Y axis issue?  

Re: Y axis issue?

I would check two things, if the belt pulley on the Y axis didn't slip on the motor shaft, Just move the axis with LCD a bit and then try to move the axis with your hand. It will either slip on the shaft or motor will skip steps.

Also glass can slide on the pcb, can you verify this didnt happen?

Re: Y axis issue?

Hi Josef

Thanks for a quick response.

The glass did not move; I have it positioned level with the front of the heated bed, and I checked it was in that position when I noticed what happened.

I turned on the printer and moved the Y AXis to engage the steppers. I then attempted to move the carriage by hand. It was very difficult, but possible. The motor turned - not just the pulley.

Peter

Re: Y axis issue?

OK, so it happened again, this time 3 hours into a 36 hour print.

This time, the Y axis skipped in the other direction. I was sitting close to the printer when it happened, and there was a strange grating noise, very much like when moving the Y axis with the motors engaged.

Again, the glass did not move, and the pulley is tight.

Moving the Y axis via OctoPrint: near the 0 position, something rattles quite a bit, but past the 50mm mark and through to 200, it sounds quite normal.

I also checked the belt alignment; this is almost perfect, with the belt running through the centre of the pulley.

I am just about to take the motor off and replace; somehow, I don't think it is a motor problem.

Peter

Re: Y axis issue?

So I changed the Y motor and it is now a lot quieter, but it has not resolved the problem.

In the slicer, I turned the model 90 degrees and now the shift is on the X axis. This is a very strange issue and I note that Scott has also reported a similar issue.

On this model the shift has now occurred twice on the same print at exactly the same height, in both X and Y axis. I have checked out the gcode and not found any reason for it to happen.

My next step is to use the same gcode to print the model using a different printer to see if it happens with that. If it does happen, then I will try a different slicer and test both printers.

The problem for me is that it takes 11 hours of printing before the shift occurs, wasting a good amount of filament.

Peter

Re: Y axis issue? [Possibly solved]

Hi

Just for reference. I split the part into 2, re-sliced and printed and the shift happened again at a similar height, but not as pronounced as previously.

Then I decided to try slicing with Cura. This time, the prints worked out fine. I had previously been slicing these parts with KISSlicer; I have to presume that this slicer created some gcode which instructed the printer to move the extruder out of bounds at some point, possibly because of a mistake in the model.

Peter

Re: Y axis issue?

Glad you got it to work.
I printed out 2 different walls of similar design as the one that shifted and they printed fine too.
I think you are right in that the slicer generate the shift. Might be the model in my case, sounds like KISSlicer did it for you.
Glad you got it though. Worse thing as after an 11 hour print to see an issue kinda hurts lol.

Re: Y axis issue?

Hi Scott

Yeah, it is a real pain and I have some 30 to 40 hour prints to do, so hopefully the issue will be gone now.

The weekend after I received my first printer, I was doing a 12 hour print of a T-Rex. After 7 hours, we had a power cut; I was scheduled for the village, but I hadn't noticed. That was the day I installed a UPS for the printers...

Peter

Re: Y axis issue?

OK, so it's not fixed. I sliced a part with Cura. Estimated print time of 12 hours.

I watched it for the first 4 hours and then had to go out.

I came back 6 hours later to find this:

This time the shift occurred in both X and Y axis initially at the same time and then in steps, with the Y axis being 50mm out and the Z axis 45 mm.

The Y axis skipped in the opposite direction to what it had skipped previously.

Don't know where I can go with this now; at least I still have my old clone printing...

Peter

Re: Y axis issue?

Ouch. Sorry Peter.

Re: Y axis issue?

If I remember correctly, you use the SD card to print. Is it possible the card or the reader is having issues?

Re: Y axis issue?

Hi Richard

No I use OctoPrint.

I am not wondering if I should revert to firmware 2.1 (before the motors were made quieter) to see if hat makes a difference.

Peter

Re: Y axis issue?

I have just had the printer on my desk to check it out.

The only thing I could find was that both X and Y end stops had been pushed back as far as they would go..

Peter

Re: Y axis issue?

did you ever get to the bottom of this? I am still on the first print with the bed working but I have noticed y axis has shifted as well, but no where near as much as yours.

Re: Y axis issue?

Right, here is a view from the top.

