Shift to left at the same point in the print

My first job since building my Prusa i3 has been to print replacement parts for my Prusa i3 as I'm not especially happy with the quality of the ones I purchased but I'm having issues with the X-Axis Motor mount - I'm using the Rework version. At exactly the same point in the print 1 hour 40 ish minutes the last 4 times I've tried to print it it has skipped off to the right by a couple of mm. On one attempt I let it almost complete before stopping it and it the remaining layers were perfectly in line with the new position.

I've got motors that are 68oz-in @ 1.2Amps and the motors are cool to the touch at all times. The stepper drivers are a4988's set to 1/16th steps and are putting out just under 1.2A I've got heatsinks on the chips and a 40mm fan blowing over the whole ramps board. So I don't think it is the stepper motors or drivers overheating and if it was mechanical then why does it only happen once during the print ?

I've been using Pronterface & slic3r on a Mac but printing from an SD card. I tried reslicing it for this last attempt but still exactly the same fault and as I had "Randomise starting points on" it is a completely different set of gcode instructions this time which should in theory reduce the chance of it being resonance related. I've just started attempt 5 with some different settings in Slic3r ... I've removed the "Randomise starting points" option in the hopes that this may avoid some of the long travelling movements, and I've reduced the travelling speed from 130mm/s to 60mm/s.



Can anyone help shed some light and point me to some other things I can try ?

ok so slower travel didn't help ... I had a brainwave ... I rotated the object 90 degrees in slic3r before generating the gcode so the long travel was on the Y-Axis rather than the X-Axis ... so using a different stepper driver and motor and different mechanical axis so in theory if it is a mechanical problem then I'd either see a shift to the left as per every other print or no shift at all since the travel in the X-Axis had been much reduced.

The result ... a Shift in the Y-Axis ... the opposite direction to the shift when it was happening on the X-Axis but the shift still occurred along the long edge of the object.


Could it be a Jerk speed value problem ?

I had this problem earlier this week. I ended up upgrading to the latest Slic3r (1.0 I think). Seemed to solve the issue.

Also, check to make sure your belts are tight.

---

greghoge.com

HUGE 3D PRINTER PARTS SALE!!!

I too had the same problem... particularly with some geometry (using small detail areas or this wall cross sections).

After chasing my tail with checking the mechanics (gears, pulleys etc.), changing slicers, and print speeds, changing stepper motor driver current etc. to no avail, I new I was on the wrong track. The offset problem was seemingly random (but was not random) and drove me nuts, until I ran across a post on Jerk.

Turns out it was the Jerk speed value was way to high, I radically reduced the Jerk Value from 20 to 10 in my Marlin Firmware (located in Configuration.h)
// #define DEFAULT_XY JERK 20.0 // (mm/sec)
#define DEFAULT_XY JERK 10.0 // (mm/sec)
NOTE : at 15.0 I was still having the offset problem.

Reducing Jerk to 10 has turned my printer into a much more friendly sounding printer, less jerky of course/smoother and NO More offsets in my print.
My printer now runs smooth as silk and doesn't sound like it's rattling itself apart while printing.

Hope this fixes your offset problems also.

__________

FYI... NOT sure why anybody is still using Slic3R beyond v7.02b (v7.02b is a great simple and competent slicer).

I laugh when I see people trying to fix problems with this continual problematic slicer >Slic3R v7.02b.
I finally gave up after Slic3R v9.8; recouped my life, and no more wasted time and plastic, and moved forward away from Slic3R.
Every new Slic3R upgrade version has a new set of unresolved problems. I tried EVERY Slic3R version v7.02b - v9.8 extensively and utilized every smart so called fix for the problems... result garbage can full of melted plastic, hundred of frustrating hours endured/wasted, and saw my life passing me by. I'M DONE, DONE, DONE and DONE with Slic3R !!!

I used Skeinforge v41-v50 as my primary slicer for the last 2 years, but it has a memory limit which restricts you from printing more than simple trinkets and objects.

Over the last year, I print very complex and close tolerance engineering prototypes (0.1mm Layer Height @ +- 0.05mm tolerance) all day long everyday using KISSlicer, IMO it is the best slicer solution out there. NO more Slic3r wasting my time and memory limitations using Skeinforge, I have my life back now and time to do other things other than melt plastic in to useless shapes and fill the garbage can.

Note : KISSlicer is seemingly and sadly going defunct it appears, and is no longer maintained by it's creator/owner, but is still the best overall slicer ATM.
Download from here:
www.kisslicer.com

Information here:
- Kisslicer Refugee Camp - Google Groups
[groups.google.com]

_____________________

Your problems should be all be history now.

