Rails & stepper motors

[norego_masterbuilder]

Hey Trolls.

I've been doing a deep dive for building a DIY lathe in Brisbane, Australia. I'm starting to order the pieces to build a setup and one part in particular that I haven't had experience with setting up is the rails/stepper motor situation. I'm looking at several options in the links below.


https://sea.banggood.com/Machifit-150mm-Slide-Stroke-CNC-Z-Axis-Linear-Motion-Linear-Actuator-Engraving-Machine-with-Stepper-Motor-p-1124302.html?rmmds=pageNotFound__2&cur_warehouse=CN

https://www.ebay.com/itm/323883078491?_trkparms=amclksrc%3DITM%26aid%3D1110006%26algo%3DHOMESPLICE.SIM%26ao%3D1%26asc%3D235844%26meid%3D16879a00c39643eaa3d1b398c475e265%26pid%3D101113%26rk%3D3%26rkt%3D12%26sd%3D383830681010%26itm%3D323883078491%26pmt%3D1%26noa%3D1%26pg%3D2563228%26algv%3DDefaultOrganicWeb&_trksid=p2563228.c101113.m2108

I'm wondering what other people have used and if there are any options that have been used before that I haven't come across yet. I understand the amount of sound the motor driving the rails could be problematic as well. Are there any other things I should be considering for this part?

Cheers

[markrob]

Hi,

Both options look good. Personally, I like the Open Builds stuff because its modular, open source and easy to customize. What are your goals for the build? Are you trying for a level of performance equal to a professional setup, or can you live with some tradeoffs? The biggest issue is the coupling of the stepper torque ripple to the system. I have found that this is mostly an issue during lead in and lead out. You can mitigate that by physically de-coupling via shock/resilient mounting of the motor to the lead screw and also the plinth. I have experimented with adding some inertia to the lead screw via a external flywheel. The choice of stepper driver plays a big role. I don't think you need to go with a ball screw since any backlash is taken out once you drive the rail in its cutting direction. Also having some friction in the system due to a tight nylon nut actually helps dampen some of the motion. There aren't heavy forces applied to the rails as would be the case in a CNC application, so you don't need much torque. I see they both come with a stepper motors. That seems to be a drawback to me as I would rather select the stepper to be optimized for the driver and power supply. In either case, you can start with the assembly as delivered and test. I suspect, you will probably need to tinker a bit with the stepper mounting and coupling to reduce the noise level. Let us know how you do with the build.

Mark

[farmersplow]

I have already carried out a large number of attempts with stepper motor drives. The following findings:

1. small motor (Nema 17) = small maximum torque
larger motor (Nema 23) = larger maximum torque
Differences in quality and torque in different models!

2. Low supply voltage of the motor driver (12V) = small maximum torque & smoother movements
high supply voltage of the motor driver (48V) = high maximum torque & harder movements

3. The higher the engine speed, the lower the torque

4. If the torque is too low, the spindle stops or steps are "lost". If steps are "lost" (the stepper motor is activated, but it does not turn any further), then the whole system no longer knows exactly in which position the unit is. - It should be avoided at all costs. Here you can help yourself with a "counting wheel" on the stepper motor. This checks the movement progress.

5. It is important to take into account the translation of the spindle (1 revolution = 1mm or 2mm or 5mm). The higher the gear ratio (5mm), the lower the necessary engine speed and the higher the available torque.

6. But! If the spindle translation is high, i.e. one spindle revolution moves 5mm feed then: Then the individual steps of the stepper motor must be very, very small!

7. The more steps per stepper motor revolution you choose, the lower the available torque! - Yet again. Most stepper motors have a resolution of 1.8 °. That means 200 steps for one revolution. This is far from accurate enough when you consider that a spindle with a 5mm pitch then moves about 25 micrometers forward in one step.

