Sturdy and Fast all Steel CNC , my first build

3 NM steppers from Zappautomation coupled with AM882 drivers , driven by 650w Toroidal Psu at 70VDC / 2×50v rectified by power regulator board with 4 big capacitors that gives 70VDC at the output/.
Old Dell PC 2.8Ghz with 512 Ram, stripped reworked and hacked to be extremely light and fast + stable version of Windows Server 2003, 25khz Mach3, 1/8 microstepping, 4.2A motor selected by jumpers on AM882.
Deflection at any part of the machine at any direction less than 0.00mm under 100kg load/me/
20t HTD 5M Pulleys, 270mm belts 15mm wide. 1610 screws on all axis, 1605 on Z, 1:1,
Velocity 10m per minute,acceleration 800 at the motor tuning in Mach 3.
Temperature of motors at X and A 55 Celsius, at Y and Z 35 Celsius, when here 30 Celsius.

Yes, Dean and Johnathan are the guilty ones that all went that good. Not only them but all who criticized the initial design and followed the build which inspired me when I was tired. Great Forum. I bow to all the wisdom here.

Once again, brilliant. Not one of the advisors, but an enthusiastic onlooker!!! G.

Wow congrats Silyavksi that's a great looking machine!
If I may ask: Was it very difficult to align the top and bottom rails to get them parallel in the Z direction? Did you just epoxy cast the top and then bottom from the same level surface and the two rails were aligned or did you have to perform other magic tricks as well?

Quote:

---

Originally Posted by mitchejc

Wow congrats Silyavksi that's a great looking machine!
If I may ask: Was it very difficult to align the top and bottom rails to get them parallel in the Z direction? Did you just epoxy cast the top and then bottom from the same level surface and the two rails were aligned or did you have to perform other magic tricks as well?

---

Thanks!
read please post #93 , point 1 to 4, especially 4

It was a lot of thinking but once i figured the correct steps how to do it, it was easy. You will need a straight edge though, long enough that can lay on both long axis rails perpendicular. + Big enough square. Without that its not possible to align the machine correctly. Both can be obtained for around 100euro and this is money well spend. the links are somewhere in the thread.

silyavski, this is a very impressive build. Thanks for sharing.

Beautiful build, and a lot of great info. Congrats on the final outcome. Must have been hard to part with it :).
Thanks again for sharing.

Hi Silyavski,

I see you have used HTD pulleys and belts on your machine. Did you get those from Zapp?
Also I see they do key cutting which makes it easy to mount the pulley to the servo but how did you mount the pulleys on to the ballscrew. I know its a newbie question but that is what I am. I am thinking to use 3 to 1 reduction and trying to figure it out. Also do you need any special machining for the ballscrews when using pulleys? HTD belts are a good option for me as they are rated to -25 Celsius.

Thanks,
Vass

Hi,
All pulleys and belts are from beltingonline.com
Nothing special, ball screws machined normally. 2 grub screws on pulley. A drop of instant glue on the shaft. If I was doing it for myself i would file or drill the shaft where the grub screws are. Now it's more forgiving for mistakes

On the big machine I am building now for myself I use only aluminum pulleys as it would be geared 20:30 t and the rotating nut is heavy so I don't want to have too high inertia. These I bought from China as here prices are prohibitive. Aliexpress.

Normally you don't need any special machining given that the motor plate is not thicker than 10mm.
On the Z however is good idea not to make it too strong especially if geared for some reason. Cause hitting the bed can do some damage especially with strong motor.

Hi,

Thanks for the info. I am thinking 3 to 1 on X and Y and 2 to 1 on Z as my Spindle will be large about 15 kg. I see they offer grub screw and key cutting service as well so will have key cutting on the servo side and grub screws on the ballscrew.
I understand from your info given ( 20:30) that you calculate the gearing ratio by the tooth count?

I see you are using on your new machine the west systems 105+209 wich I know is ideal for this application the only problem beeing that here in the uk the weather isn't ideal :) Do you think I will be ok with the 105+207 or 206?

