Problem with the X axis carriage movement (drag/stall)

Hi,

I post a new topic to avoid polluting the "Ormerod2 feedback thread" anymore.

So I have a broblem with the X carriage dragging while moving, leading to X steps from a layer to the next. I spent few hours yesterday trying to set the things right.

I disassembled and reassembled the X axis, the "stall" problem remains... on the whole course. I checked the belt tension, trying to tweak it, without any result. I got sure the motor gear was correctly aligned with the belt position (for it not to scratch against the printed part guide), I tried to tighten or release pressure on the bearing at the end of the belt, I verified that the bearing at the back of the carriage was effectively rotating, I put some light grease on bearing, cariage linear bearing, I put off the cable ties in case the friction rigidified it to much, tempering with the carriage movements... No results.

The only thing that would explain the behaviour as far as I can see is the return path of the belt being maybe to close to the axis structure, but I don't know how to put space between both except by adapting the motor's gear position, which I tested with no more result.

I'm pretty clueless now, so I think my next move will be to try switching 2 motors to see if another motor behaves differently, but if you have other ideas...

Matthieu

Yes it could be the motor, stepper motors occasionally have debris inside that causes them to jam.

Can you feel excessive or varying resistance when you move the X carriage by hand with the power off? If so, does the resistance disappear when you disconnect the belt? With the belt disconnected, there should be very little resistance. If the resistance remains, then perhaps you have the X runner bearing clamped tight against the X plate so that the X carriage is touching and scraping along the other side of the X plate. Or I guess the linear bearing could be faulty.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I don't know what "excessive" resistance would mean here because I don't have any other experience... When I disassembled the axis I felt no particular resistance in the linear bearing alone, but I found the resistance with the full carriage a little too much for my taste (though I don't know what a normal resistance would be), I *think* I also felt a slightly varying resistance with the full carriage while moving it by hand with the power off.
I have not tried to move the carriage manually without the belt so I don't know how much resistance it gives. I'll give it a try this evening but I tried to move the bearing a bit during my tests and it didn't change the behaviour (I made sure the carriage was not touching/scraping along the X plate during my tests, as I tought it could be the cause of my troubles).

A precision I forgot to mention (but given your answer I think it was clear enough): when dragging the motor does stall, it's not the belt that occasionally looses contact with it.

Looking at your print from the other thread [forums.reprap.org] , it looks like there is backlash in X - if there was a problem with the motor movement (motor problem or Duet stepper driver problem), I would expect it to be worse. Most likely cause is a loose motor pulley (tighten the grub screw onto the flat of the motor shaft), then that the bearing is sticking on the smooth rod (lubricate the smooth rod - [reprappro.com] ). Another possibility is that the belt is rubbing on the x-idler end; you can fit an extra washer between the bearing and the printed part, to space this out a little.

I posted the best way of testing motor, wiring and stepper driver function here: [forums.reprap.org]

Ian
RepRapPro tech support

Hi,

Thank you both for your advices. Good to know that there are options
I'll test it ASAP (but I have very little free time untill this week-end).

Quote
droftarts
... Another possibility is that the belt is rubbing on the x-idler end; you can fit an extra washer between the bearing and the printed part, to space this out a little...

belt rubbing, the visual





x-axis, notice Pointy's brilliant printable fix for that in this tread:
[forums.reprap.org]

Erik

I haven't seen (yet? ) that kind of rubbing on the belt so I can't confim the problem is the same (but I think there was temporary improvement when I pulled the belt away from the axis at the bearing level).
Problem with the printed part solution, brilliant as it may be, is that you need a working 3D printer to implement it :/

Quote
mat_fr
I don't know what "excessive" resistance would mean here because I don't have any other experience...

Just did a friction test of the hotend with a hook hanging spring weight, as far as I can see it takes a force of 600-700 grams to move the hotend with belt and step motor (motor current off)

This was with a 10kg spring weight so not very accurate, will find a more suitable weight with a finer scale and try again tomorrow

These are my motor current setting from config.g
M906 X1000 Y1200 Z800 E1200

Erik

I couldn't *not* test something so I quickly made a verification of resistance, tightened the motor pulley as much as I could (my allen does not perfectly fit ). It seemed better for a quick test.
As I don't have much time before leaving home I launched the snowman model, I supposed it was quicker to print than the ormaxis and it would do for a quick test.

