ZAY7045 CNC SERVO

[RalcoBe]

Hello,

Recently, I bought a new ZAY7045 from a company insolvency.
The goal is to convert it to CNC, to make some hobby parts for myself and friends (Historical car racing)

I played with the live simulation in Linux EMC2, the other parts, I think of buying them at homeshopcnc.com

-Servo Z axis 850 oz/in 4200 rpm with 1/5 or 1/10 ratio, motor placed behind the column.
-Servo Y axis 850 oz/in 4200rpm with 1/1 ratio, motor placed under the machine and Yaxis.
-2 nema23 Servo's X axis 350 oz/in, 4700rpm, one right, one left, placed in front of the table, with a belt, ratio 1/1, with a cover over them.
Only 1 encoder used, DC motors in parallel. Two nema23 here instead of one nema34 gives me more place and will make the front servo cover smaller.

The servo motors will not be visible, the machine won't become larger, placing the X servos in front instead of behind the table nowhere limits X/Y movement.

The XY spindle ratio is 3mm/rotation, so 500 IPM is possible.
With the 250 lines encoder, I will need a 70kHz processor.
Z spindle ratio is 2.5mm/rotation.

I looked at the new Gecko G320X, with the pulse multiplier (1/10) I only will need 7kHz, this is no problem for EMC2 under Linux.
At 7kHz, the accuracy will be 0.03mm at 4200rpm (1/10) 500IPM or 0.003 at 420rpm (1/1) 50IPM. Good enough for me.

I measured the Backlash on X and Y. It's 0.2 mm. I won't be replacing the spindles at the first moment. Does backlash compensation on EMC2 work good?


Comments and remarks are welcome !



.

[SpeedsCustom]

I use EMC2 and swear by it. The compensation is pretty decent, and is simply placed in the HAL "BACKLASH = .02" , minus the quotes and not in SI unless set-up for SI units.

-Jason

[SCzEngrgGroup]

A 1:1 ratio on the Y axis will give you FAR more speed than you can possibly use - more than 100IPM. And why two servos on X? Why not simply use the same servo as Y and Z? I use those same 850 oz-in servos on all four axes of my knee mill (X, Y, knee, quill), with 2.5:1 reduction and 4-pitch screws, they are un-stoppable, and will run 400 IPM (on all but the 700# knee), which is twice what I can really use. If it was me, I'd go with the 805 oz-in motors and 4:1 reduction on all axes. That way you can use the exact same belt reducer assemblies on all three, with only the mounting adaptors being different.

Regards,
Ray L.

[RalcoBe]

The two little motors on X was an idea to make the front cover somewhat smaller,
I know, not necessary, but doesn't harm.

What do you mean with 4-pitch screws? Is it the same as 6.35mm/turn?
With my 3mm/turn in direct drive 1/1 I will have +- the same IPM as you have now.

I know you can't mill 400-500 IPM with my range of machines,
but on the other hand, does it have disadvantages?
Is it less controllable at, say 40 IPM?
Or does a higher reduction give a solution for a needed bigger precision?
For me, I don't need this (I think) , wit a backlash of 0.2mm.

[Starleper1]

That kind of rapid is scary fast for a home shop machine. Think about how much you actually will rapid, not to often. I had my machine set at 220IPM rapid and I turned back to 150IPM because at this speed I was more comfortable to rapid down to the .100" i typically use for clearance.

It is overkill IMO.

~Scott

[RalcoBe]

[Strike]everyone' s against me, saying 1:1 's not good[/Strike]

Nice, thank you all for advising me to take 1:2 :-)

This brings me to the Pulleys and Belts. I didn't really find the info I need with a search on this forum. Sometimes I order here:
https://www.misumi-ec.com/euro/egrme...300405510.html
https://www.misumi-ec.com/euro/egrme...300413640.html

I don't know anything about dis/advantages of the form factors like MTS8M, S3M, H, XL, Alu, steel, ... For the belt: Rubber, PU, no idea.

Altrough have studied in this domain, I prefer relying on the expierience of others and examples.
For X and Y, my technical intuition says me to use XL-type (pitch 5.08mm) 12 / 24 teeth (dia 20/40) with 19mm width as XL050. Anodized Dural.

Wrong?

.

[BobWarfield]

Scott, on a machine with more travel, like an RF-45 or a Bridgeport conversion, I'd opt for more than 150 IPM rapids if I could get them to perform reliably. It's amazing how long you can wait when working on a big workpiece to position things even at 100-150 IPM.

400 is likely overkill, but 200-250 would be about right from my perspective.

I too would not bother with 2 X servos. It would worry me that you create some problems for yourself if they don't work and play well together.

RalcoBe, if I read your post right, you plan to live with the 0.2mm backlash and just use backlash comp. I wouldn't do that. That's actually quite a lot of backlash for a CNC machine. Backlash comp will help, but it is just a bandaid. Getting a set of ballscrews on the machine will be worth it, particularly if you fancy the kind of performance you seem to be aiming for with servos running at 400 IPM. I think you'll be very disappointed without the ballscrews.

Lastly, make provision for a one shot oiling system as long as you are disassembling the mill to make the conversion. They really make a big difference, especially on a mill this size. It's very easy to add one:

http://www.cnccookbook.com/CCMilllOneShot.htm

Best,

BW

[SCzEngrgGroup]

The two little motors on X was an idea to make the front cover somewhat smaller,
I know, not necessary, but doesn't harm.