Thanks again,
Vass

Quote:

---

Originally Posted by HipoPapi

Hi,

I understand from your info given ( 20:30) that you calculate the gearing ratio by the tooth count?
Vass

---

Yes, cause 1:2 , 1:3 or whatever seems not to reflect reality , misleads the mind. The reality is that 16t is the smallest one and geared 1:3 means 48t for the second one which is in fact 75.25mm in diameter and at 100mm separated center from center they have only 6t in mesh. So imagine the inertia if from steel. Even from aluminum.

I assume you will be using servos. So the encoder count + ballscrew pitch will dictate your gearing, as it will determine resolution, max speed of the machine.

Quote:

---

Originally Posted by HipoPapi

I see you are using on your new machine the west systems 105+209 which I know is ideal for this application the only problem being that here in the uk the weather isn't ideal :) Do you think I will be ok with the 105+207 or 206?

Thanks again,

---

No, it will be a problem. Its not only the time, the faster it cures the more heat generates, the more it distorts. Whats the problem with the weather in UK? humidity is not a problem if its not raining directly on it. And where i live its much more humid.

Hi,

I am planning to use 2x 750w servos on the X, 1x 750w servo on the Y and a 400w servo on Z.
The shaft on the 750w servo is 19mm dia and 31mm long so in fact that is what determines my smallest pulley.
I am planning to use 25mm HTD belts as my gantry will be close to 200 kg plus the belts are rated for -25 degrees Celsius.
http://www.mycncuk.com/attachment.ph...id=13398&stc=1
In the Case of a 5 pitch pulley for 25 mm belts F= 30.5 which means that that is my whole 31mm shaft. So if I would want to use the whole L which is 38 I would have 7mm of the pulley hanging of the servo shaft. If I want to use grub screw I would be limited by Dm and an even larger Dm if I want to use key. So I have decided to cut of F from L and use just F and key cutting therefore I can use a smaller pulley and I am limited only by (De) and have nothing hanging of the shaft.
So basically 24 t pulley on the servo where Dp=38.20 and 72 t pulley on ballscrew where Dp = 114.59 giving me
a ballscrew rpm of 1000.09 on the ballscrew with 3000 rpm on the servo side so essentially 3:1.
With a 2010 ballscrew my critical speed is 1128 so would be able to get 10m/m rapids or I can go with 2020 ballscrew and get 20m/m.

With the resin my worry is that here in the uk we are getting around 21-25 degrees Celsius during day time and it goes down to 15 during the night. But I did not look into it so deeply yet and maybe that isn't a problem with the 105+209. If By any chance you already know it please let me know.

Thanks,
Vass

Vass you cannot just rely on the Key you will need Grub screws as well to stop the pulleys coming off the shaft. You can buy pulleys with larger Boss so can get more screws in but also not uncommon to put the screws thru the teeth and provided they are below the teeth and deburred correctly it's not a problem.

Regards the servo's the rated Continous speed will probably be 3000rpm which is the Max speed they will provide continous rated torque but in practice they often will reach higher speeds at lower torque 4 or 5K is not uncommon.
Really thou you don't want to go past the rated speed/torque esp with a heavy gantry so your calcs are correct but with 3:1 ratio and 750W servos's you'll have much more torque than you need so won't be a problem within reason.
To be honest you could go lower on the ratio and increase the speed and lower inertia slightly at same time. Like I said in emails the critical speed isn't critical and you can run above it without any issues so 2010 run at 2:1 would give 1500rpm and 15mtr/min rapids which is more than enough. You won't be running above critical speed all the time and infact rarely will be for more than 2-3s at a time so it never becomes an issue. My own machine as been running 2005 screws with 1:2 ratio so double speed and well above the critical speed for 6-7yrs without any issues.
I would only use 2020 screws if the screw length was long and needed to lower screw speed to resist whip. In this case I'd go with 3:1 ratio to increase torque and make up for lower mechanical advantage from larger pitch.

Hi Dean,

I was thinking of the grub screw through the teeth option but wasn't sure thanks for confirming that.
What do you mean by the larger Boss?

Thanks,
Vass

Quote:

---

Originally Posted by HipoPapi

What do you mean by the larger Boss?

---

Boss is the bit grub screw would normally go thru or the difference between F and L

Hello Sir! Amazing work! :thumsup:
How did you adjust Y axis bearing blocks against the Y rails? Did you weld the hole assembly of 10mm lasercut steel?