It seemed alright untill the motor began to scratch like crazy, not moving as it should have. I don't know if it falls in the same problem or if it is now more "a PLA dragging the hotend" problem... It made a beautiful hill of PLA.
Settings (I suppose relevant) were M906 X11000 Y1100 Z800 E800
hot end 185°C, heatBed 65°C (default on the web interface, though pronterface is set by "default" to 200°C/45°C for PLA I think)

Quote
mat_fr
It seemed alright untill the motor began to scratch like crazy, not moving as it should have. I don't know if it falls in the same problem or if it is now more "a PLA dragging the hotend" problem... It made a beautiful hill of PLA.
Settings (I suppose relevant) were M906 X11000 Y1100 Z800 E800

Do you really mean X11000, or do you really have X1100 in config.g? The absolute maximum you should have is 1200. I find X800 Y1000 Z800 E800 sufficient. The Y axis needs a little more than the others because of the inertia of the bed.

Also, have you tried running test patterns such as circle.h and square.g? They don't do any extrusion, so they can be used as very quick tests of the X and Y axis movement.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

It's possible you have faulty wiring, and that one of the motor wires is making a partial connection. Probably the best thing to do is swap the x-axis and y-axis wiring looms over (they are about the same length) and see if the problem stays on the x-axis, or moves to the y-axis.

Shame, as your snowman looked pretty good up to then!

Ian
RepRapPro tech support

Quote
dc42
Do you really mean X11000, or do you really have X1100 in config.g? The absolute maximum you should have is 1200. I find X800 Y1000 Z800 E800 sufficient. The Y axis needs a little more than the others because of the inertia of the bed.
Also, have you tried running test patterns such as circle.h and square.g? They don't do any extrusion, so they can be used as very quick tests of the X and Y axis movement.

Sorry, I meant X1100 of course. I don't know if I need 600g of pressure to move the X carriage, but I feel it's way harder to move the X carriage than the plate itself (which I find surprising).

Quote
droftarts
It's possible you have faulty wiring, and that one of the motor wires is making a partial connection. Probably the best thing to do is swap the x-axis and y-axis wiring looms over (they are about the same length) and see if the problem stays on the x-axis, or moves to the y-axis.

Shame, as your snowman looked pretty good up to then!

I agree, very cute indeed
I'll check the wiring also.
What occured to me is that while the carriage seems rather difficult to move, every time I thought the problem was settled it ended up in failure after some time. So I wonder if the MOSFET is not trying its best to win against the drag but ends up turning in thermal shut down. Would it give that kind of behaviour or would it shut down completely for a while?

Edit: yesterday evening I didn't have the time to switch either the motors or the looms, just to add a washer between the X end bearing and the mounting point (X plate side). Once again it seemed to improve things significantly, with the carriage being able to go all the way to X=220 (~215 actually because it's the actual end of course), which is a great improvement as it only went to ~200 before, then was unable to go further smoothly.
I tried the circle and square gcodes, no problem at all.
But printing the snowman again, all was well up to the last 2 layers or so, the last layers were off a bit (no awful scratch eard, just a little offset that appeared on the last runs).

Edited 2 time(s). Last edit at 10/30/2014 04:50AM by mat_fr.

Hi,

I'm getting cautious about declaring victories, but after adding a second washer on the X axis end bearing (which released the drag a litte, I think), switching the X/Y looms (no effect), and finally switching the X and Y motors, it seems to work OK. As I got no issue on the Y axis I suppose the motor itself is not faulty, so I can only assume the X motor mounting was not sufficiently tight.

When testing the extruder began to feed eratic quantities of PLA, I thought I was becoming crazy, but it seems that all the manipulations ended up in untighting the nyloc nut on the extruder big gear.

And, finally, the snowman came up OK, what a relief.

Now I have to put on the shielding and close the duet box, test again the ormaxis and see if I can measure and calibrate the print. See that this week-end.

Thanks again for the help.

That's a perfect snowman, congrats! - You seems to have got the extrusion and retraction just right in the very first go! - me not so much

Your print..



.. vs my very first completed snowman of 6th jan. 2014



Erik

Thanks, for the extruder part I just followed the calibration documentation, but I can assure you that yesterday when the extruder screw got loose it was pretty horrible. So no, it's not the very first go
Trying to print the ormaxis again to calibrate the printer properly now. Hope it works.

