A few OPS questions

I've had some not-quite-good experiences with Repetier-fw's OPS and decided to go back to Slic3r's retract while i figure out what i'm doing wrong...

Question #1: With OPS on, fast mode, set to start at 0% move, the retract starts about the same time the next move does (as expected), but the re-prime occurs after the move is finished, with the head stationary, before the next move in queue begins. Is that normal?

Question #2: The OPS retract/re-prime feed rate / acceleration are rather slow. I don't mind slow retract (since its speed is proportional to the feed rate and distance of the move), but slow re-prime is a pain. At slow speeds it results in the (hot) nozzle/block hovering stationary above the object while the re-prime finishes, and re-melting the previous layer. Is there a way to make it faster?

Apologies for being annoying with all the "issues" and questions, i've finally decided to switch from Marlin to Repetier and just can't seem to get it tuned to the same level i've had Marlin. Thanks for all the hard work, love your software!

I think you have set the max extruder speed wrong. It is a bit misleading that the max value is in mm/min while start feedrate is in mm/second. The re-prime uses max speed for computation, which would also explain that you don't see much re-prime in your case.

If you are coming from Marlin you might like the new developer version. There pid values are like in Marlin and also many marlin g-codes now work in that firmware. But be aware that the config has changed a bit - now all speeds are set in mm/s.

---

Repetier-Software - the home of Repetier-Host (Windows, Linux and Mac OS X) and Repetier-Firmware.
Repetier-Server - the solution to control your printer from everywhere.
Visit us on Facebook and Twitter!

Forgot to mention that i'm already on 0.8x.
I'll recheck the speeds, thanks for the tips.

You were right, i had a leftover value from 0.7x (due to EEPROM transition, probably) in mm/min.
Working fine now.

But i have another issue now.
The very last layer always ends up offset by 1mm along Y for some reason (in + direction).
It's always 1mm and always the very last layer.

I doubt it's skipped steps due to the nature of it (it's regular and it happens on the last layer, but not the one before the last, which in my test scenarios with a 20mm "cube" are identical).

Any ideas on what might be going wrong?

Good to see that at least one error is gone.

Having a skip always at the same position makes me think it is something in the code. Wrong code or wrong interpreted I can't say without knowing the code.
Does the host show the offset as well? Are there some unusual commands in the area of the shift. If it is always at the layer change, have a look there what commands you find and compare with other layer switches. Commands should be similar except the one causing your problem (well i hope).

---

Repetier-Software - the home of Repetier-Host (Windows, Linux and Mac OS X) and Repetier-Firmware.
Repetier-Server - the solution to control your printer from everywhere.
Visit us on Facebook and Twitter!

Took a look at the gcode for the last two layers and found nothing out of the ordinary.
The preview in host looks fine.

(BTW, highlighting the moves in 3D while highlighting the code is a BRILLIANT feature! )

Giving Cura-generated code a go now. If it happens again, at least i know i can exclude the slicer.

Okay, so the verdict is - no issues with Cura.
But Slic3r consistently gives me the same result - Y+1mm offset on the final layer.

I've attached the Slic3r (0.7.2) generated gcode, in case you want to take a look.

The part i'm finding weird is that it started happening after i've killed the Slic3r retract and switched back to OPS.
I've reduced some acceleration and max feed settings too. But no changes have been made to Slic3r settings since it started behaving this way.

Could it be some weird interaction with OPS?
Current OPS settings:

Quote

< 20:29:08: EPR:0 103 2 OPS operation mode [0=Off,1=Classic,2=Fast]
< 20:29:08: EPR:3 99 30.00 OPS move after x% retract [%]
< 20:29:08: EPR:3 43 6.00 OPS min. distance for fil. retraction [mm]
< 20:29:08: EPR:3 87 0.50 OPS retraction length [mm]
< 20:29:08: EPR:3 91 0.00 OPS retraction backlash [mm]

< 20:29:08: EPR:3 204 10.00 Extr.1 max. feedrate [mm/s]
< 20:29:08: EPR:3 208 5.00 Extr.1 start feedrate [mm/s]
< 20:29:08: EPR:3 212 10.00 Extr.1 acceleration [mm/s^2]

Edited 1 time(s). Last edit at 12/08/2012 02:38PM by orcinus.

I printed it with the same OPS setting with success. There is also nothing special at the last layer, except that it starts with a move in y direction, which is a move of 1.2 mm. When your motor would stall there you have your shift for the last layer. In this case I would try reducing the z-jerk, because it influences the start speed which seems too high in this case.

---

Repetier-Software - the home of Repetier-Host (Windows, Linux and Mac OS X) and Repetier-Firmware.
Repetier-Server - the solution to control your printer from everywhere.
Visit us on Facebook and Twitter!

It's quite low already. I've brought the acceleration and jerk down significantly (1000 m/s^2, 10 mm/s) in order to decrease shaking and vibration artifacts in object surfaces.
But good point, i've probably botched something in the same way i have with the extruder feedrate.

At a guess, i probably left a default value in EEPROM during the switch to 0.80 somewhere (travel acceleration possibly).

Thanks!

Z Jerk is a extra parameter and gets combined with normal jerk. This is normally very low like 0.2 compared to xy Jerk which is in the range of 10-40.

---

Repetier-Software - the home of Repetier-Host (Windows, Linux and Mac OS X) and Repetier-Firmware.
Repetier-Server - the solution to control your printer from everywhere.
Visit us on Facebook and Twitter!