Best solution for UMO with 'end of step' uneveness; TLSmooth, ROSC jumper, SSSDriver?

[mastory 44]

I’ve been doing a lot of reading regarding the uneven X-Y travel caused by micro-stepping at the end of full steps, and am trying to determine the best solution to address the problem on my UMO.  To test, I printed out @Torgeir 's test piece and indeed I have the lines caused by the 'end of step' uneveness.  I've seen it in other prints forever, but just accepted it.  I’m now confident it can be improved.

I've read quite a bit about 3 different solutions, and they are all fairly cheap; TL-smooth adapters as discussed in this thread , tying the stepper driver ROSC pin to ground discussed here, or by upgrading to SIlentStepStick Drivers as discussed here.

After all the reading, I am uncertain if the upgraded SSSDrivers have the skipped/uneven step issue at all.  If not, this would be a good way to go, except that the motor voltages need to be adjusted, and motors reversed via firmware or wiring.  The work here would be a stretch for my capabilities, but when facing a choice, I prefer simpler solutions.

The ROSC jumper to me would be the ideal solution, but does require some deft soldering.  What I have not been able to determine is, which driver chips the UMO's shipped with.  It has been mentioned with measure of doubt that they have A4988 chips or possibly A4983.  The data sheets indicate that the two chips do not have the same function in regard to the ROSC pin.  By my reading, the jumper won't work with the A4983 chip.  I looked at my chips, and can't tell because the heat sink covers the chip.  I don't want to break them up unless I have the fix ready.

The TLSmooth solution might be a good answer, but here again, it may depend on which A49XX chips I have installed.  I’m not wild about the additional heat issues that should be dealt with.  The chips do all have heat sinks already with the stock UMO fan set up.

I’m undecided.  Can anyone vouch for the fact that UMO shipped as standard with the A4988 stepper drivers?  do the SSSDrivers exhibit the ‘end of step’ unevenness?  Would the stock heatsinks and fan setup be sufficient with TLSmooth to dissipate the additional heat?

Sorry for the novel, and thanks for your attention and help.

Matt

Edited November 1, 2017 by Guest

[tommyph1208 55]

I run SSSDrivers on my UMO, but am not aware of the issue (maybe that means I don't have it? Or as you also mentioned it, maybe I just accepted flaws like that, writing it off as the bed shifting slightly, or similar...

Regardless I think the SSSDriver swap is a no-brainer for any UMO owner, as it just overall makes for a much nicer (primarily quieter) printer.

You mention it being a stretch for your capabilities, but speaking from experience it is really not that difficult... There was a lot of confusion at first as to how to properly install them, but in the end they are extremely close to being a drop in replacement for the stock drivers (which I do believe are A4988 ).

The current adjustments you are talking about are done on the drivers themselves with a little screwdriver... There are vias for measuring ref voltage with a multimeter on the driver as well, but you can also just go by the trial and error approach and adjust until they are running right.

As for reversing axis you can get around the firmware update issue (if that frightens you) by simply turning the stepper motor cables 180 degrees.

If you prefer one of your other solutions that depend on A4988 drivers, but are unsure the ones in your machine are A4988 or A4983, why not just buy a new set of A4988 and drop them in? They cost next to nothing... You will have to adjust current on those as well.

Edited November 1, 2017 by Guest

[conny_g 251]

I am convinced if you have the white UM Mainboard you should have th A4988 drivers.

If you don’t I can’t tell.

[mastory 44]

Mine is an UMO from early 2012, with a green motherboard ver 1.5.7

[conny_g 251]

This?

http://reprap.org/wiki/Ultimaker%27s_v1.5.7_PCB

[conny_g 251]

It seems Ultimaker used A4988, schematic here

https://github.com/Ultimaker/UltimakerOriginal/blob/master/1098_Stepper_Driver/A4988.pdf

[mastory 44]

Thanks for digging Conny.  I'll take that as a confirmation for A4988 chips. I looked pretty hard and could not find it.  I'm still interested to know if the SSSdrivers have the end of step issue or not.  At this point, I have no evidence to suggest yea or nea.  Lack of it being mentioned may prevail.

Quieter is always good. Not so significant for me - I print in the garage.

Matt

Edited November 1, 2017 by Guest

[gr5 2,167]

Quieter is always better. I suspect they don't need the tl-smoother but it would be great if you could do a before and after print with the SSS chips and let us know.