Quote:

---

Originally Posted by toomast

Hello Sir! Amazing work! :thumsup:
How did you adjust Y axis bearing blocks against the Y rails? Did you weld the hole assembly of 10mm lasercut steel?

---

Thanks!
Yes, i welded the whole assembly, in fact all on the machine is fully welded except the Z assembly, which is spot welded almost fully-spot by spot each next to the other.


You mean how i squared the gantry? Y is my 2 long rails. After the epoxy they were on 1 plane. Then using straight edge and precision square i squared them. Then mounted the bearing blocks. Then mounted the gantry on top of the bearing blocks with bolts loose. Then using again straight edge touching the bearing blocks i squared the bearing blocks both sides. Then tightened the bolts that hold the gantry over the bearing blocks. So the gantry was now square.
Now have in mind all was was welded <1mm at least. Also the rails on the gantry were not drilled yet. Later i squared the rails on the gantry, fixed them with clamps and then drilled on place. Look at post #93

Actually I meant X-axis then.
I understand that you ensure with epoxy that the rails are parallel. But how did you adjust HIWIN bearing blocks to match the rails?
Surely you did not weld the laser-cut parts so precisely?
http://www.mycncuk.com/attachment.ph...id=14634&stc=1

Quote:

---

Originally Posted by toomast

Actually I meant X-axis then.
I understand that you ensure with epoxy that the rails are parallel. But how did you adjust HIWIN bearing blocks to match the rails?
Surely you did not weld the laser-cut parts so precisely?
http://www.mycncuk.com/attachment.ph...id=14634&stc=1

---

The machine when long rails epoxy is poured must be glued to the floor so you will not bump it later. Then long rails mounted. Then first i poured epoxy on the gantry top , the gantry in normal position as near as it would be in final setup. Then flipped the gantry 180 degrees and the leveled top epoxy now down resting on the already leveled long Y rails. Then i poured epoxy on the low gantry rail and on the legs. So now we had everuthing parallel, gantry legs, low rail and upper rail.

Then i squared the gantry top rail as described bellow. In short i squared a straight edge on the long rails. Then using 2 equal sized aluminum bars and checking them square against the straight edge i squared the top rail. Then additionally using only one piece at both sides i carefully finished the job feeling the scratching . Drilled and mounted the same way. See first picture
Now i had 1 top rail ready and square.

The bottom gantry rail.
I mounted the plates on both rails. The plates that screw to the bearing blocks. The Z was still only pieces, not soldered. Then via 2 clamps i mounted the lower rail more or less. But clamps not tightened very much. Then using the front Z plate that will be later soldered to the both plates that screw to the bearing blocks i checked and aligned perpendicularity to the squared straight edge resting on the long rails , picture below.

Note should be taken here that when i bolted the 2 plates from the Z i pushed them from behind till they touched the screws, so i know they are alligned to the bearing blockes, as the holes are laser cut. I mean because of this i am sure that their edges are parallel with the corresponding rails. Later i will have some additional opportunity to align them on one plane/the z face plate/ as i can move one of them a bit back


So i mounted the lower rail. Z was carefully spot weld fixed while all was bolted on its place. No paint anywhere on the machine where bearing blocks contact with frame, metal only.

Quote:

---

Originally Posted by silyavski

4. Bolting the gantry to the frame and aligning the gantry rails.

Needless to say - a lot of problems.

a/ First to say i have dry bolted the gantry before painting all. It was ok. So what was my surprise when everything was painted, the machine fixed on the welding jigs for comfort and the gantry will not fit. So i called a friend that helped me and 3 times i had to lift the gantry and lower it, mean while widening the holes. The reason. This is crazy- the machine is outside, the sun heats the frame and it widens. cause i made the machine in the winter...
So basically now i cover all the time it so it could not heat . My garage is small i mean to work inside.

So at the end i squared perfectly the gantry and bolted it


b/ Now next unsuspected problem. How do i fit the rails on the gantry to be perfectly square with the long rails on the frame? I knew i could do it somehow but in reality it was a big challenge.
I know how to do it if say i make 10 same machines, a fixture and no problem. But how to do it on the cheap. Cause this machine really ate a lot of money spend on little things.
After a lot of thought i fixed the straight edge perpendicular to the base rails and decided to start with the upper rail first.
Once the upper rail was square, there came the problem with the lower rail. A friend of mine says we have all necessary stuff always at hand. So i was in no rush, thought it a whole weekend and just have done it.
Another difficulty was that i do all alone at home with no help, so first i more or less squared it using clamps and then drilled and tapped 2 holes only. Then aligned and bolted until happy. Then drilled the rest of the holes directly on place and bolted. Photos speak better than words how all of this was done.

