Presto 6N/8N belt-drive conversion

[handcut]

I'm currently weighing-up options for moving away from the idler-rim drive, for various reasons. Conversion to direct drive seems to be prohibitively expensive and would mean modifying the lathes in ways I'd rather not, so I'm looking a belt drive setup with a BLDC motor driving a belt around the platter. I'm sure I saw a post on here - years back - where someone had done this to a 6N, but can't seem to find it now...

Anyway, I've been speaking to an engineer froma motor manufacturer. They seem to think that the high inertia ratio (calculated to approximately 24.5:1) between the motor pulley and the platter could cause oscillation.

Just wondered if anyone who has been down this road could offer any pointers? Even if it's 'don't bother'!

Cheers

[markrob]

A high inertia low friction system can cause stability problems especially if you try to push the closed loop servo bandwidth. However, if as long as you go with a simple PI loop and target a very low closed loop bandwidth (maybe 0.5hz or less), you should be fine. Most designers try to optimize the settling and response time, but in this application, you really don't care if it takes 5 or 10 seconds to settle to speed. You also do not need to have fast response due to external loads. The servo is just a slow correcting "auto Pilot" as far as speed control is concerned. The high system inertia acts as an effective filter to keep the speed constant against loads variation. I would make sure that the controller you go with allows for you to adjust the servo tuning parameters manually rather than relying on an auto tune method as these will typically try to get the fastest response out of the system. I went down this road with the Teknic ClearPath product line of all in one servo motors. They are very nicely designed and low in cost, but I they use a proprietary servo algorithm that is not suited to end user fiddling. The auto tune is the only approved way to tune and I found that at the end of the day, I could not get the wow and flutter down to an acceptable level. In fact, I'm doing better (although not perfect) with a cheap NEMA 17 stepper and Trinamic (see my other posts) driver as compared with the ClearPath.

[handcut]

Hi Markrob,

Thanks for this info. After some further discussion with the engineer, they've recommended the integrated system below, which is fully tunable. They've also said that oscillation can be avoided by keeping the inertia ratio below 20 if I reduce the small pulley. It's definitely not a cheap solution, but looks like a good starting point.

https://en.nanotec.com/products/1623-pd4-cb59m024035-e-01

[Jccc]

i remember awhile back seeing Flo modify a presto to make it belt drive.
he also installed a amp inside of it as well.

i cant seem to find the pictures or page where he had pictures of it .

[Teknic_Servo]

Hi markrob,

I’m an Applications Engineer at Teknic and I came across your post. Hopefully I can provide some helpful input here.

You make a number of good points regarding how relatively unique a record lathe application is for the motion control industry. In this case, minimizing the wow and flutter often requires balancing position tracking and disturbance rejection with relatively slow corrections. This scenario is in stark contrast to the more typical applications in the motion control industry that require fast corrections.

Couple these somewhat unique motion requirements with a mechanical design that includes a high inertia load driven by a compliant belt and it is not surprising that the auto-tuner did not provide optimal tuning results. The auto-tuner is optimized to provide high bandwidth, high precision motion control by using square-wave position stimuli, which works very well for the vast majority of applications. For your somewhat unique scenario, an advanced tuning procedure is most likely required to achieve the motion control objectives. We can elaborate more if you contact Teknic directly.

It is also true that the MSP software does not provide an interface for the user to make tuning adjustments. This is because most users simply do not have the experience to make good decisions regarding these adjustments. Making the proper adjustments can even be challenging for experienced servo users because we have some proprietary parameters that other systems do not have. Without proper insight, setting these parameters optimally would be at best happenstance.

If you want to try using the ClearPath on your vinyl cutting machine, you can contact Teknic directly for technical support. We have worked with other customers on similar applications in the past and there may be steps that we can take to help optimize the overall machine performance.

Please feel free to reach out to us directly by phone at 585-784-7454 M-F from 8AM -6PM ET or use our “Contact Us” form at https://www.teknic.com/contact.

Best regards,
Matt C. -Teknic Servo Systems Engineer

[handcut]

i remember awhile back seeing Flo modify a presto to make it belt drive.
he also installed a amp inside of it as well.

i cant seem to find the pictures or page where he had pictures of it .

Ah ha! Found it on Flo's site: https://www.floka.com/lofi/portable_lathe.html

That's exactly what I'm trying to achieve

[markrob]

HI Matt,

Technic support, went way above and beyond to help me try and tune the Clearpath servo for my system. At the end of the day, I could not get the results I needed. I will say that these are super nice, cost effective all in one servo drives for most any normal application. The auto-tune process is very well though out and even has the ability to set the aggressiveness of the tuning. For 99% of industrial motion control applications, this is a great product line. I was given access to the internal tuning parameters, but because they are not in the standard PID form and no documentation is provided, they are not very intuitive. I suspect that with the right settings, the servo might work, but at the end of the day, I had to move on.

[dmills]

Interesting, I was rather planning to use one of these for the carriage drive in my machine, thru a 50:1 harmonic drive reducer, which has a fair amount of friction to help damp the thing.
Time to look elsewhere i guess.

The number of servo drive system vendors who do not do speed rather then position as a first class citizen is rather aggravating, NSK looking at you.

Their Megatorque motors are cool but I am having to play with writing my own motor driver because the stock one just does not do high mass, low friction as a combination that works well for speed stability. It is tempting to just gorilla the bearing preload nut to increase the friction, and hence the damping, but that seems like an excellent way to trash a very expensive set of super precision bearing assemblies.

It is sort of tempting to just over excite the damn thing with low frequency three phase and let the rotor angle fall where it will, or pull something like the technics SP10 drive concept (But with more modern silicon - If I can buy it, another rant), pretty sure the SP10 style loop would work out ok.

[markrob]

It might work for the carriage drive. One nice thing about Technic ClearPath line is that they do have a 30 day eval trial. In my case, I let that pass. But I'm sure I'll find another use for them down the road.

[dmills]

Yea, but mine came from ebay....
Nice gear generally, but as you say this application is just a little out there.

I did actually contemplate a setup with a strain wave reducer (Yay for zero backlash!) actually driving the turntable directly, at 50:1 a reasonable mass of flywheel on the high speed side has 2500 times the stored energy it would have on the low speed side, and even a 400mm diameter mild steel disk is still well within the hoop stress limit at 3000RPM, bit scary because of all that stored energy, but it won't actually fail. It is the sort of thing that makes control loop stability iffy however!