Strange problem with Y-axis working and not working

Hello all! I've got a really strange problem that I'm trying to figure out, and I'm hoping someone here can shed some light on it, or perhaps at least point me in the right direction.

The short version is that the Y-axis won't work right. The long version is that it works (albeit incorrectly) using a 6-month old version of Sprinter firmware, and doesn't work with the current version of Sprinter or Marlin (and other firmwares I've tried don't seem to work at all). It doesn't (on the surface) appear to actually be an electronics problem, because I can (sort of) make it work with a certain version of Sprinter, but if it's a firmware problem I haven't been able to figure out what the issue could possibly be.

The RepRap design is a Prusa Mendel with Sanguinololu 1.3a electronics (Atmega644P) and mechanical endstops.

The long explanation:

I initially loaded Sprinter a352585 ( [github.com] ) and using Printrun/Pronterface ( [github.com] ) I verified that all axis and sensors appeared to be operational. When I attempted to print something, it just kind of ran the Y-axis bed into the endstop and sort of attempted to print half of the Reprap minimug (it managed a bit of a half-circle, so the Y-axis wasn't completely stationary).

After messing with it a bit (thinking that I had the axis set up backwards or something), I tried running straight G-code, and discovered that so long as only one axis is being controlled at once (ie, "G1 X20" or "G1 Y20"), it works fine. But if I attempt to run both axis at once ("G1 X20 Y20"), only the X-axis actually responds; the Y just sits there. A call to M114 shows that the firmware believes that both axis have moved (says X:20.00 Y:20.00). This would seem to suggest a current problem, however I can run X, Z and E at the same time ("G1 X20 Z20 E20") without a problem, so unless X and Y are run off of a different current rail than Z and E, I don't see how current could be the issue (and boosting the current control pot on the Y-axis Pololu doesn't seem to make any difference). I also tried switching the X and Y motors and wires to make sure it wasn't something wrong with them, and it does the same thing (just with the other motor).

So, I thought maybe it was a firmware glitch of some kind. After all, it was working if I ran the Y-axis by itself, and I couldn't see why it would be a current problem if the other axis could work together okay. So I tried the most recent version of Sprinter, and also the most recent version of Marlin, and the Y-axis just doesn't work at all with them. I can hear a small noise like it's trying to work, but nothing happens. (I also tried Teacup; it didn't work at all, and I've had problems getting anything else to compile).

I have also tried swapping the Pololus for X and Y axis (no difference), and also using a stepstick I had available from a previous attempt on the Y axis (still no difference). I've also messed with the endstop settings, checked (and changed and uploaded and re-checked) the settings in the firmware configuration file, and that didn't do anything either.

I'm out of ideas. At this point my best guess is that it has something to do with current, but that seems iffy. If anyone has any ideas for anything else I can check or look into, I'd sure appreciate it. Thanks!

Is your tension on the Y axis cable to high?
Can you smoothly move the y axis by hand when the motors are off? Does the Y axis catch on anything?
Are you using linear bearings or bushings?

You stated that swapping the X/Y motors caused the issue to move to the other axis. perhaps ther is something wrong with micro stepping.
Microstepping is very dependant on the current and torque. Try adjusting the jumpers to not use microstepping. this will hurt precision in the long run, but can be a good check in the short term.

Thanks, but, sigh... no dice. Changing the microstepping sounded like a good idea (I hadn't thought of that), but it didn't make any difference. I tried it at 1 (no jumpers) and 1/8 (jumpers 1 and 2), and also adjusted the AXIS_STEPS_PER_UNIT settings in the firmware when the X-axis didn't work right after changing to 1 (I just divided the default settings by 16, and then divided by 2 when I reconfigured for 1/8 microstepping), but it's still doing the same thing. X works, Y doesn't, swap X and Y hookups and then X doesn't and Y does.

The Y axis belt is about the same tightness as the X axis belt, but I took it off just to make sure. No difference. And again, plugging the Y motor into the X hookup and the X into the Y hookup causes the problem to manifest on the X axis and the Y axis moves fine. So I don't think the motors or wiring are the problem... it's got to have something to do with the Y axis hookup/pins/current/microstepping/whatever.