Edited 1 time(s). Last edit at 01/25/2014 01:57PM by ShawnT98027.

I'll try with a lower jerk setting as mine is set at the default of 20.

I'm using the latest version of Slic3r and Marlin downloaded from Github on the 14th of Decemnber.

I have made some progress though ... I sliced the part using Cura with as close to the same settings I've been using in Slic3r as I could and while I don't want to temp fate the print has so far progressed far beyond where I was getting.

I think the quality of the print from Slic3r is better than Cura on the whole ... the edge is straighter and the layers are more even and finer in. But I like the way Cura has been building the support structures better ... Slic3r was putting far too much in as far as I was concerned when I'd set the angle to auto ... I guess I'lll have to experiment to get it better and I wonder if by tweaking that I'd get round this problem too.

OK so back to square one ... got one good print out of it and since then every other part I've tried has resulted in the same basic behaviour ... it gets to a specific part of the print and then you can hear the stutter, actually it is more of a clunk kind of sound ... I've tried slowing everything down to 30 or 40mm/s and reducing the default acceleration down to 100 in config.h and still the same ... getting this behaviour with objects sliced in Cura too ... really ******* frustrated at the moment ... got a £400 paperweight ... just wanna chuck the whole thing in the bin.

Edited 1 time(s). Last edit at 01/29/2014 06:24PM by BarneyG.

Is it the same EXACT layer every time? It might be something in your axis or something getting in the way at that specific point.

---

greghoge.com

HUGE 3D PRINTER PARTS SALE!!!

Quote
gmh39
Is it the same EXACT layer every time? It might be something in your axis or something getting in the way at that specific point.

Or it might be a specific move in the print at that point (like a very short infill) that makes the motor almost vibrate and oscillate at a frequency that causes it to skip a step. I had this before with my older printer. It was solved upping the stepper driver power and what solved it for me was to not print with honeycomb infill but just using the rectilinear infill.

I fought this issue for quite a will too...

If lowering the Jerk Value to 10 did not solely eliminate your offset problem, and upping the stepper motor driver voltage still does not fix the problem of offsetting (did nothing for my offsets), the next obvious step to me was find a better slicer (KISSlicer). KISSlicer seems to control the infill acceleration properly in thin walled geometry (<1.75mm wall) better than other slicers, hence reducing jerk that can create offsets.

One change at a time.. to monitor effect of change...
Step #1 - I too initially bumped up my stepper motor drive current (still offsetting prints everytime with a particular STL)
+
Step #2 - Reduced Jerk to 10 (still infrequently offsetting prints with the same above particular STL)
+
Step #3 - Started using KISSlicer.... (unable to produce offsetting prints with the same above particular STL), and never saw an offset since in 100's of future prints.
End of Story.

Don't get stuck on Slic3R... (and chasing your tail around forever)
I laugh when I see people trying to fix problems with this continual problematic slicer >Slic3R v7.02b. I finally gave up after Slic3R v9.8.
Note: Skienforge (and Skeinforge based CURA), and of course Slic3R have never produced stable results when printing complex geometry with thin walls.
IMHO these slicers are for printing small simple trinkets and toys only.

After the dust had settled (with stable non-offset prints every time for a few weeks), I then reduced by stepper motor back to it's originally pot setting/voltage setting and still producing perfect prints everyday a year later. So for me the problem was too high of Jerk value and a POS Slicer.

And yes, I agree with Ohmarinus, with using the rectilinear infill.

So for my setup, reducing Jerk to 10 and using a non-junk Slicer (KISSlicer), ended this offset problem.

Best of Luck !

Edited 2 time(s). Last edit at 01/30/2014 02:15PM by ShawnT98027.

Same here, I went to Cura, now I'm a happy camper

Cura was giving me similar problems in different places to Slic3r.

I've tried KissSlicer now and had 2 successful prints out of 2 attempts. I'm not quite as happy with the finish of what it produces at present though ... much more stringing between parts and on one print it kinda looks like the infill goes outside the perimeter if you know what I mean.

And thanks to documentation on how to use KissSlicer I now understand how to adjust the zero point for for my z axis so I don't have to get it 100% correct via manual adjustment of either the endstop or the bed

I had this same problem, but it was not the fault of Slic3r, Cura, etc. It was not mechanical in nature either. What was causing this for me was my acceleration settings in the Marlin firmware and in Slic3r. I have a home built Prusa i3, and I had been using an acceleration speed on the Y axis of about 1500, both in the firmware and in the Slic3r software. I reduced that to 1000 and also lowered the "jerk" speed on the Y axis down to 5. This took care of this problem for me.