Another Shifty Y-Axis (Page 1 of 2)

Yes, I have tried it on some other known-good small solid parts and the same shift effect occurs.  However, when printing a single walled cylinder that was 4" high, it printed flawlessly with almost zero banding and no shift.  But basically anything with infill gets screwed up.

My travel speed during prints is 70 mm/s, so I don't think this should be the issue.

To have that small of a shift seems a little strange...things to check and try:

retighten all setscrews in the drive pulleys (y rod and x,y steppers)

double your rapid move speed and see if it gets worse

Nothing else comes to mind that would cause such a small shift like that???

I have same issue but much worse shift. I did the alignment to for a week and get it work like you side small cubes etc but once a big parts starts it steps out every 15-20 layers. Now I think some is binding but I have seen the infill on small parts causes something.

I run at 30-45 perimeters and I think 45-50 for infill and 17mm for the first layer (never a problem)

Am I running too SLOW? I wonder if slower speeds do not have enough power from motor and accel where it causes some binding, a little more speed and accel will add some torque to push it through any slop. And this is after alignment and yes my voltages check out.

What are your Accel settings and speeds?

My feeds are XY = 500 Z = 5 Extruder = 45
Max Accel = XY 9000 Z 100 E 10000
XY jerk 20 Z .5 E 5

I also notice it occurs at the start of the layers as well. It does appear the head jerks when it has an X-Y shift to go from the infill or inner perimeter then jump to the outer perimeter or next layer it jerks when moving both direction it seems like a rough transition and the rounds make this work (like your vase) so squares are easier I guess.

Not sure why but maybe always there but now I am acute to the problem

I also notice this: My RH stops communicating occasional especially after sitting for a few hours or on long prints and screws things up. It stills is connected but the motors don't respond I have to remove the usb and go back again. i wonder if this is causing skips too.

I am using a MAC and a USB 3.0 port I wonder if this is causing issues?

Yes I had to do that after a long print. The board still does it, after sitting on (not running even) the motors stop responding like the port times out. I unplug the USB and it works. But the whole time it never disconnects it remains connected. Also recently it di the same but the motors no longer responded and the eerpom would not load the Mcodes all defaults were zeros. I had to manually type them in in Pronterface and M500 save then RH worked again.

What about USB 3.0 for MACbook pro think gives problems?

I read somewhere on macs you had to issue some commands in the CLI in addition to the power saving stuff... I'll see if I can find it in my history....

Ok, Found it ---> http://www.cultofmac.com/206291/keep-yo … os-x-tips/

The overheating was in response to losing the connection, not the skipping.

Overheating will cause it to lose connection was the theory.

I have made some more progress on this issue recently, although the end result is that the problem looks stranger still.

Yesterday, I completely re-lubed the system, re-aligned, re-tensioned, and sanded down the burrs/nibs off the y-axis rod, and the system rolled much smoother.  I printed off a couple things and they turned out well, so I thought the problem might be fixed.  To my disappointment, it was not.

The weird part is that the y-axis shift was completely repeatable in a certain part.  It is best explained in order of events:

-I printed a test run with two "heatsink" looking parts.  These parts have just a row of double-perimeter walls to easily detect y-axis shifts.

-The two heatsinks are arranged with one in the back of the platform, the other in the front.

-I examined them from the side, and found that the heatsink printed at the back had absolutely no y-shifts whatsoever.  However, the one at the front had a multitude of seemingly random displacements.

-I repeated the same exact test and examined the parts again.  Turns out, the exact same pattern of defects was duplicated on the "forward" heatsink part.  The attached images show this.

What conclusions can I draw from this, or what causes can I eliminate?

-It cannot be slipping of the set screws or pulleys.  It seems impossible that they would randomly repeat the exact same slippage and correct themselves when they go to one side of the stage.

-It has nothing to do with belt tension.  The two heatsink tests were conducted with very different settings of belt tension and still yielded an identical result.

I am growing increasingly suspicious of firmware or software issues, at this point.  One thing to note is that when I fixed my extruder, I uploaded marlin again with a different thermistor code.  However, I don't know if it is the same marlin version that I was using previously.  It says Lawsy's in the code, I think.  More importantly than what version I am using, what could the firmware do to cause this?

Could it be something like the steps_per_distance for y-axis is incorrect now, maybe?  So perhaps it discretizes it incorrectly when processing g-code?

Thanks for the help, everyone.  I think I am making some progress, at least...

Just a quick note about swapping the x,y,z,e plugs on the main board, You WILL damage the driver boards if there is any power going to them and a plug going to a stepper motor is removed.

So, make sure the printer does NOT have power going into it when switching around plugs.

I had same exact problem, very sensitive to any loading. I had loose tight etc. but after week of playing my Y stepper finally fried I suspect it is same for you. Reading around here it is over driven and over heats in the y at .521= 1.42A and the board at best even on web site says 1A is max without cooling and the motor is only rated to 1.33A The X is right under the max at .44=.88A at 1am motor max and the Z I assume is same

I am gonna replace with stronger board that handles cooling and load better and order a new y stepper that can handle more torque and at 1.2 only from qu-bd.

with more torque at the right Amp limits of the electronics this sensitivity to binding issue should go away and allow for some mis alignment. My carriage I feel is totally smooth and dialed in. with my current belt tension. Unless the tension is way off, when it is smooth, no power, you should be fine.

Then the only adjustment you should have with belt tensions it to adjust to get the perfect circles little tight/loose to dial it in. BUT if these little adjustments are causing failed prints randomly and over sensitive it is more then belts it is probably the hardware.

I have seen at least 4-5 threads here started by different users all with same symptoms I can not imagine all 5 of us can not execute the same circle procedure to get machine right. It must be failing hardware

I doubt that skipping would chew the belt up that much. The amount of tension for that to happen would load the bushings up or the carriage to bind and the alignment procedure fixes any skipping teeth and help them find a new index for the left/right sides of the carriage.

So after alignment should eliminate skipped teeth. But again the skip steps in the servo I noticed like you over infill and extruder head popping and the stepper driver could have been under or over driven and caused it and probably just crapped out and over heated.

I also noticed that my belt problems started after the Extruder Ribbon was popping on the Y belt on the left hand side after long runs for a few days. I think that pushed the rod to the left causing my binding. And after realignment I realized how sensitive that back rod needs to be in right position to keep the pulleys aligned.