Of course i managed to chip the paint here and there so it seems when everything is working i have to retouch it at the end. No big deal.


2 equal pieces of aluminum square bar were used to fix more or less the correct distance to the upper rail.

http://www.mycncuk.com/attachment.ph...id=12401&stc=1


Both were precisely squared to the straight edge which was squared to the base rails. 2 pieces of rail were used to support them, a kind of parallels . Not seen on the pictures at the end only one was used at both sides so to make sure the fit is perfect and is really squared.

http://www.mycncuk.com/attachment.ph...id=12402&stc=1


Holes were marked, 6mm drill used to clean the epoxy so it will not unglue or raise during drilling the beam, then 4.2mm drill to make the holes and then tapped. With the help of the nice little block i have.

http://www.mycncuk.com/attachment.ph...id=12403&stc=1


On all drilling and tapping the gantry was clamped so it will not move around. here is the fixture which helped me align the lower gantry rail. Dont laugh at my square, had to clamp a rule to make it longer. I said we have all we need at home, just some imagination is needed.
Used thickness gauge though to align it perfectly /yes i love that word/.

Note should be taken here that when i bolted the 2 plates from the Z i pushed them from behind till they touched the screws, so i know they are alligned to the bearing blockes, as the holes are laser cut. I mean because of this i am sure that their edges are parallel with the corresponding rails. Later i will have some additional opportunity to align them on one plane/the z face plate/ as i can move one of them a bit back.

---

I swear I read this section many times before! But could not understand...
I guess my brain refused to believe that it is possible to fix the bearing blocks and then weld the parts together and end up with perfect alignment...
Even hammering some 10mm steel plate to perfect plane seems unbelievable. Guess it depends who is holding the hammer or welding torch :)
Sorry you had to copy hole page from your thread!

All the best,
Toomas

The key there is to leave possibility for alignment.

I mounted the plates so all bolts touched the holes from one side, and as they were laser drilled they were more or less straight. But later i drilled the holes 1-1.5mm wider. So at the moment of mounting i had not only the margin of the hole itself , but of the widened hole. So its easy to adjust cause there is space that allows it.

The other key is to have 2 precision squares and 1 straight edge. So at all times you can check and assure perfect fit. You can not measure how perfect, but if you insist and try that no visible gap is left at any place, you will finish with very precise machine. And not be lazy, cause you need sometimes to check a simple thing 10 times untill you are definitely sure its straight.

And when i say check, i mean check with strong led torch

Silyavski, do you think 10mm steel is minimum for Z axis?
I hink that I will at least try to build my Z axis from steel. If I dont succeed that I will go for aluminium :)

Steel is cheaper, faster and easier if you don't have a mill or other cnc. 10mm steel and that having in mind the way i did it, where the rails and the rectangular bars that the rails lay on, play the role of ribs that strengthen all. You will be surprised how flexible is 10mm of steel if not strengthened by bracing plates that provide strength in contrary direction.

I managed to mill with that Z/180mm fully extended/an aluminum plate- 1.5mm deep pass, and lack of cooling and proper bed fixture was the problem to go deeper per pass.

Hi dear friends, do u thing its good ide to use ballscrews RM2505-L=2600mm on both sides of my y-axis. Could it be too long and cause vibrations ? Or should i use R and G instead?

Quote:

---

Originally Posted by masinecc

Hi dear friends, do u thing its good ide to use ballscrews RM2505-L=2600mm on both sides of my y-axis. Could it be too long and cause vibrations ? Or should i use R and G instead?

---

Welcome to the forum. It would be better if you started a new thread with this.

You could use rotating ball nuts so that the screws remained static. ..Clive

Thanks for ur reply, would u like to post a picture from that rotating ball nut? Just to see how does it look like. Thanks !

