E-step difficulty

[JimG]

Calibrating my E-steps is not going according to plan. Based on my results, I feel there must be some slippage going on in the extruder. But I cannot detect the slippage by watching/listening.

After applying the linear corrections predicted by measuring a 100mm gauge length, the new E-step values were bouncing all over the place. So I abandoned the linear correction formula [new = old / (100 /measured extrusion) ] and just tried some various values. My results:

E-steps = 900 , caliper measurement = 27.4, extrusion length = 92.6 mm
E-steps = 910, caliper measurement = 25.8, extrusion length = 94.2 mm
E-steps = 960, caliper measurement = 13.6, extrusion length = 106.4 mm
E-steps = 935, caliper measurement = 31.8, extrusion length = 88.2 mm

The trial with 935 extruded significantly less than the trial with 900, so something is not right. And subsequent trials with E-steps set to 935 yielded caliper measurements of 31.8, 25.7 and 14.5.

I have tried cranking down very hard on the guidler screws, and I have tried backing them off. The bearing in guidler turns freely. What else can I try?

Jim

10" Prusa i3V, 3mm natural PLA, 185F, 0.4mm Hexagon

[JimG]

Apologies to all. I now see that this topic is well-covered in a recent thread and seems to be related to 100mm/min being a little too fast for the 0.4mm Hexagon.

Jim

[JimG]

When asking for 100mm of extrusion through pronterface I still see variations of more than 25%. I can't get my estep calibration done. And yesterday a long print was ruined on the last few layers because (apparently) the extrusion rate dropped very, very low to finish the print.

I cannot see slippage of the large gear on the hobbed bolt. I cannot see slippage of the small gear on the shaft and I have tightened the small set screw as much as I dare.

I don't see gouging of the filament by the hobbed bolt, and I am seeing the variability across several different spools of 3mm filament, both PLA and ABS.

The pot on the stepper controller is adjusted (I did not record the final voltage readings when I checked and did very minor adjustments, but they were all very close to the middle of the recommended range). There is nothing obvious to me that says the stepper is missing steps.

The only clue, besides the wildly varying extrusion lengths, I have to go on is a lot of squeaking when I extrude PLA. I tried a little vegetable oil and this reduced the squeaking but did not fix the extrusion problem.

Unfortunately, all of things on my list to check next involve disassembling the extruder. These are the suspects I have identified:

1) The filament path is constricted passing through the extruder body? This explains the squeaking. I did not drill out this path before assembly.

2) The nut on the back side of the hobbed bolt is not tight enough allowing the large gear to slip?

3) The small gear is slipping on the shaft?

Is there anything else I can try before disassembling the extruder?

Thanks in advance.

Jim

[printbus]

FWIW, I noticed the consistency of my extrusion tests went up after I started discarding the first lengthy extrusion after heating up the extruder. I figure the filament might be *cooking* a bit in the nozzle during cooldown from prior use, and that the subsequent extrusion helps clear out whatever gunk might be in the nozzle.

One thing you could check before disassembly is to remove the filament from the extruder and see how easy the large gear rotates with steppers disabled. On one extruder build, I evidently hadn't tightened the second nut on the large gear enough, and with use the large gear was slowly rotating itself tighter to the extruder. By the time I noticed that, I could barely rotate the large gear by hand when I tested it.

[printbus]

Another easy thing to check before disassembly - looking down at the top of the extruder, how well does the filament groove in the guidler bracket line up with the cut in the hobbed bolt? I know that on my printer, it took an unusual combination of a washer and R/C axle shims under the head of the hobbed bolt to get them lined up, and usarmyaircav has had a significant alignment problem with his. If the alignment isn't good, perhaps the filament moves around, sometimes riding on the deepest part of the hobbed bolt cut and other times riding on the shoulder of the cut. It seems like that could cause some extrusion variability.

Is the hobbed bolt free of any filament build up, as best you can tell?

[Roxy]

If the alignment isn't good, perhaps the filament moves around, sometimes riding on the deepest part of the hobbed bolt cut and other times riding on the shoulder of the cut. It seems like that could cause some extrusion variability.

This can also cause the filament to twist as it comes off the roll. Or actually, it twists going into the extruder, but its connected to the filament coming off the roll and it can look like the filament is coming off a roll that was wound with twists in it.

[JimG]

One thing you could check before disassembly is to remove the filament from the extruder and see how easy the large gear rotates with steppers disabled. On one extruder build, I evidently hadn't tightened the second nut on the large gear enough, and with use the large gear was slowly rotating itself tighter to the extruder. By the time I noticed that, I could barely rotate the large gear by hand when I tested it.

When I checked the large gear in the manner you suggested, I found that it can move a few degrees either way without the hobbed bolt rotating. Pretty clear that I did not tighten the jam nut tightly enough in the back. I can get rid of the play by holding the head of the hobbed bolt and turning the gear. I am going to run that way for a bit and see if it will stay snug. But I am making plans to disassemble

Thanks for the insight.

Jim

EDIT - nope, running temporarily not going to work. When reversed, the gear loosens again immediately. This also explains why the problem just showed up in a print rather than only in calibration. The particular object I was printing needed a bunch of retractions. So I am guessing that the bolt became very loose during the print, resulting in an almost complete loss of extrusion.

[JimG]

Main source of the problem is that the nut well on the large gear is 1 - 1.5 mm deeper than the nut thickness. So the jam nut cannot make contact with the nut in the well. I noticed this during assembly but failed to act on it then.

The two M8 nuts in my kit were both thin nuts. The well in the large gear is sized for a standard M8. Fortunately, I had a standard M8 in a spare parts box. When I swapped out the standard nut for the thin nut, there was enough projection outside the large gear for the jam nut to do its job.

Estep calibration complete now at 880, with fairly consistent extrusion lengths. I am cautiously calling this a solved problem.

Thanks again for the help!

Jim

[old man emu]

What numbers are you getting for steps per 100mm? I hadn't calibrated my extruder and it had the factory setting of 831 steps per 100mm. I ran a calibration test and found that instead of 100mm it only pulled through 87mm.

According to the gurus, Calibrated E steps(100) = (100/measured feed) x original E step(100) value

For me this is Calibrated E step (100) = (100/87) x 831 = 1.15 x 831 = 955.

OME

[JimG]

I ended up at 880.

My factory default was 841, but this extruded a little less than 100mm in my calibration test.

Jim