The axis slides on bushings (Prusa design), but again, I disconnected the Y axis belt and the motor still doesn't move when it should (and moves fine when I swap X/Y motors), so I don't think it's a torque/resistance problem.

Any other suggestions?

Perhaps the Y step pin is not connected or shorted to something else. Check continuity between the pin on the micro and the pin on the Pololu.

---

[www.hydraraptor.blogspot.com]

I checked the Y step pin continuity as suggested, it appears fine. Also checked Y dir, it seems fine too. I'm not sure how I could check for a short.

Again, the motor DOES work, and it works CORRECTLY (at least it sure seems to), but ONLY with Sprinter a352585 and ONLY when doing a single-axis move. That's what's so weird. If it was a simple electrical problem, I would expect that it would not be working ALL of the time.

Any other ideas?

Look at where the tracks run and look for shorts to adjacent tracks. If what one axis does is related to what another does it seems like there must be some erroneous connection between them.

---

[www.hydraraptor.blogspot.com]

I visually checked the board (looks fine on the surface), and I used a continuity tester to see if I could detect a connection between the Y dir and Y step and any IC pins that it should not have a connection to, and I couldn't find anything.

Also, I'm pretty sure I checked and it wasn't just running Y along with the X axis that was the problem (ignoring for the moment the fact that it doesn't work at all on the newer firmware versions, with or without running X at the same time). I'm pretty sure I tried G1 Y10 Z10 and it did the same thing (moved Z but not Y).

Thanks for your help so far! I take it that your opinion at this point is that there is something wrong with the board? (That's certainly where I'm headed.) I was really hoping to avoid getting yet another board (already had problems with my first one), but if that's what I need to do, then I'll just have to do it.... Any other ideas I could maybe try before I do that though?

Nophead, I think you've found the problem! Ignore my last post (I'm an idiot)!

I just discovered while reading around looking for ideas that my multimeter has a beeping continuity tester (yes, I'm that new to all this; I thought the little 'audio' icon had something to do with testing buzzers). I had been using a $3 1.5v light-bulb tester mostly likely intended for house wiring that wasn't nearly sensitive enough.

Using the PROPER equipment, my multimeter detected a 75ohm path between the x-dir and y-step pins on the chip. So it's not the board that's bad, it's the chip. Luckily I have a boot flasher from screwing with my last board, so I probably just need to get a new chip, load it up, and I should be good.

Thanks for all your help you guys! I'll post here again and let you know if it works after I get the new chip.

Quote

I had been using a $3 1.5v light-bulb tester mostly likely intended for house wiring that wasn't nearly sensitive enough.

If by that you mean a 1.5v battery and a light bulb, not only would it not be sensitive enough, it could also damage semiconductors. Light bulbs have about 10 times less resistance when they are at room temperature compared to when they are white hot, so take a big surge current when starting. Also a 1.5v bulb probably takes 100's of mA to get a reasonable brightness, so when multiplied by ten that is a lot of current to put through a signal pin of a micro.

---

[www.hydraraptor.blogspot.com]

Quote

If by that you mean a 1.5v battery and a light bulb, not only would it not be sensitive enough, it could also damage semiconductors.

Thanks for pointing that out, I hadn't considered that (near-total electronics newbie here, software programming is my normal thing). I don't think that's what caused the damage, since I didn't start attempting to use it to test the board until after you suggested doing so, but I definitely won't be using it near any components again (I originally got it to test that my motor hookup wires were attached to their plugs correctly, which I'm sure it's fine for so long as the wires aren't actually plugged into anything).

Edited 1 time(s). Last edit at 06/25/2012 09:30AM by jdrgoldeneagle.

Yep, indeed that fixed it (unsurprisingly enough)! Got a new chip and now the "G1 X10 Y10" moves both axis at the same time, as it should.

So, for anyone else who comes across this thread, there were two lessons learned by me:

1. If your device is doing something weird, try checking circuit continuity (such as between pins/pololus and pins/pins), and make sure that there isn't a short of some kind between two different pins (such as a 75ohm circuit between the Y-step and X-dir pins in my case).

2. Don't use a cheap lightbulb continuity tester, they could cause damage. Continuity testing is what the little audio/waves icon on your multimeter is for. (This is for dummies like me who didn't read the manual for their multimeter.)