8. A driver is therefore necessary to increase the steps per revolution to 400, 800 or 1600 steps per revolution (microsteps). Experience has shown that it is necessary to use the best possible driver. (None of my Arduino drivers made it). A good driver module also provides "softer" control. With a control of 800 steps per revolution, the feed rate improves (decreases). Then a spindle with a pitch of 5mm has a forward movement of about 6.25 micrometers in one step. However, with a lower torque. A good driver can counteract some loss of torque, but not work miracles.

9. The feed (axial movement per motor step) must be as low as possible because a stepper motor, as the name suggests, makes step-by-step movements (jerky). If the width of the cut is about 20 micrometers wide, then a deflection of 6.25 micrometers is already a loud sound! And every step would ultimately be heard as a deflection when played back as a sound. After a few attempts, I reduced the deflection to about 0.4 micrometers using a gearbox. That was then no longer to hear. However, a gearbox has the disadvantage that the drive can only be used in one direction when cutting (Rheinsche's filler writing method is then not possible). Another thing to consider is the control of a pause groove (between two songs) which require a much higher speed of the stepper motor.

Summary of my attempts:
Nema 17 motor too loud at higher speeds (5mm spindle)
and too weak (without gear). With a 1:50 gearbox powerful enough, but too slow.

Nema 23 motor loud at higher speeds but with gear 1:15 very quiet when cutting (33.45.78rpm), OK with pause groove, OK when moving quickly to the position (5mm / sec) and loud at 10mm / sec. (but possible)

No vibration transmission with a good driver module (the motor is only slightly dampened with rubber)

I have not made any attempts with transmission belts or gear transmissions.

I hope it is understandable (my english is not very good)

Nice greetings from Vienna
Thomas

[markrob]

Hi,

I had no problems with torque using a NEMA 17. The Trinamic driver I used is interpolated to 256X microstep. I had no problem at all using an Arduino to feed stepper pulses to the driver and achieve nice linear ramps and rock stable pulse streams. To make that happen, I take control of the timer hardware registers directly and do not use the stock Arduino libraries. I I/O ports that are tied directly to the timer. I use a feed screw with 8mm lead per rev. At 200 step/rev and 256 microstepping, the theoretical step distance is 156nM. In reality, that is not achieved, but the stepper motor current wave form is very close to a true sine, so the stepper runs quite smoothly. On top of that, the feedscrew RPM is so low at 100-300 LPI that any torque ripple is too low in frequency to be an issue. When you move at lead out speeds, the stepping frequency is in the audible range and you can hear it at a low level in the silent grove. This would not be acceptable if you are going for professional grade performance. But for me, it fine. If you want the top level performance, you will need to spend much more on a true servo drive and perhaps need some sort of gear shift mechanics to get good solid 100:1 speed range.

Mark

[farmersplow]

Hi,

..... At 200 step/rev and 256 microstepping, the theoretical step distance is 156nM. ....

Mark

Hi Mark, good results and perfect that you use the internal timer clock of the Arduino. 200 Steps/rev and 256 microsteps = 156nm and it means to have 256 microsteps of each step of the 200steps/rev! that would mean 51.200 steps per rev. what ist realy not possible with a Nema 17! I ran out of torque from 800 steps per revolution with my Nema17 an 3.200 steps with the Nema23. What I didn't mention is that I'm driving the entire turntable, not the cutting head.

When I did the first test with the Nema17, I was still using the standard Arduino libraries. It wasn't until much later that I also accessed the internal clock of the Arduino with the controller. But by then I had already switched everything to Nema23. After that I didn't try again whether the Nema17 would work.
Obviously Mark did it well with good hardware, internal clock controls and Nema17.

Thomas

[norego_masterbuilder]

Thanks everyone for the advice.

I'm looking to see what is available and I'm struggling to find a decent stepper motor driver in Australia for the Nema 17 or 23. I did find a full kit for the Nema 34. I'm imaginging that it might be overkill.