Quote:

---

Originally Posted by masinecc

Thanks for ur reply, would u like to post a picture from that rotating ball nut? Just to see how does it look like. Thanks !

---

If you do a search on the forum you'll see that Jonathan uses them and so does Silyavski, they have pictures in their threads...

Great thread! plenty of useful information!
Cheers
Clayt

Quote:

---

Originally Posted by Skipsoaring

Great thread! plenty of useful information!
Cheers
Clayt

---

Great thread indeed!
I was wondering if there is a way to print this thread in its entirety? It is so long to read on the PC that I would like to print and highlight key info. Or maybe it can be saved as a file and imported into MS Word? Or maybe one of you has distilled it down to the basics already? I am open to ideas on this.

Randy

Just looking at this quickly - go to the top of this page - you'll see 'Thread Tools' which has a 'Printable' option that you can select. From the printable version you can print to a .pdf. You'll still need to do this for every page, though...

Quote:

---

Originally Posted by Wal

Just looking at this quickly - go to the top of this page - you'll see 'Thread Tools' which has a 'Printable' option that you can select. From the printable version you can print to a .pdf. You'll still need to do this for every page, though...

---

Nice to see you on again Wal. Also there is an option on the printable version to show 40 posts on one page (at the top right hand side)

Good call on the '40 posts' Clive. Yeah, I've been lurking a bit... I'm around one of the weekends in Feb (8th?) I'll give you a shout and maybe drop in for a cuppa!

Wal.

Hi,

I'm new to the forum but want to build something similar to this (very similar) I have read the thread a few times and am in the process of re-drawing the design in NX as I am not familiar enough with Sketchup to be confident releasing parts from it for manufacture. Also I think if I redraw it I will have a better idea of exactly how it all works and whether it will still be suitable for my requirements.

Could you please provide exact part numbers for the ball screws and linear rails you have used in this design. I have got the Hiwin catalogues but am unsure of the exact components that have been used. Once I know exactly what is used I can then start modelling these also.

Really enjoy reading the thread and have learned a lot already.

Thanks in advance for any help you can be.

Thanks,
Sy

Quote:

---

Originally Posted by f1sy

Could you please provide exact part numbers

---

You forgot to ask what he would do different if he was starting again :single_eye:

Quote:

---

Originally Posted by Robin Hewitt

You forgot to ask what he would do different if he was starting again :single_eye:

---

You can see that for yourself by looking at the next one!! Boyan's going to have built 6 or 8 machines by the time I've done one I recon!

The main difference is that i would have done all from aluminum plate, cause i spend too much time cleaning, , aligning, welding, painting, etc. When you draw the line is not that much cheaper and aluminum is more beautiful to look at.

Speaking especially about a machine made for sale , not for one for my self.

Querido Silyavski, have you some detailed info, or some thread that point me to the right way about how to level with epoxy top and bottom of Y Gantry? i think that wil be a little hard to get it perfectly leveled :/

Quote:

---

Originally Posted by samsagaz

Querido Silyavski, have you some detailed info, or some thread that point me to the right way about how to level with epoxy top and bottom of Y Gantry? i think that wil be a little hard to get it perfectly leveled :/

---

Read from here if you want to see all the troubles and experiments #84. Or jump directly to #121 to #126 to see how to be done properly on any size machine. There is no escaping buying the West System epoxy with the slow hardener. If you read all you will see i tried other epoxies with poor result.

Quote:

---

Originally Posted by Boyan Silyavski

Read from here if you want to see all the troubles and experiments #84. Or jump directly to #121 to #126 to see how to be done properly on any size machine. There is no escaping buying the West System epoxy with the slow hardener. If you read all you will see i tried other epoxies with poor result.

---

impossible to get that brand in Argentina :/ btw, reading your posts now, thanks for the link!

looks like some of your pics dont work anymore :( http://www.mycncuk.com/threads/6619-...7113#post67113

Quote:

---

Originally Posted by samsagaz

impossible to get that brand in Argentina :/ btw, reading your posts now, thanks for the link!

looks like some of your pics dont work anymore :( http://www.mycncuk.com/threads/6619-...7113#post67113

---

Then you have to download the technical sheet of West System 105 resin and 209 hardener, understand all details and search for similar stuff there. Or buy from USA. I bought from UK, in Spain it was veeery expensive