X-axis skips. Out of ideas... please help

My X axis skips steps, giving me prints with misaligned layers every once in a while. On average this happens once every object. Sometimes twice or more, and sometimes I get lucky and a print finishes without faults.

I cannot really put my finger on when the probability of skipping increases. It seems more or less independent of the speed (feedrate or travel feedrate). My Y axis runs fine with the same microstepper board. The Y axis has a lot more moving mass.

I thought that switching to microstepping would decrease the amount of skipping because there would not be as much vibration in the system and AFAIK in general stepper motors can run at higher RPM when you use microstepping.

I played with increasing the power. Now the X-motor can hardly be stopped by hand when it's enabled. Only perhaps during fast 'erratic' movements (infill in small spaces) you could stop it. I suspect that these are moments where things go wrong.

I reduced the moving mass further by making a Bowden extruder. Now there's only the X-carriage, some acrylic, a heating element, the timing belt and a few other parts that need to move. No more bulky motor having to move back and forth, but still, nothing better! (see my blog for details)

I don't see the head bump into the printed parts, but this may be it. Yet there don't seem to be specific objects where it goes wrong more often.

This really is a problem that just doesn't go away... maybe you have some thoughts or tips of what I should try next?

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

I've removed the line in the firmware where it would disable the X-stepper. Not very efficient in terms of power usage, but this might be what's causing the problem... tested it, but it also seems to skip. I could actually hear it skip, and looking at the plastic pattern, it seemed to happen during the infill, but there was nothing special about the movements (just normal turns of direction). These turns do involve some (de- and) acceleration but it's nothing this axis shouln't be able to handle.

I'm now putting an extra capacitor on the PWM output that goes the heating element. Maybe that's a source of a power 'dip' that causes the stepper board to receive too little power once in a while?!

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

Are you sure the step pulse is long enough and the direction signal meets the stup and hold times. IIRC there is an RC filter on step so you need to take that into account. Also make sure the step signal is the correct polarity.

I had a similar problem on HydraRaptor early on and it was lack of setup time on the direction signal due to a firmware bug.

---

[www.hydraraptor.blogspot.com]

Thanks nophead, for your suggestion!

I didn't think of that. I got the enabled input inverted, because you need that to get things working at all. The direction and enable can be inverted from the parameters.h, but I forgot to invert the STEP/CLOCK pin!

I looked at the datasheet again:
[en.nanotec.com]

The delay between HIGH and LOW was 5 usec, but should have been from HIGH (default), to LOW (for minimum of 5 usec), back to HIGH.

I'm doing another print now to see whether this was it! It doesn't go wrong all the time, so I need a couple of prints to really have some confidence. At least this setting has never been correct and is now corrected! At least I don't thing that it will make things worse.

Keeping my fingers crossed...

Thanks again,

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

I was using a cheap extension connector on my X axis power connection, and vibration seems to work the ground connection loose over time. Usually manifests as skipped steps in the X axis. I was able to track it down by wiggling all the connections during a print, and watching for errors.

Good luck!

I got my hopes up, but then still one layer went where it shouldn't be. It took so long on such a big object that it almost seems that it is much better, but still not completely solved.

Step required a > 2 usec (microseconds) pause and got 5 usec.

For Dir, 120 usec is required if it changes. But this operation is in a different section of code, so I expected there would be enough delay. I will firmware it again with a 120 usec change. Not sure how often it would call that and if the delay would disturb anything, though. 16 MHz (16 MIPS in this case) is 16 instructions per usec, 16*120 = 1920 instruction is actually quite a lot. So I added the delay minus 1 per instruction I saw that would at least execute.

Up next is a taller object with little volume: the business card holder that just came up on thingiverse.com.

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

Another thing I would try is to swap the driver and stepper. Watch out for the connections too, as the X is moved back and forth by Y, while Y and Z are fixed.

1 hour ago I just had a faulty Z stepper connection (my Repstrap has Y axis moving the toolhead which includes the Z axis), causing my toolhead skipping a layer or two, making squeaking sound and crash into the object being print. Luckily my extruder are all metal and could survive some minor crash...

Thanks. It seems to skip much less frequently already. Today I printed 8 chess pieces and 4 whistles without any hickups!

I will also do what you suggested, sam0737, thanks.

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

Is the power to the steper controlers connected in a dasiy chain of power connectors?

If the are just try swaping the order on the dasiy chain.

Or run sepearte power leads from the PSU.

I have mine all connected on seperate power leads from the PSU.

---

Bodge It [reprap.org]
=======================================

---

BIQ Sanguinololu SD LCD board	BIQ Stepcon	BIQ Opto Endstop
BIQ Heater Block PCB	BIQ Extruder Peek clamp replacement	BIQ Huxley Seedling
BIQ Sanguinololu mounting	BIQ standalone Sanguinololu or Ramps mounting	Print It Stick It Cut it

My rep strap: [repstrapbertha.blogspot.com]

Buy the bits from B&Q pipestrap [diyrepstrap.blogspot.com]
How to Build a Darwin without any Rep Rap Parts [repstrapdarwin.blogspot.com]
Web Site [www.takeaway3dtech.com]