I was wrong earlier. The X axis shifted first, then Y, Y, X, X, Y, X and Y then the print was stopped.

@Bryan, I have just downloaded the previous firmware 2.1. I think I remember reading somewhere that the first "quieter printer" firmware went too far and caused axis skipping, but I can't be certain of that. I will have a look to see if I can find it.

And now, with 2.1, I realise why people were complaining about the noise...

Just wondering if I do a compare of the source files for 2.1 and 2.2 whether I will be able to find where the changes were made, and maybe go somewhere in between...

Peter

Re: Y axis issue?

Yes, I knew I had read it somewhere: http://shop.prusa3d.com/forum/prusa-i3-kit-building-calibrating-first-print-main-f6/quieter-printer-new-fw-t231.html

The original "quiet" release caused user marco.p to experience shifts.

I guess it's still happening; hopefully the noisier motors with more current will be happier. Will let you know in the morning.

Peter

Re: Y axis issue?

Hi guys

I have been doing a bit of research into the stepper motor current and I don't think that reverting to firmware version 2.1 will help.

The firmware source files (file: pins.h) determine what current is fed to the stepper motors via a PWM signal on 3 of the AT Mega pins. One pin determines the current for the X and Y motors, one for the Z motors and one for the Extruder.

The amount of current required is determined by the stepper motor; give it too mush current and it will melt. The stepper motors shipped with the i3 are capable of handling 1Amp, and so this should not be exceeded.

The current is limited with current sensing resistors (0.1Ohm) on the stepper motor driver, with the voltage drop over these resistors being compared to the voltage present on the AT Mega pins mentioned above.

On a Melzi controller board, a 1 Amp stepper current would be generated by a 0.8 Volt signal; the RAMBo appears to be slightly different.

Firmware version 2.1 and earlier outputs a current limiting voltage of 1V for the X and Y motors, 1.375V for the Z motor and 1.625V for the E motor. Version 2.2 has the following values: 1.125V, 1.125V and 0.675V

Therefore by reverting to firmware version 2.1, I have lowered the current to the X and Y motors and therefore they are more prone to skipping...

Now, this is where things become more complicated. There is a later version of the Marlin firmware which has default settings for the Mini RAMBo controller. In that version, all the current control voltages are set to 3.25Volts.

Taking all the above into account, I believe that the maximum voltage I can output to current limit the motors is around 1.625 Volts. In the morning, I will increase the X and Y motor currents by increasing the voltages to 1.5 Volts. Hopefully, that will not only prevent the skipping, but will also quieten the motors significantly.

In the meantime, I expect the print which is currently (and noisily) happening will fail overnight.

Apologies for the lecture...

Peter

P.S. Just found this in relation to the Mini RAMBo:

The A4982 is set to be limited to 2A. This means the adjustable voltage range is 0 to 1.66 volts. The following formula will yield the 8-bit binary value, Wv, to be entered into firmware: Wv = (VRef / 1.66) * 255

Still trying to work that one out! I think it may mean that values up to 2.4V are OK... But I really don't want to blow up another RAMBo.

Lastly for this evening, I have just noted that the VRef values output by the ATMega are fed through resistors which decrease the voltages by two thirds. So outputting 2.4V on the ATMega is right on the limit.

Re: Y axis issue?

you know far more then I do, it was hard to just follow all that but I'm learning, hope you can figure it out! my print as finished and it seems to have shifted back to where it should of been about 1.5 inchs higher tho

Re: Y axis issue?

Peter PJR:

Hi, did you copied the start g-code to the alternate Slic3r you use completely?

I set the acceleration in there with "M201 X1000 Y1000 E600"

Also you can try increasing the current. Latest FW is here https://github.com/prusa3d/Prusa-i3-Plus the settings for RAMBO mini are in variants folder and look for "#define DEFAULT_PWM_MOTOR_CURRENT {450, 450, 270} // {E,Z,XY}" This will definitely help.

Bryan:

From pictures it seem it is on the X axis, and id it is slowly leaning and not jumping a lot between layers, I would check if the belt pulley is tight enough 🙂

Re: Y axis issue?

Hi Josef

Many thanks for the reply. I was relying on the acceleration settings from the firmware; I will amend the slicer accordingly, thanks.

Now I have resolved the pins conflict, I will push the XY motor up to 450 and test.

Thanks for getting the latest source code uploaded.

Peter