This is what is readily available
https://www.vevor.com.au/product/010367219430/closed-loop-stepper-motor-nema-34-12nm-servo-motor-hybrid-driver-1712oz-in

This seems closer to what I would need based on the advice above.
https://www.vevor.com.au/product/010330830849/nema23-3nm-stepper-motor-driver-servo-motor-stepper-driver-4-2a-hybrid-brand-new

This is an option if I can afford to wait until mid January.
https://www.alibaba.com/product-detail/Nema-23-2-phase-hybrid-step_1803796503.html?mark=google_shopping&seo=1

Are there any alternative stepper motor drivers that anyone could reccommend?

[norego_masterbuilder]

I found this one that is available nearby.
https://www.cnc3d.com.au/product-page/dm542-stepper-driver

[farmersplow]

Thanks everyone for the advice.

I'm looking to see what is available and I'm struggling to find a decent stepper motor driver in Australia for the Nema 17 or 23. I did find a full kit for the Nema 34. I'm imaginging that it might be overkill.

This is what is readily available
https://www.vevor.com.au/product/010367219430/closed-loop-stepper-motor-nema-34-12nm-servo-motor-hybrid-driver-1712oz-in

This seems closer to what I would need based on the advice above.
https://www.vevor.com.au/product/010330830849/nema23-3nm-stepper-motor-driver-servo-motor-stepper-driver-4-2a-hybrid-brand-new

This is an option if I can afford to wait until mid January.
https://www.alibaba.com/product-detail/Nema-23-2-phase-hybrid-step_1803796503.html?mark=google_shopping&seo=1

Are there any alternative stepper motor drivers that anyone could reccommend?

All three motors would certainly fit (even the smallest is still powerful enough).
Thomas

[farmersplow]

I found this one that is available nearby.
https://www.cnc3d.com.au/product-page/dm542-stepper-driver

The DM542 microstep driver is good enough. I also use it and it does a good job.

Thomas

[norego_masterbuilder]

Thanks guys for all the advice. I wanted to check about the voltage of the PSU. The specs for the Nema 34 are

60V DC 5.9A output.
AC input voltage range: 93~132V/176~264VAC.
115V/230V AC selected by switch.
High efficiency low cost.
Forced air cooling by built-in DC fan.
Low output ripple and yawp.
Over current, over voltage, short circuit and overheat protections.
215*115*50mm (L*W*H)

Meanwell Australia has an HRP-300-48 that can be wound up to about 58V. Would this be problematic to use? I've seen some kits including the Nema 34 with a longer generic looking PSU but I can't find the exact one. Would the loss of voltage be an issue or is that just splitting hairs?

[farmersplow]

I've never seen a 60V NEMA34 before. I assume you mean the stepper motor controller is 60V. In this case, check whether the control has "up to" 60 V or "fix". Normally the controls are designed for a voltage range "from-to" (e.g .: 12V - 48V or 24V - 60V). A higher voltage means that the motor receives a higher holding torque. (This often has a negative effect on the smoothness of the run. Higher tension = harder steps. Lower tension = softer steps). I tested my Nema34 with 24V and 48V and didn't notice any difference! Presumably, 12V would also be sufficient to carry out this drive (moving the cutting head). According to your request, 48V or 58V or 60V would be hair-splitting.

[norego_masterbuilder]

Thanks

[norego_masterbuilder]

I'm getting stuck into this and I'm wondering if anyone has advice regarding the coupling of the motor. I have a nema 34 with a keyed shaft that is 11mm x 15mm (along the keyed slot). Is there a better method for mitigating the noise bleeding into the cutting head? I've seen both designs with gears or a belt that have the motor mounted separately below and I have see others that use a direct coupling with the leadscrew which have the motor mounted along the rails. There seems to be quite a few options for couplers and some that have nylon in them to dampen the vibration.

