HOORAH! I won!

Awesome!!!

Congrats, Lloyd. Not a bad prize! I hope it works as stated.

Please keep us posted. I used to be a big fan of GeckoDrives, but currently I only have one installed (G540 (4-G201s as part of an assembley)) and its probably been a year since the last time I used it. I would suggest keeping it at about 60% of rated voltage and currrent. It might last longer. My experience has been when you push G201s or G320Xs to near their ratings they don't last all that long. Haven't used any other GeckoDrives myself, but I've been thru a few of those.

Congratulations. How does the controller detect resonance?

I ask because I tried to write an assembly-language pulse coding routine that critically damped the stepping of a print head on a long cable drive without positional feedback to the computer, though I could see the head movement on an oscilloscope with an optical sensor.

All I "won" last week was two used garage door openers. They might be precise enough to paint lines in a parking lot, or position logs on my sawmill.

--jsw

What do you mean "awesome"? I thought he was talking about something real, like the Lottery (not un-claimed freight).

"Jim Wilkins" fired this volley in news:npeod1$lp5 $ snipped-for-privacy@dont-email.me:

Jim, I don't know, but I'm ready to learn.

On one of my customers' machines, I have a 4-chemical feeder/screener/weigher device that simply CANNOT be used in a fairly wide range of speeds that are all within the steppers' and drives' capabilities, because of the low force but really high inertia of the screening drums.

I 'programmed around it' in software, but I'd really like to gain back those intermediate speeds I lost.

LLoyd

snipped-for-privacy@hotmail.com fired this volley in news:f98f5bf4-0039-4141-bdca- snipped-for-privacy@googlegroups.com:

If he means what I do, it's 'real like the Lottery' for guys who design stepper-based machinery.

Lloyd

Slow the control loop down. It's trying too hard given the inertia of the screening drum.

I'm assuming that the screening drum cannot be designed to have sufficiently lower inertia, and that it's OK for the drum to take time to get up to speed, and to react sluggishly to changes.

Joe Gwinn

Lloyd - is that a continuous weigher or a stop & go type?

Hul

Lloyd E. Sp> "Jim Wilkins" fired this volley in news:npeod1$lp5

Hul Tytus fired this volley in news:npihlj$mgm$2 @reader2.panix.com:

Stop and go. And contrary to advice, it's already accelerating/decelerating the drums as slowly as necessary to avoid inertia-induced resonances at all speeds, except in a 'band' in the middle speeds. I can even 'jump' from the either end outside of the band to the other 'smooth' speed area in one step, and it will run smoothly both below and above... just not IN the band, no matter how slowly it's approached (from either side).

It's presently running on Gecko 201s, and Gecko themselves feel that the new 241 will fix the problem. I sure hope so!

LLoyd

The drives may not be the problem.

I can see the transition between rotating around the geometric center below the critical speed and around the center of gravity above it in the spin cycle of my washing machine, and on a paint stirrer in a variable speed drill. At the critical speed where vibration amplitude is theoretically infinite the agitator lurches sideways and the stirrer shaft bows out. Above it both settle down and spin smoothly again.

--jsw

This was my next thought as well.

The drum may need to be made far stiffer.

Joe Gwinn

Good luck with the new driver. Being able to skip over the bothersome speeds is a good position to be working from.

Hul

Lloyd E. Sp> Hul Tytus fired this volley in news:npihlj$mgm$2

Joe Gwinn fired this volley in news:240820160924339438% snipped-for-privacy@comcast.net:

It couldn't possibly made 'stiffer'. It's heavy material anchored securely at both ends with large bearings. There's no 'wobble' in this device. If you spin the drum to 3600 rpm, you can't even feel the imbalance.

That's not it. You'd have to know all the 'dynamics' in the system to fully understand, and by both contract and limited time, I can't describe it in enough detail to do that.

Lloyd

Hul Tytus fired this volley in news:npl533$mjo$1 @reader2.panix.com:

In operation, if it needs to go 'fast', it's accelerated from zero to the bottom of the resonance band, then the step rate is 'jumped' to the top of the band, and it continues to accelerate (with an obvious single 'bump').

Same thing winding down.

As you say, being able to manage that in software is a plus, but not having the problem would be "plus-er"!

LLoyd

Well, to be precise, if critical speed is the issue, the resonant frequency in bending needs to be raised. This is a combination of mass (lower is better) and stiffness (higher is better).

Well, if we cannot get the details, the magic suggestion may be a long time coming. Critical details are slowly emerging.

I would do the computation for critical speed, and see if it lands anywhere in the normal speed range.

Joe Gwinn

There is a thread that may be of interest: "Re: compensating phase shift" in newsgroup sci.electronics.design during August 2016. Lots on how to handle resonances in servo loops.

Joe Gwinn

Joe Gwinn fired this volley in news:270820161012557499% snipped-for-privacy@comcast.net:

Thanks, Joe. That might be useful.

And, THAT I can accept. The idea proffered that the assembly is flexing was so far afield I gave up on the ideas here.

LLoyd

Welcome, Lloyd. Also look at the parallel thread with a similar name.

Umm. Don't be so quick - the vibration may not need to be visible to mess with the servo.

Joe