Need someone with RAMPS and Marlin..

.. for confirming a bug in Marlin. It's a very simple test:

1. go to starting position using G1 X84 Y-45 Z30 F1000 (Z and F are arbitrary)
2. reset Extruder position using G92 E0
3. move slowly to second position while extruding 17 mm using G1 X-54 Y51 E17 F180 (feed unit mm/min)

The effector should start moving with 180 mm/min and soon accelerate like in this [1] video. We are affected with the bug since the end of last year and would appreciate any help. Further information can be found here [2] and here [3].

[1] [youtu.be]
[2] [github.com]
[3] [forums.reprap.org]

Edit01: G29 -> G92

Edited 1 time(s). Last edit at 06/02/2016 02:16AM by ribaldos.

Marlin firmware uses the Bresenham approximation to compute motion and also uses an approximate acceleration algorithm. As a result, it has minimum movement speeds. I suspect that 3mm/sec is below the minimum movement speed of Marlin for some towers at some XY positions. Does the problem go away if you increase the speed to 10mm/sec?

RepRapFirmware used to use the same algorithms and also had a minimum speed. I rewrote it in late 2013 to calculate the step times precisely, which allowed me to get rid of the minimum speed.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Hi dc42, yes, the behavior is still present when using higher speeds, however it is best visible with lower speeds. The weird thing is that it does not occur when the 4th axis (E) is not involved. In that case even lower speeds are possible.

Does it vary with the number of segments per second that you configure?

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

G29 is for auto-levelling (Z probing).

Did you mean G92 E0?

Quote
dc42
Does it vary with the number of segments per second that you configure?

The behaviour expresses differently when going as low as 20 - 30 segments per second but it still persists (see upper [3] for further info)

Quote
frankvdh
G29 is for auto-levelling (Z probing). Did you mean G92 E0?

Thank you for the correction!

The topic is still open. Can someone with Marlin and RAMPS please test the three commands? It should not take more than 3 minutes.

Are you sure that the feed rate is in mm/min? Usually it's in mm/s.

Marlin has a problem in that when the effector is near the edges, the speed is a lot slower than when the effector is near the centre. I've certainly seen this behaviour before when using Marlin.

Moving to Repetier alleviates this problem.

It is indeed mm/min in the gcode.

The behavior occurs also in the middle and at higher speeds. Certain ratios of extruder- and motor-speed are causing it. Sometimes, very low speeds (e.g. for the first layer) even work better than faster speeds.

I guess that stepper pulses coincide at certain feedrate ratios and cause mess within the firmware (accumulation of interrupts?).

Had no luck with Repetier, either. No more strange behavior, however, printing at low speeds is no more possible (Bresenham approximation) and we had loss of steps (thread).