[Dogtemple]

You need some sort of isolation, a belt would probably be most affective, that’s the route I have gone down at least. Depends upon your setup and how you want it all to fit together

[Mr.Inventor]

I too am about to build my own.
Everything you basically need outside of amazon

https://www.automationtechnologiesinc.com

You need a WAY bigger motor, over 1000oz torque to get what you want at full speed.
https://www.automationtechnologiesinc.com/products-page/nema-34/nema-34-12-dual-shaft-with-flat
More like that to drive a cutting table.

I'm coming at this world from a lifetime of printers and computer engineering.

Get dampers for your steppers and use belts to transfer the drive torque to your motion.
https://www.amazon.com/Zeelo-Stepper-Vibration-Dampers-Creality/dp/B07NVB2N2M/ref=sr_1_1_sspa?

It seems counter intuitive but the rubber dampers on the motor mounts themselves prevent them causing ringing in the structure they are attached to. Belts and dampers will isolate the motor, also set your steps much higher and engage micro stepping on the controllers to make less heavy winding noise.

[grooveguy]

A couple of quick questions for the 'stepper brain trust' if I might. In the cases cited here, is the stepper coupled directly to the leadscrew, or is there either a gear train or belt between the motor and the load? Also, regarding those Zeelo 'vibration dampers,' do they simply isolate motor vibration from what the motor's mounted to, or do they provide any sort of 'softening' of the rotational torque? Both questions relate to reducing 'sharp edges' from the inevitable cogging of a stepper-type motor... a low-pass filter inserted in series with the mechanical load. I was thinking of using a somewhat-stretchy O-ring rubber belt between pulleys, but was afraid that any stiffness in the leadscrew might translate to uneven groove spacing. Thanks for any experiences and advice.

[boryo]

A couple of quick questions for the 'stepper brain trust' if I might. In the cases cited here, is the stepper coupled directly to the leadscrew, or is there either a gear train or belt between the motor and the load? Also, regarding those Zeelo 'vibration dampers,' do they simply isolate motor vibration from what the motor's mounted to, or do they provide any sort of 'softening' of the rotational torque? Both questions relate to reducing 'sharp edges' from the inevitable cogging of a stepper-type motor... a low-pass filter inserted in series with the mechanical load. I was thinking of using a somewhat-stretchy O-ring rubber belt between pulleys, but was afraid that any stiffness in the leadscrew might translate to uneven groove spacing. Thanks for any experiences and advice.

About the zero vibration dampers - both peaces of metal are simply "glued" with some rubber. The motor is screwed to of them by 2 screws and the other 2 screws are connected to the base of your motor. Hope this helps!

Best,
Bob

[norego_masterbuilder]

I'm back at it. I've found so far that the NEMA 34 is harder to source accessories for and that the NEMA 23 might have been a more economical choice. I'm attempting to source couplings, dampening mounts and various other parts with a fair bit of difficulty. The motor itself is pretty hefty and the 60V power supply is pretty intense compared to some of the modular synth power supplies I have built in the past. I'll post some photos once I've gotten my setup together.

[dmills]

One thing I found is that any kind of recirculating ball linear bearing is the kiss of death, you really want something more in the line of a linear cross roller slide, but getting one long enough can be a challenge.
Same goes for ball screws, good for precision machining, but NOT the right thing in a recording lathe, the issue in both cases being NASTY torque fluctuations.

My current experiment has the carriage held back by a length of piano wire running over a pully with a weight hung on it, and the thing pulled toward the centre by the wire running over another pully and down to a winding drum driven thru a 50:1 harmonic reducer from a small servo motor.
This has the advantage that the servo can be mounted away from the lathe bed and there is no backlash because everything is held under tension.

Has anyone else been playing with 'epoxy granite' as a way to cast extremely well damped lathe beds and swan necks? Messy as hell, but it does work.

[jjwharris]

My first carriage was 3d printed - it had some crazy resonances, you would get 1/2 through embossing a 12" record and the carriage would start bumping up and down...I ended up printing it as a shell and then packing that was epoxy granite, it stopped it bumping around, but the heat from curing meant that it warped. I'm currently using a laser cut steel skeleton frame that's welded together