless you take other steps to correct it.

All these little things will bite and like I've said your lucky your using round rail. Profiled linear rail would just not tolerate this level of in accuracy and bind up solid.

Edit: Just to be anal and prove my point look at this picture and tell me whats wrong.?

Attachment 20107

Quote:

---

Originally Posted by Slixxor

When did I make that claim? This is my first build and I don't see that claim made :/? I think you grossly underestimate my realism.. [emoji14]

Sent from my SM-G920I using Tapatalk

---

If you read it again Dean was not referring to you with the above It was a claim made by another on the forum.

Quote:

---

Hence the reason I will see how it goes when it's done. I've endeavoured to make it as square as possible on all axes and feel comfortable that I've managed to achieve that within the best of my ability.

---

At the end of the day that is all anybody can do. Looking forward to seeing the first cut :friendly_wink:

You sir have a much better personality. I think JAZZCNC needs a holiday by the sounds.. Clive, legend as always ;)

Quote:

---

Originally Posted by JAZZCNC

Crossed purposes wasn't talking about you with that statement.



Because I'm not blind and can see there's been no machine work done on any of those surfaces or any thought regards rails being on same plane etc.
I've never yet seen piece of angle iron that is perfectly flat or 90deg and you are bolting straight to the face of one.

Then we have the heat put into the metal which can clearly be seen. You can Stitch it with pigeon Shit and beat the shite out of it all you like but it will still be warped or twisted in some way unless you take other steps to correct it.

All these little things will bite and like I've said your lucky your using round rail. Profiled linear rail would just not tolerate this level of in accuracy and bind up solid.

Edit: Just to be anal and prove my point look at this picture and tell me whats wrong.?

Attachment 20107

---

You are referring to me shaping it? I don't see the problem? what are you suggesting? (genuinely asking)

Having a good read of this thread I'm rather speechless and even more speechless with the commentary that goes with your posts Slixxor, you can educate yourself so well on the internet of the basics it' just takes a bit of time to learn and absorb. I'm not going to pick up on individual things because would be here until the 1st of January. This is not a good advert for someone to have a go at home. You are one brave man to post this publicly.

Quote:

---

Originally Posted by spluppit

Having a good read of this thread I'm rather speechless and even more speechless with the commentary that goes with your posts Slixxor, you can educate yourself so well on the internet of the basics it' just takes a bit of time to learn and absorb. I'm not going to pick up on individual things because would be here until the 1st of January. This is not a good advert for someone to have a go at home. You are one brave man to post this publicly.

---

Im confused what bewilders you? Can you elaborate? I dont see how my build is different to any other. And my comments are far from offensive. Do you just not like the design. Im confused

Sent from my SM-G920I using Tapatalk

In post #32 the Z bearings are trapped between the fixed distance between the angle iron sections. This means they cannot sit naturally, or have any pre-load control, and you can't adjust it either.

I'm guessing you welded it together with the angle iron bolted to the bearings? If you release the bolts into the Z bearings do they slide out easily from under the Z axis (which is what you want), or does the Z axis 'spring away' or 'close up'? Ideally this should be a nice gentle fit onto the bearings , or possibly have a slight bit of pre-load.

Quote:

---

Originally Posted by routercnc

In post #32 the Z bearings are trapped between the fixed distance between the angle iron sections. This means they cannot sit naturally, or have any pre-load control, and you can't adjust it either.

I'm guessing you welded it together with the angle iron bolted to the bearings? If you release the bolts into the Z bearings do they slide out easily from under the Z axis (which is what you want), or does the Z axis 'spring away' or 'close up'? Ideally this should be a nice gentle fit onto the bearings , or possibly have a slight bit of pre-load.

---

When welded they werent bolted on. Just the outer frame clamped onto both sides with pressure applied. When loosened there is slight rubbing on the bearing blocks but minimal force is required to remove them. If on an angle they scrape and fall straight out. The bearing holes were drilled afterwards with bearing blocks on the rails with no tension. The rear of the bearings were coated with steel blue and the top casing was slid over the bearing blocks creating the guide for my bearing block drilling die I'd made.


Sent from my SM-G920I using Tapatalk

Quote:

---

Originally Posted by Slixxor

You sir have a much better personality. I think JAZZCNC needs a holiday by the sounds.. Clive, legend as always ;)

---

Oh so your YES man well tough luck because if it's shit you'll get told it's shit by me.! . . . . Like wise if it's good I'll say so as well but I see nothing here that's good.!

To be fair I've been very kind on what I've commented so far, mainly because not got full week spare to list all that's bad.!

Quote:

---

Originally Posted by Slixxor

You are referring to me shaping it? I don't see the problem? what are you suggesting? (genuinely asking)

---

Nothing to do with shaping and every thing to do with being warped and surfaces not on same plane. Then generally just weak poor rough arsed engineering which for CNC won't cut it.
Just because your using steel doesn't mean it's stronger. If the design is weak and poorly built which this is then doesn't matter if built from 2" steel material or 0.1" aluminum (note Ref in American just to so can understand . .lol) you'll never succeed in achieving accuracy or reliabilty.

Look again at this Pic for just few bits of whats wrong.!

Attachment 20111