X-axis lost steps. (Page 1 of 2)

Well made it worse tonight. not the idler pulley is clicking while moving and steps are screwed up for the entire print.  I could really use some more help then just support saying redo the x/y calibration that says nothing at all about the X-axis.

Reading other threads people talk about the clicking, and mention set screws in the idler pulley.  But I see none in any of my pulleys except the ones that are driven, is this normal?  If not what Size do I need?

Whats a good tension for x-axis belt?
How do we check tension?
Do I need set screws in the idler pulleys?

1, y rods are not lose, push pull on them nothing changes.

2. tried with belts tight and lose, still clickgin on small/slow x-axis changes.

3. groans from the right side y carrage when moving through the middle of its range.

3. good to know on the pulleys its what I thought.

Here is some video of the print.

So spent most of my spare time today tweaking and retweaking, lubing, readjusting y-axis, x-axis belt tension, and got back to getting prints that are only shifted at the layers with lots of x-axis travel.  Once out of the areas withalot of x travel the part prints fine.  seems to go away after the bottom 2-4mms are done printing on most prints. 

Checked voltage on the x-axis driver was a little low .438, tweaked to .444 and still having same issues. The biggest puzzler is the drift always shifts away from the x-axis stepper, and stops part way thru the print. tomarrow going to add an electronics cover with fans to cool the drivers and see if that does any good at all.

Well I added my cooling fans and same issue.  Moved the driver boards around between the x-axis and y-axis same issues.  prints start out losing steps in the x-axis direction, they striaghten up after the firs 2-4mm.

Here are some pictures of the fan mount, and the results from moving the driver board.

And here I highlighted in blue what is going on.

how do you calibrate the x-steps?  no one has ever told me or shown me a link to do that.

This happens on all parts weather they be small circles or big sqaures or whatever shape you can think of.

I will give that a shot, but that does not explain why the shifting goes away after the first 4 millimeters.

The picture looks like it might be variation in shrinkage that can happen in lower layers near the bed.  Try turning the bed temp as low as you can and still stick, maybe 80.  Be sure you use the same location on the bed and see if it makes any difference.

Well I got my SD2 up and running after tension and replacing both step drivers with SureStepr. and made a huge difference but still had shift but now in the X. Last thing I did which fixed it the most. Was cleaning my nozzle? Not sure why but after still have shifts in Y and X with new drivers and tension I cleared my nozzle, which didn't appear to be clogged. I heard some popping. SO try cleaning nozzle. Of course I went back and leveled bed and cleaned everything again and checked voltages with both POWER and USB plugged in that made a huge difference.

So after all that and calibration I had good runs on larger diameters parts within .1MM at 50mm or 0.001" on 2" parts which was good. But on small diameters I was still .1mm-.2mm off and some with shift.

I think it is now backlash, maybe too loose, but hate to play with it more and make it worse. But I also noticed it like you at the top of my prints. Looks good but then shifts near the last few layers. to the right.

I think it could be small diameters and lack of cooling so the head drags a little. I turned off retract on new layer in SLic3r .99(new feature) which helped. But I am thinking that with small parts and my speed/accel I am getting overshoot.

Anyone have good accel numbers? I am running the stock 9000. I am thinking they are too hi and getting overshoot?

Well I say the top. But maybe I am looking at it wrong, could be bottom is over and adjust after about 30 layers and then goes straight too? In small circle case the last few layers shifted to the right very obvious. It almost appeared like the plastic was still molten and at the end of the print as the head finished the last layers it deformed and pulled the wall with it a little. no drag marks.?

I noticed this on the Z wobble cal cubes. Thin 1 wall prints at the top on one side after complete I went to big up and it appeared the one wall sagged in deformed. I printed again and it did it on a different wall and I realized it looked like my thumb. Like when I went to ull off my thumb pressed the wall in slightly. Again maybe not cooled yet? Wonder if on small prints I am seeing this because I am printing too fast and or accel?

I have a 2" diameter port I print for test, i can post the stl. On that part it comes out dead nuts with nothing wrong.

Well if not tension or cooling as you have that covered. Didi you clear the nozzle to be sure then last I would maybe try the USB voltage, Try to go direct off PSU power and see if voltage is fluctuating?

rincewind just posted where he had voltage problems using the USB. USB and PSU both share the same 5v rail. So I think if the PSU fluctuates (which form the PC it easily can when it goes to sleep gets busy or taxed with cpu loads) It effects the voltage. The voltage is critical to the stepper outputs. And easily can cause shifts.