Hi,

I was pretty busy these last weeks, but not away from the printer all the time. I got luck with some prints, and less with others. The X carriage problem seems less present.
But there's still an annoying problem with the first layer only it seems : the first layer external perimeter is doubled, but at one end of the X axis I have effectively a double perimeter, and at the closer end (motor's side) I have only one perimeter, both layers are superposed. As I used black PLA it's not easy to se on the picture... After that the layers seem to stack correctly, vertically. I tried to add some oil on every bearing, but it doesn't seem to be a mecanical problem this time. I tried other temperatures (I'm using 185°C/55°C), didn't change a thing. The issue is systematic and only for the first layer, at a time nothing drags on the nozle. I tried lifting a bit the Z 0 position as the nozle ended up dragging on the print due to the staked parts, no luck.

Now the print for which I join pictures is the latest and most successful (I made like 5 or 6 tests), and it's not easy to see the issue I'm talking about. So I suppose that given the result the question would be more "Can it get any better?"
This was only a test with a reduced width to test the screw holes' position and openings sizes. But as you can see in the last picture the print is not perfect and I'm affraid it might not fit in the box I intend to print also.

Another issue, much less of a problem, but a little annoying, is that the Z homing is never stable, I always have to recalibrate it, and the IR value to give it (G31 Z3.2 P579...) is never the same from one day to the next. Is that a normal behaviour?

I just tried to print the real, full model... It's a failure.
I just don't know what to do at this point. I followed the advices from the troubleshooting guides, but I can't get any better than this

If the displacement in your photos is along the X axis, then that axis is definitely skipping. The 3 most common places where a skip can occur are:
1) The belt is too loose and slips a tooth
2) The carriage or nozzle is being mechanically obstructed and the motor stalls, skipping a step or two (this is a common occurrence if part of a print curls up and the nozzle or another part on the X carriage hits it, for example.)
3) The grub screw in the drive cog is loose or not oriented to the flat on the motor spindle and is slipping on the shaft.

Less common causes are the firmware trying to accelerate the motor too much, causing a motor stall and missed steps. Early RRP firmware had a bug that would sometimes do that, but so long as you are using the latest firmware this will not be the cause. An electronics fault is another possibility (may result in low drive current for example), and another possibility is a loose motor connection that briefly loses contact every now and again. A faulty motor is the only other cause I can think of.

From your previous problem I would guess in your case it is likely to be (2) and your X carriage is meeting mechanical resistance somewhere. Maybe you did not find the root cause of that issue. If you slide the X carriage by hand (with the power off) it should move smoothly the whole length, and as a comparison it should feel and sound similar to, but somewhat easier to move by hand than the bed (Y axis).

Given the history, if nothing else becomes apparent and you have not already tried swapping motors, that would be the next step I'd recommend. Yes, I am aware that it is not a quick and easy thing to do, but I think that the motor needs to be eliminated as being the culprit. Besides, while you are disassembling & reassembling you can have a good check over the rest of the assembly, and the act of re-making that part of the printer might even fix the problem.

Dave

Hi,

I use the 0.78C Git version of the firmware.
I don't know if a mecanical failure/problem is involved, but I already swapped the X and Y motors/looms so it doesn't come from there. I don't think it's the belt tension either. It was relatively tight and I tightened it a bit more (afraid it might be too much after that).
The motor's grub screw seems tight too, on the flat. I suppose the X skips seem to be all in one direction, I suppose if this screw was loose it would go either way, not always the same.

The whistle gets out OK, it seems, so maybe it's a drag problem agains the fill of my model?

Quote
mat_fr
Another issue, much less of a problem, but a little annoying, is that the Z homing is never stable, I always have to recalibrate it, and the IR value to give it (G31 Z3.2 P579...) is never the same from one day to the next. Is that a normal behaviour?

Not your main problem, but to answer your question: Yes, I too see this every time I turn on the printer.

Quote
MikeyD

Quote
mat_fr
Another issue, much less of a problem, but a little annoying, is that the Z homing is never stable, I always have to recalibrate it, and the IR value to give it (G31 Z3.2 P579...) is never the same from one day to the next. Is that a normal behaviour?

Not your main problem, but to answer your question: Yes, I too see this every time I turn on the printer.