What do you mean with 4-pitch screws? Is it the same as 6.35mm/turn?
With my 3mm/turn in direct drive 1/1 I will have +- the same IPM as you have now.

I know you can't mill 400-500 IPM with my range of machines,
but on the other hand, does it have disadvantages?
Is it less controllable at, say 40 IPM?
Or does a higher reduction give a solution for a needed bigger precision?
For me, I don't need this (I think) , wit a backlash of 0.2mm.

Are you planning to use the stock leadscrews for this? If so, be prepared for your backlash to increase *rapidly*. And running very fast rapids will make it increase faster still. 400 IPM rapids will save you only seconds on a large job. I've never heard of anyone actually running such high speeds on a benchtop machine. There's just no advantage, and it greatly increases your odds of breaking something major. Things happen very quickly at those speeds, and you have almost no time to react. In fact, I don't even see the logic of using servos on a machine with such large backlash - You're putting high performance motors on what will be a low-performance machine. You'd be FAR better off putting some money into ballscrews. With 0.2mm backlash, you *will* have very visible artifacts in your work.

Running two servo motors on one axis also is not a great idea. Twist in the screw, flex in the mount, runout in the pulleys *will* have them fighting each other at times, which will cause you all kinds of grief.

Good machines are developed by measuring the performance of the base machine, carefullly matching the drive components to the machines performance and intended operating conditions, and following good engineering and system design practices. Just grabbing a pile of random parts, and doing no analysis, does not generally get a good result.

Regards,
Ray L.

[RalcoBe]

With the 1:2 ratio, I will end up with a maximum of 200 to 250 IPM.
When I use it there will still be the possibility to limit to 100 IPM or lower.

The leadscrews won't be replaced on the first conversion. Replacing the screws stays possible after that.
This conversion is mainly for myself to learn about CNC programming, Gcode, CAM and my 3D scanner. Time to put the theory in the real world.
It won't be a production machine. I beleave you when you say they will wear out, but if you know the amount of work/year I have for it, and the needed resolution, you'll understand.

For the servos, I surely don't want to buy too small, because I have a bigger mill, witch comes to to this adress when my garage is built. I's a real one, no toy like this one.
That one will eventually produce "sellable" parts.

Good machines are developed by...

asking the advise of people who know :-)


Anyone for the pulleys and belts?

.

[SCzEngrgGroup]

With the 1:2 ratio, I will end up with a maximum of 200 to 250 IPM.
When I use it there will still be the possibility to limit to 100 IPM or lower.

The leadscrews won't be replaced on the first conversion. Replacing the screws stays possible after that.
This conversion is mainly for myself to learn about CNC programming, Gcode, CAM and my 3D scanner. Time to put the theory in the real world.
It won't be a production machine. I beleave you when you say they will wear out, but if you know the amount of work/year I have for it, and the needed resolution, you'll understand.

For the servos, I surely don't want to buy too small, because I have a bigger mill, witch comes to to this adress when my garage is built. I's a real one, no toy like this one.
That one will eventually produce "sellable" parts.



asking the advise of people who know :-)


Anyone for the pulleys and belts?

.

I used GT2 belts in mine, and am very happy with them. Smooth, quiet, and amazingly strong for such skinny little things. I had to use XL on the knee, to get a higher ratio than I could get in GT2, to assist in lifting its 600# dead weight. All pulleys and belts from www.sdp-si.com. They have a nice on-line design tool for selecting pulleys and belts, and determining correct center distances.

You might be surprised how fast brass nuts will wear, even at "only" 200 IPM. Also, keep in mind that leadscrews have FAR more friction than ballscrews (as much as 10X), so performance (both "thrust" and rapids) will be very significantly affected.

Regards,
Ray L.

[RalcoBe]

The prices are really good at sdp-si, they have more choice and an easyer site. Nice.
At misumu there is an incredible expensive zero-backlash series of belts/pulleys.
Are they really needed in combination with the right ballscrews?

Everyone seems to have actions of ballscrews factories here :-)
You convinced me, I'm starting a search for the right materials...
Suggestions are welcome.

Another question: Servos are known to be better for Speed, while Steppers are better for starting and holding Torque.
Is it maybe a good idea to use servos for XY, and a stepper for Z? Who needs Z-speed?
My extrapolating, intrapolating and simplifying mind thinks I won't need a brake or counterweight in that way. Just one stepper driver instead of servo driver.

[SCzEngrgGroup]

The prices are really good at sdp-si, they have more choice and an easyer site. Nice.
At misumu there is an incredible expensive zero-backlash series of belts/pulleys.
Are they really needed in combination with the right ballscrews?

Everyone seems to have actions of ballscrews factories here :-)
You convinced me, I'm starting a search for the right materials...
Suggestions are welcome.

Another question: Servos are known to be better for Speed, while Steppers are better for starting and holding Torque.
Is it maybe a good idea to use servos for XY, and a stepper for Z? Who needs Z-speed?
My extrapolating, intrapolating and simplifying mind thinks I won't need a brake or counterweight in that way. Just one stepper driver instead of servo driver.

If you're going to be doing much drilling, especially deep drilling (>2 diameters), you *want* a fast Z axis. Also, at least with the current version of Mach3, you can't do helical moves in CV mode unless the Z axis acceleration and velocity are the same as those for X and Y.

Regards,
Ray L.