Deceleration problem, not acceleration CoreXY
March 23, 2015 01:04PM
Hello.

I have an issue.

I've tweaked and tweaked acceleration and jerk values and motor current again and again but not resolved it.

If I issue a move to X or Y at more than 7500 mm/s It starts off ok, accelerates at any rate ok, but at the end of travel it 'clonks' and skips on another 2 or 3 mm

This is on a CoreXY set up.
If I configure it not as CoreXY - ie each motor moves my 'thing' diagonally, then I can set whatever feedrate I wish with no problem. Each motor sounds fine.

It 'feels' like some sort of timing issue between the 2 motors when they think its time to stop.

(Having 'feelings' about 'thinking' motors is not scientific, I know)

I've swopped motor drivers (8825 on RAMPS) around with no effect.

Any thoughts on that?

Thanks

Alan
Re: Deceleration problem, not acceleration CoreXY
March 23, 2015 02:33PM
Quote

If I issue a move to X or Y at more than 7500 mm/s
Is 7500mm/s a typo?
Re: Deceleration problem, not acceleration CoreXY
March 23, 2015 04:28PM
Ahum!
Scuse me.

That should be mm/minute.
Re: Deceleration problem, not acceleration CoreXY
March 23, 2015 04:36PM
In my experience, clonking sounds from the motors are typically caused by having the maximum jerk rate or maximum acceleration of the motors configured too high, or by firmware bugs (i.e. firmware not respecting those values).



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].
Re: Deceleration problem, not acceleration CoreXY
March 23, 2015 05:35PM
At 7500mm/m it is ok with Acels up to 3000 and Jerk up to 20

At 7501mm/m it gives the clonk even with Acel at 200 (wow, so slow) and Jerk at 1

Is there anything significant about 125mm/s 7500mm/m?

I'm starting to wonder if its firmware which is latest Marlin of 2 days ago.

Ta.


-A
Re: Deceleration problem, not acceleration CoreXY
March 23, 2015 06:01PM
Quote
alan richard
At 7500mm/m it is ok with Acels up to 3000 and Jerk up to 20

At 7501mm/m it gives the clonk even with Acel at 200 (wow, so slow) and Jerk at 1

Is there anything significant about 125mm/s 7500mm/m?

I'm starting to wonder if its firmware which is latest Marlin of 2 days ago.

Sounds like a firmware bug to me. I have read that Marlin has to make some compromises to generate high step rates on the 8-bit Arduino/RAMPS platform, and perhaps that interacts with the CoreXY support. I don't use 8-bit controller electronics or Marlin, so I can't help. Assuming you don't want to switch to more capable electronics, I suggest you either try to get the bug fixed in Marlin, or try an alternative 8-bit firmware such as Teacup (if that supports CoreXY). Or you could just live with the speed limitation - I doubt you will get good print quality at more than 125mm/sec anyway.



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].
Re: Deceleration problem, not acceleration CoreXY
March 23, 2015 06:36PM
Thanks dc42.

I did wonder if the wee Arduino may be struggling with the number crunching and quite agree about the print speed.

Just checking for something I may have missed.

Regards

a
Re: Deceleration problem, not acceleration CoreXY
March 24, 2015 11:36PM
I experienced a similar problem on my CoreXY, but more random. If I gave a single G1 command, even up to 650 mm/s, there was no problem. But when I was running a file, there were random places where the machine would give a "clunk" sound and lose steps. I finally gave up trying to achieve anything higher than 250mm/s. At 250mm/s, even at 10,000mm/s^2 and jerk of 50mm/s, I have no problems. Right now I have infill set for 125mm/s and rapid travel at 250mm/s and it is very reliable.

I have the same 2560/RAMPS/8825 combination, but I have my drivers set to 1/16 steps instead of 1/32. So assuming you have a 20 tooth sprocket, the 125mm/s you get would match the 250mm/s I'm using without a problem. So there does seem to be a limit somewhere around 20,000 steps/sec. However, I saw that you posted elsewhere that changing to 1/8 steps didn't help. It should have allowed you to go 4 times as fast, if it really was a firmware processing limit.
Sorry, only registered users may post in this forum.

Click here to login