Y axis missing steps and running jerkey. What to test?

I have a DIY cnc router running GRBL 1.1 on an Arduino UNO R3. Been working fine for several years but yesterday I noticed that the Y axis (a 570 in/oz stepper) running a 3-1 T5 on T5 belt drive) was missing steps and acting jerkey. BAAAD PARTS. Grrr What could it be. I don't detect and impingement with the bearings. The belt does not appear to be slipping or skipping. I can actually see the stepper jerk. Bad stepper? Bad driver?? I don't have a spare motor but I could buy one.
Any help would be appreciated,
Thanks

Hi,
what drives are you using for the steppers?

I would guess that 570 oz.in steppers will require 5A or so, rather more than what I would expect from a GRBL board. Additionally what voltage are the drivers? If it 24V or 36V then you
can expect your steppers to crap out beyond a 100rpm or so. If you want them to go fast then you'll need over 60VDC drivers and 80VDC preferred.


Try reducing the max speed of the steppers (the G0 speed) to 1/4 and see what happens. Steppers lose torque the faster they go. If you ask that they go faster than the circumstances
(driver voltage, current, mechanical load) allow it will start missing steps if not outright stall. If reducing the speed helps then you know what areas to look at.

Craig

Quote:

---

Originally Posted by joeavaerage

Hi,
what drives are you using for the steppers?

The steppers and drives are from automation direct and are matched per their support.
5A on a 48 volt PS.

I would guess that 570 oz.in steppers will require 5A or so, rather more than what I would expect from a GRBL board. Additionally what voltage are the drivers? If it 24V or 36V then you
can expect your steppers to crap out beyond a 100rpm or so. If you want them to go fast then you'll need over 60VDC drivers and 80VDC preferred.

I built this machine in 2019 and it has been running with no issues since then until this week. I was at the end of a thousand or so piece run when it started acting weird. I get 400 in/min rapids and I was cutting with a .125 compression bit at 40 in/min.


Try reducing the max speed of the steppers (the G0 speed) to 1/4 and see what happens. Steppers lose torque the faster they go. If you ask that they go faster than the circumstances
(driver voltage, current, mechanical load) allow it will start missing steps if not outright stall. If reducing the speed helps then you know what areas to look at.

Craig

---

See imbedded responses.
I tried running the program above the work piece with no cutting and it ran fine. I turned on the spindle(2.2kw) and ran again above the work and again no issues. I even tried putting my weight against the axis while it was running and it never skipped. Only when it was actually cutting did it mess up.
Hope that helps.
Thanks

It could be radio-frequency interference (RFI) from your spindle leads. Try running the program "in air" above your workpiece without running the spindle. If it works perfectly, check the grounding and shielding on your spindle wires (if any).

I tried running the program above the work piece with no cutting and it ran fine. I turned on the spindle(2.2kw) and ran again above the work and again no issues. I even tried putting my weight against the axis while it was running and it never skipped. Only when it was actually cutting did it mess up.
Hope that helps.
It is a Chinese spindle and it was not grounded when I bought it but I grounded it according to a post on these forums. The spindle wire shield is also grounded.
Is there a way to test to see if the grounding is working?
I have confirmed that it is ONLY the Y AXIS.

Hi,
do you have either/and/or both a two stage EMI filter or line reactor on the AC input side of the VFD?
This sort of thing:

https://nz.element14.com/schaffner/f...sis/dp/1191381

The intent is to keep the electrical noise of the VFD, for which they are famous at producing copious quantities of, confined to the VFD and not 'leak back out' onto the local AC supply and upset
other more sensitive devices.

Craig

Yes I do but just to be sure can you tell me exactly how it needs to be wired. I am going down every path.

Hi,
its a line filter. The wiring is usually printed on it. It will have two input terminals, phase and a neutral, and two output terminals, a phase and a neutral.

Craig

Here is another possible piece of the puzzle. Ever since I built this machine I had problems with it randomly tripping a hard limit when it was no where near any limit switch. Got so bad that I eventually turned off the hard limits. Soft limits seem good enough as long as I home the machine before use.
Could that be another clue to this issue or are they independent issues?
Thanks

Ran some test cuts in particle board and it is still loosing steps on the Y axis. I can actually feel it jerk at times instead of being smooth all the time.

I can jog the y axis at speeds between 40 and 600 inches per minute and it never glitches. When it is cutting at 40 in/ min I can see the y stepper stutter. I can move the axis manually and can't detect any binding. I have seen no issue on any of the other 3 steppers( 2 on the x axis).

Hi TKMS - These are two different conditions. One is loaded, jogging is unloaded. For instance the pulleys are held to shafts with grub screws? If so this has happened to me in the past. If the grub screw is loose yet has worn a small groove or hole in the shaft it will jog and move by hand as normal. Yet under load it will behave badly. The only way to find this is to pull things apart and check. Peter

I checked all the pulley to shaft connections and I don't see any issue there. When I say I can see the stepper stutter I mean I have a flag on the motors back shaft and can actually see the stepper motor itself stutter, not just the axis. If there was some binding in the linear motion assembly would I not see that when jogging slowly?
By the way. My Y axis is using cold rolled steel plate and skate bearings. I have wipers on all of them so no grit gets under the bearings. Found out the hard way on that one. My X axis is 20mm linear rail. Lets not get into why I didn't do that on the Y axis. Just a dumb decision and I am stuck with it. I really don't think that is the issue.

Then disconnect the motor so its free and see if it stutters by itself. If it does not stutter then its a function of the linear system its driving. If it does stutter when free its a motor issue or an electronic issue. Unfortunately with this sort of fault you have to pull things down to basics and eliminate things one by one. Frustrating but you will find the fly in the ointment eventually... If the grub screw has worn a groove on the shaft it will not show up driving one way. Peter

if wires long enough you can also switch X and Y electronics and see if the opposing drive does the same thing, this will tell you if its the mechanics or not...

It only stutters when it is cutting. I can jog it all over the place at any speed from 40 to 600 in/min and it dies not stutter.

Thats why I believe it's a mechanical issue, you have to check everything and it will be a simple hidden thing... - Peter

Hi,
I unlike Peter think its a noise problem. It is electrical noise that is intermittently triggering you limit switch circuits, and that same noise could easily affect one or more stepper drivers.

How have you routed the spindle cable? You need to have it separated from signal cables as far as possible Where they cross each each try to ensure they cross at right angles. Any time
a signal cable is lying along side and parallel to the spindle cable is the opportunity to contaminate that signal.

Craig

Hi again. My spindle cables and signal cables are in the same energy chain. They have been since the beginning and I have always had issues with the limit switch trip. Only last week did the erratic Y axis start. So, I was thinking on it and I realized it started after I installed a new bit from a new set of 5(amazon). So, I swapped out the bit and installed a new one and immediately snapped that one off. So I installed another one and that one I used to cut 5 copies of the pattern that was giving me all the problems with NO ISSUES.
So now I am thinking bad bits and an axis that just cant push them.
Your thoughts?

An extremely dull bit could account for lost steps and jerky motion while cutting. It would also explain why the program worked okay with the spindle running in air. But the limit switch errors are probably due to RFI. Are those wires also in the same energy train as the spindle leads? If so, re-routing them might be worth trying.

Yes. All the cables are running in the same energy chain. How far away from the spindle cable do they need to be? Are there other options?
Thanks