If you live in USA and do decide to get tl-smoothers consider shopping from my store at thegr5store.com.

[mastory 44]

If you have what I'm looking for gr5, I will gladly buy from your store. You do a lot for this community, and deserve more payback. I've looked before, but couldn't find a reference to the store. You should plug it more often, maybe in your profile if its possible.

I do appreciate your help here as always.

For my issue currently, I'm gong to read a bit more about the SSS driver chips and how they handle current decay - and/or - go ahead with the ROSC to ground jumper.

[conny_g 251]

How about doing both to learn - ground the rosc today and order the other drivers in parallel ;-)

[mastory 44]

I completed the ROSC ground jumper.  Actually the surface mount resistors unsoldered accidentally, so I bridged the resistor location with solder (or at least I think I did - it's hard to see)  The resistors are lost, and can't go back on.  I'm fairly good at soldering, but this is very small work and difficult for 47yo eyes.  I did this to the X and Y chips only, and then reprinted Torgeir's test piece.  The longer straight side of the piece (side A) was basically parallel to the Y axis.  This side has greatly improved in regard to the vertical lines discussed in the Zebra Stripe Thread.  The most curved side (side B) however was basically diagonal across the bed, and has what I think are pressure ripples (can't think of the common term for these) that emanate from the Z seam.  The first print had this, but to a much lesser extent.  Of course, I used the same gcode, same filament, same temps etc on both pieces.  This is natural colored ABS, layer height .15mm.

Pre jumper part is on left labeled "1".  Post jumper is on right "2"

Side A was printed parallel to Y axis:  The lines marked in the first piece are definitely gone on the second print, but the ripples are new.

Side B was diagonal toward rear left:  Major ripples

Side C was parallel to X across front of printer:  Way better on the first pass.

According to info found in the Zebra Stripe Thread , R4 is to be jumped to ground for the desired effect:

Not sure what to think.  I think I'll try again tomorrow.  If I got the chips mixed up, would it affect the current settings?  Maybe one of my solder bridges is not really connected?

Time for bed and the printer is out in the garage. I wonder if I bumped the X motor loose when I tipped the printer over for access?

Edited November 2, 2017 by Guest

[conny_g 251]

Yes, double check everything tomorrow.

For that kind of work I can recommend this kind of magnifying glasses:

http://www.eschenbach-sehhilfen.com/de-DE/278/produktuebersicht/detail/82/laboCOMFORT

with 2x or 2.5x.

No issues with small SMD anymore.

Edited November 2, 2017 by Guest

[mastory 44]

After sleeping on it, I will first check my belt tensions. If loose, will correct and reprint. If not, then print another test piece with the x & y chips swapped to see if the problem changes...

Is there a reputable source for the SSSDrivers pre assembled for UM's?

[drayson 75]

I bought the SSS for my 2012 edition UMO - the TMC2100 - directly at Watterott.

http://www.watterott.com/index.php?page=product&info=4107

It was a straight forward change with only one hurdle at the second extruder.

Just make sure you buy the right heat sinks :-)#

I can fully recommend the SSS as my UMO got that quiet and also the zebras have been nearly gone (but I never discovered them as THAT annoying...).

If it also has effect on the end-of-step-uneveness, I have no idea, but keep us updated...

Edited November 2, 2017 by Guest

[mastory 44]

I guess I am guilty of Re classifying the Zebra stripe to something more name precise based on the findings in the Zebra Stripe thread, so same thing as I see it, but more specific name? Sorry to make more confusion.

The symptoms (lines) in my first print mimic the 'zebra stripes'. I think they were corrected in one axe of my second print with the ROSC ground jumper added, but something else or questionable continuity of my jumper in the other axe caused the bigger waviness. Watterott is across the pond for me, but that could be ok

[gr5 2,167]

@mastory please update your profile to indicate what country you live in. Even if you tell me here I'll forget again later.

Definitely get some cheap reading glasses. I got my first pair at CVS. Get the strongest they offer. Seriously. The strongest. You want around "10X" (meaningless number) which is about 5 diopters. This allows you to focus much closer so you can look at a part from just a few inches/cm away and see great detail.

Also I'm not sure what you are printing but this is the best zebra-stripe-test https://www.youmagine.com/designs/curve-test-print

[mastory 44]

Gr5, that's the same part that I'm printing