If you have an Ormerod 1 with the original proximity sensor, then it is very sensitive to ambient IR from bright sunlight and incandescent artificial light. To get consistent results, you need to shield it from those light sources. The modulated IR sensor supplied with the Ormerod 2 is a lot better, but not perfect. The differential IR sensor that is included on my hot end board (see [forums.reprap.org] and [miscsolutions.wordpress.com]) is even better, and does not need white targets on the bed.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Have you tried new Allen keys on the grub screws. My first set really didn't fit well (the 1.5 mm was actually 1.75 when I measured it). Got a better set and then only could I get the screws really tight.

I've got an Ormerod 2, so I imagine the IR sensor is good enough, I won't use the printer every day so I suppose I can live with calibrating the Z axis each time

The Allen key I've got might not be exactly the right size, but the screw seems quite tight. I'll try and get one just in case.

I seems also that the heat bed moved with all the tries. I had to set the bed plane manually again. It could very well be the cause of my latest issues as the plane was a little higher away from the motor side. Besides, I'm not sure the automatic bed plane compensation is to be trusted given the aforementioned IR sensor problems... I'm not too confident the value read by the sensor would be the same for each corner at the same height.

I'll give the print another try when getting home.

Thanks again for the advices.

Quote
mat_fr
Hi,

I use the 0.78C Git version of the firmware.
I don't know if a mecanical failure/problem is involved, but I already swapped the X and Y motors/looms so it doesn't come from there. I don't think it's the belt tension either. It was relatively tight and I tightened it a bit more (afraid it might be too much after that).
The motor's grub screw seems tight too, on the flat. I suppose the X skips seem to be all in one direction, I suppose if this screw was loose it would go either way, not always the same.

The whistle gets out OK, it seems, so maybe it's a drag problem agains the fill of my model?

Did you try moving the X carriage back and forth by hand (with power off) and comparing the feel with the Y bed movement? It should be consistently smooth with a similar drag felt on both carriages. There is always the possibility of a faulty linear bearing that occasionally binds or some other intermittent mechanical interference.

Dave

I tried. Before adding some oil (WD40) on the bearing and linear bearing the X carriage seemed smooth but heavier than the "Y plate", which seemed rather surprising to me.
I honestly don't know if there were little drags, it's possible I have felt slight differences at some random spots, but too slight for me to be sure it was a real issue or if I was imagining it. In any case the carriage was lighter to move after adding WD40 on mobile parts.

Hi Mat

Can you post some pictures of your X-axis? This might help us see the problem.

Another place that can cause excess friction is if the pulley is not aligned well on the motor shaft; the belt can drag on the plastic part, where it goes into the x-motor-mount. This will be worse one end than the other. Usually there will be black dust (from the edge of the belt) around this area.

Ian
RepRapPro tech support

OK, I'll try first to get a new Allen Key, tighten the grub screw as I can and try to relaunch the print with the plate plane set correctly when I get home.
I'll take pictures if the problem is still present. There is no rubber dust present as far as I can tell.

Hi Mat,

Quote
mat_fr
I tried. Before adding some oil (WD40) on the bearing and linear bearing the X carriage seemed smooth but heavier than the "Y plate", which seemed rather surprising to me.
I honestly don't know if there were little drags, it's possible I have felt slight differences at some random spots, but too slight for me to be sure it was a real issue or if I was imagining it. In any case the carriage was lighter to move after adding WD40 on mobile parts.

You shouldn't really use WD40 on bearings as it dissolves the grease in the bearings. A light oil such as 3 in 1 wiped on the shafts should suffice.

Pete

Quote
Peterdl
You shouldn't really use WD40 on bearings as it dissolves the grease in the bearings. A light oil such as 3 in 1 wiped on the shafts should suffice.

Oups
I'll try to find another more suitable oil then.

I just made another run. At first I thought it would be ok with the plate plane set correctly and the grub screw really tightened. The first layer externernal perimeter of the model seemed alright. But again, after a few layers it again wraps...
I join some pictures of the axis. Sorry for the poor quality it's taken with my mobile in poor light condition.

Edit: the new thing (the printing is still ongoing) is that the wrapping goes both ways (I made screws holes in the model and on the last picture we can see it going the opposite way as the side of the model)

Edited 1 time(s). Last edit at 11/14/2014 03:19AM by mat_fr.

From the photos, it looks to me as if your print head might be sagging, making the fan duct too low. Is the nozzle perpendicular to the bed? With the nozzle touching the bed, how much clearance is there between the bottom of the fan duct and the bed?

Edited 1 time(s). Last edit at 11/14/2014 04:00AM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].