uneven steps on my Kossel Mini

I noticed with the original RAMPS that very small jogs on the vertical axis would sometimes not even move and I blamed the limited motor/math involved. I assumed it would go away with the Duet.
Today I am jogging up and down and I can visibly see the head zig-zag as I jog it straight up and down. In one case I put my fingers on the stages and jogged .01 down. The X and Y towers did no move at all but the Z tower moved significantly, which is clearly why I see sideways moves.
Now the math precision should be good with the duet so what gives? Vertical moves don't even need any math because the the towers all move the same amount (unless there is some kind of compensation going on but I only powered up this morning and have not done any auto-cals yet)

I should get better precision when I upgrade to 0.9 degree motors but even 1.8s shouldn't have uneven behavior like this.
Ideas?
thanks.

edit: I actually get the above mentioned behavior with .05 steps. I tried to read the stepper count with Get Position but apparently the RepRapFirmware doesn't not return those because they all have a value of zero. If all the steppers do not change the same amount then it seems like a firmware issue.

Edited 1 time(s). Last edit at 08/21/2015 07:20PM by shadowphile.

You should take a look at this: [www.micromo.com]

What motors do you have?
Do you feed them enough current to actually move when microstepping?
Do the carriages move equally easy or do one or two require more force to move?

Please don´t refer to the micromo blog.
It is already proven, that he is wrong. No one would build a CNC machine with steppers providing only 10% at 1/16 of the full torque.

Better point towards the "decay hack" page. When I did this mod with my DRV8825 drivers, it opened a whole new world for my delta.
-Olaf

+1 for the previous reply. The Get Position gcode doesn't return stepper positions in RepRapFirmware, but I am about to do a release and I'll look at adding it.

---

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].

My thoughts are a mess after trying to read Google-translated German but here is what I think they are saying: ground ROSC pin for quiet hold and noisy moves, or connect to +5 FOR noisy holds and quiet moves. I may have this backwards from memory but the concept is what matters, yes? I will have to read the data sheet for those drivers.
The Duet so far has been much more noisy than what I had RAMPS set to. I believe I need to keep adjusting motor current and acceleration to fix the noise. Higher accelerations actually made it quieter. If electronics and theory stay the same then I should be able to adjust the duet to the same settings as I had on RAMPS and get similar noise/accuracy.
Higher currents should reduce motion hysterisis (zig-zag) at the cost of more heat and noise, disregarding the previously mentioned ROSC hack.

Edited 1 time(s). Last edit at 08/22/2015 07:08AM by shadowphile.

It was in english a couple of posts down.

I agree with the duet making the motors a lot noisier.

Edited 1 time(s). Last edit at 08/22/2015 07:15AM by Koenig.

Also noticed slightly more stepper noise with Duet after commissioning back RAMPS 1.4 (in this thread)

I only did the hack three times, but IMHO the easiest way to solder a thin wire to pin 19 is to lift up the leg from the PCB carefully.
Use a dry ( no solder blob ) tip and a cutter knife to heat up the pad and lift the leg.
Then it should be easy to solder the wire to the leg without shorting out the neighbor pins. Secure the leg with a blob of glue after the functionality is verified.

I´ve checked my A4988 modules too and found two green PCBs which already had a 0R resistor placed on R4, so the "hack" has become standard at some place.
Don´t know about the DRV modules, but someone stated he got a version with a bridge to solder the decay mode that suits best.
For deltas the noise at idle isn't relevant, all steppers are busy most of the time.
-Olaf

Quote
roboprint
Also noticed slightly more stepper noise with Duet after commissioning back RAMPS 1.4 (in this thread)

If you had exactly the same motor current set on the Duet, then it could be because the drivers on the Duet have the ROSC pin connected to ground. The reasons for doing this are to ensure that microstepping works at low currents, as explained at [hydraraptor.blogspot.co.uk] and on the A4982 datasheet.

---

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].

Quote
shadowphile
...
edit: I actually get the above mentioned behavior with .05 steps. I tried to read the stepper count with Get Position but apparently the RepRapFirmware doesn't not return those because they all have a value of zero. If all the steppers do not change the same amount then it seems like a firmware issue.

I just put a new release candidate binary (1.09g-rc1-dc42) of RepRapFirmware at [dl.dropboxusercontent.com]. This includes the following changes:

1. M114 how returns the stepper motor positions as well as the head coordinates. So you can tell whether the firmware has instructed the motors to move or not.

2. If you enable debugging but disconnect the USB port so that there is no consumer for the debug data, the firmware detects this, thereby avoiding a software watchdog timeout caused by blocking in the output routine. You indicated in another thread that this had happened to you.

---

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].

Cool! The age of software makes every new release like Christmas!