I would be very curious as to the results. Do you have the tools to accurately measure the difference by any chance?
Are you absolutely positive the pulleys on the steppers aren’t allowed to rotate on the shaft? That’s the only thing I can think of that would introduce slop into that system.
I’m pretty certain yes. I will test it again to confirm when I have a chance. However if it was the pully rotating it would not be springy and spring back so accuretly repeatedly. Either the whole stepper is rotating and flexing in the mount or the shaft is. I can easily check that to confirm though.
Bigger size, more weight, larger drivers, separate drivers, bigger voltages.
Is it possible the shaft itself is twisting without moving the stator? Seems hard to believe.
If that were happening I would think it would be a problematic stepper. However I tried two different steppers and both did the same thing. I don’t think it would have the springyness and return to the position when the load was removed? Seems hard to believe to me too but I couldn’t say for sure. That should be a totally solid connection no?
Please don’t take any of our interest or questions the wrong way, we all want to make a better machine and are simply trying to rule out all the common issues first.
At this point some videos would help best I think. Test methods and results. I know it is going to be a little hard to do but that way we can get to the right solution the fastest instead of us asking all these questions.
Just to be clear, what’s especially hard for me to believe is that the stator is rotating more than a whole step. The stepper has no reason to return if it moves that far out of place. It will just move to the closest aligned position, which would be farther away from 0. Something else is acting like a spring, or backlash.
Except you said it returns, right? So it’s not backlash.
Hey Ryan. I don’t take any of the questions negatively at all, I enjoy and appreciate all the participation. That’s what makes this fun for me, so don’t hesitate, and it’s very welcome! Please be assured mine comments and statements are all positive and in participation as well. For what this thing cost to build, it’s truly impressive. I just like learning and tweaking things.
It will be harder for videos as I’ll be doing it myself, but I will do the best I can in short order to get some new photos and video and clearer explanation of my testing.
.5mm movement will be hard to show visibly I think on a video but I’ll try.
Shoot skip the .5mm and lets tackle the 2.5mm first. Are your cuts still off? Like I said I do not have any unacceptable accuracy issues (I can’t measure any) on the parts I have been cutting.
We are all problem solvers and it turns in to a bit of a competition. Any second Barry will pop up with the obvious answer and make us all look like fools. Usually Bill gets me every time, but if Bill is here, eagle eye Barry is on deck.
Have you tried this with your beefy (looks since you don’t have specs) Zipties.
[attachment file=89581]
This prevents any springiness from a loop on an large cable tie.
Yes it returns. It’s definitely not backlash.
Yes it returns. It’s definitely not backlash.
I did some reading on microstepping and based on what I read it actually makes sense to me this can happen. The shaft can be rotated, especially when on a microstep until it lets loose when passing the full step. The loss of torque percentage with microstepping this artical shows is really high too.
They talk about loads rotating the shaft also especially with microstepping in use. This is all based on some reading and my interpretation. More testing and others verifying my results with the same test would be needed.
But 0.5mm is three or so full steps. So why would it come back to a previous whole step? It Doesn’t Make Sense.
That article is very misleading. The torque on one coil when at the first 1/16th step is 10%, but the other coil is still over 90%. Yes, It’s not as high torque from 0/16 tp 1/16 as it is from 0/1 to 1/1, but it is the same for 0/16 to 16/16.
Yeah, those micro stepping articles can be found arguing for both sides. The best I have found is some dude in his garage physically testing it and there is a slight drop in holding torque on a non full step but subtle, not drastic. What happens is your stepper would ratchet between the current step position (whatever yours is set at 1-16) it should not return. Think magnets, if forced it will move to the next position. If you do not cross more than half way it could come back but that is always less than 1/2 full step (so 0.9 degrees since we use 1.8 degree steppers). Full steps have resonance they hold so well they will spring back and forth, microstepping has merit when that is taken into consideration.
For us (16T 200 1.8 degree) if your steppers were moving it would clunk in 0.08mm increments, but not return. Any spring it could have would have to be less than 0.04mm or it would move.
I did not try that. That makes sense that it would really help. I didn’t have any of your zip ties left at the time with the yellow one, way beefier than the ones in the kit.
The numbers I have presented now are from the end with the zip tie tightened all the way with the looped belt pressed tight against the printed part and with the little 3d printed plastic in use just like your instructions. I thought that end would be much better but that is what I have tested now, and it’s using the cable tie from your shop and kit. It still has all that flex. Left side is that cable tie, right side is clamped as the picture showed. That side has no springyness left that I could detect at all.
I don’t think having both types of connection at the same time one in each end makes it difference, it just let’s one direction I push test one connection for flex, and then the other direction test the other connection. Clamp has .5mm (which appears to only be at the stepper pully) zip tieall the way snug has 1.3. I presume .5 is stepper, .7 is from ziptie.
I can’t try that extra zip tie as it is since it’s all the way tight. If you think it would be better to leave space and add that extra zip tie around the loop to compress it more ‘rectsngular’ I am happy to test that for you. But I doubt that it would be an improvement from cinched right down. The belt loop could also be adding flex maybe as well?
My current opinion is that the clamping is so easy and removes any possible locations of flex it’s ideal.
Also, the zip tie layout on mpcnc seems different than the Lowrider. Could that difference account for what I am seeing?
= spring.
No I have severely optimized the low rider cable tie path, no hard corners. That will be in the next MPCNC update.
Sorry dude, I’m at a loss as to why the stepper is going back to zero if it’s actually rotating that much. I was out in the shop fighting arduino to update the ramps firmware on the mpcnc to use the new shaft. 1.9 wouldn’t update the plugins, finally just installed the ldc drivers from the zip. After that I called it quits, by that time I was running the laptop with a remote desktop session from my desktop in the house. Had my heater running for 4 hours this morning and I could still see my breath in the shop. Can’t insulate the shop until I finish running new electrical for outlets.
I’d suggest against using the stainless zip ties. The edges aren’t exactly rounded over, they’ll cut into the belts. They’re also a pain in the ass to remove.
Actually I just thought of something. If the set screw on that stepper pulley is pointed(ish), and not flat, that could be where the movement is coming from. The pulley is just aluminum, so it has some spring to it. The side set screw could be loose, so the pulley will try to rotate around the set screw on the flat, since it’s flexing the aluminum, and not deforming, it will move back to center. Just a WAG though.
So I just went out to the shop to take a couple quick pictures, don't have time for anything else today in there, but the stepper shaft was not rotating at all now with a quick check. That is inline with what I originally thought should be the case. Everything was cold though. The other tests were done after it was running and the stepper drivers were hot. I would not think that would do it, but I would not knowm anyone able to chime in?
However. The zip tie vs clamp still showed the significant difference. I don't know if it is visible in this video, I took a quick one to try. Also a pic of each end of the belt.When load is away from ziptie, the other side of the belt goes very noticeably slack (when you can see the whole belt). When the load is the other direction straining the clamped side the belt stays taught. But I need to set up the dial indicator and strain gauge again for more precision measurements.
My videos are a few seconds but still too big to attach. What is the preferred method for sharing video here?
