Custom motion controller board, making weird PWM noises?

Hi guys,

I've created a custom motion controller for my new RepRap based pick and place / 3D printer:
[hackaday.io]
[github.com]
COMPLETELY open source, btw.... let's get that out of the way.

I'm finally getting it running, and it seems like when the motors are enabled, I'm hearing this weird repeating pattern of pwm noises. I'm watching the stepper signals on a scope, they'll do the PWM and then just go to the +12V rail for a sec and then back to PWMing. Here's a cellphone video that demonstrates the problem:

[www.youtube.com]

And here's some references for my design. It was inspired from RAMPS and Melzi, and uses the StepStick controllers, and a 20 MHz Atmega1284P. Emphasis was on completely SMT design (well almost), with a modular tool system that doesn't really matter for the context of this thread :-)

Github Repo: [github.com]
Schematic: [github.com]
It runs modified Marlin: [github.com]
And needed a custom Arduino hardware definition because sanguinololu didn't support ATmega1284p at 20 MHz: [github.com]

Anyway, I was wondering if any of the old-timers around here have ever heard anything like this, and if there's anything that I should check. I'll keep looking at it, but figured this was a great resource to bounce ideas off of.

Thanks!

-Neil

Some other info that might help....

• The /ENABLE line is solid-low when commanded to be enabled, and solid-high when not enabled.
• MS1, MS2, and MS3 are all permanently tied high, as my delta mechanism really should only be run with 16x microstepping.
• /SLEEP is tied to /RESET per the Pololu page, and the RAMPS 1.4 schematic.
• Motor power is +12V from a pretty standard switching power supply
• +5V is currently from a second +5V switching supply (was going to be powered from USB, but I didn't realize how much current my opto endstops would pull).. next version will have a +5V reg onboard.
• Stepper motors are pretty standard RobotDigg high-torque NEMA 17's that have worked great in the past on a RAMPS board: [www.robotdigg.com]
• Voltage at the StepStick potentiometer(s) is ~1.1V
• There's not really any torque or anythign on the motors while they're making the periodic weird noises.
• The stepper motors and StepSticks have all been tested "good", I've used them for quite a while before putting them in the new board.

And probably the most wierd thing: The machine will actually move around as expected, the noises are only really after it's idled for a while. The pitch seems to go down a bit after the rapids, and then it starts making the noises.

Edited 2 time(s). Last edit at 09/27/2014 10:02AM by FlyingLotus1983.

I think this was just a matter of using the wrong calculation on the potentiometer current limiting... I think the sticks were going into over-current protection, and then coming back out of it. I added heatsinks and dialed them down, and it's running great now.

Quote
FlyingLotus1983
I've created a custom motion controller for my new RepRap based [...]

What's the point of reinventing the wheel?

Quote
FlyingLotus1983
I'm finally getting it running, and it seems like when the motors are enabled, I'm hearing this weird repeating pattern of pwm noises. I'm watching the stepper signals on a scope, they'll do the PWM and then just go to the +12V rail for a sec and then back to PWMing. Here's a cellphone video that demonstrates the problem:

[www.youtube.com]

Sounds much like stepper driver chips overheating. As you probably know, their load is highest with the stepper not turning. Reducing motor current on the trimpot should help.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY