Debugging DrRob controller Board

[neilhand]

Hi All,

I have put together a variant of DrRob's controller board using PerfBoard rather than StripBoard, and initial connectivity checking seems to indicating that the wiring is correct - however all is not well.

When i power up the board it draws current on the 24V side (just under .5A from a bench supply) for the first few seconds during which the X axis stepper is held in position - then it seems to shut down. If i then use the gCode controller (under windows) to try to adjust the x-axis the stepper will make some noise, but no actual movement - just rocks in place. The bench supply seems to indicate that it is trying to briefly draw current. Is it possible that since the bench supply is doing something odd in the setup? When it transitions from no current draw to attempting to move the stepper it makes an audible click (similar to when the current limit kicks in, but it is not going into current limit mode).

Also adjusting any other axis has no impact on the associated stepper, but it does make a (lower volume) noise on the x-axis stepper.

I am using the DrRob GRBL hex file from the wiki (0.7 version) and the newest implementation of the controller board with the pull up resister. Using a scope on the enable signal seems to indicate that the polarity is correct (pulled low when attempting movements on any axis), and the step and direction commands seems to be correctly activating. The steppers i am using are the Sparkfun 400 step/rev version in the original parts list.

Any suggestions before i start more detailed debug? I am hoping its an obvious brain-dead mistake that someone has already come across.

Thanks in advance for any suggestions.

[minorthreat]

That really sounds like a current limit.

What are the spec's on the supply?

[neilhand]

I'll have to check this evening, but it should have no trouble delivering several amps at this voltage. I'll also check if the current limiting LED is lighting up - i havent noticed it, but it is easy to miss, and i may have the limit set too low.

What should be the steady state current requirement for holding the steppers under no activity - any idea?

[minorthreat]

It should only be an amp or 2 for the little sparkfun steppers. Go ahead and double check your wiring and reflash your hex file for good measure. Then double check that you have the current pot's set right. Here's the process from http://www.synthetos.com/wiki/index.php ... grblShield

Motor current for each axis is adjusted with the trimpot next to that axis.
Clockwise increases current, counter-clockwise decreases current.
Do not over-torque these pots - they are tiny and gentle and will break. The pots have 270 degrees of rotation.
You want the motor current set slightly above the range you need for your application, but not much higher.
Overdriving the motors draws more current and risks overheating or thermal shutdown.
Start by setting current to zero by turning the trimpot all the way counter-clockwise.
Then issue a very long Gcode command for that axis, something like g0x1000
Turn the trimpot clockwise until the motor starts moving reliably.
Mark this lower bound pot setting.

[neilhand]

Is the procedure for the Pololu carriers the same?

Specifically should i be measuring the voltage at the reference point at steady state, or while a continuous movement is under way?

Checking the bench supply shows it can deliver a little over 3.5 Amps at 24V, and is not going into current limiting mode while driving. The clicking seems to be due to 24V being on the boarder of switching between some ranges so i dropped it down a little to avoid the switching.

Havent tested more however since i wanted to make sure i understood the calibration procedure first. I'll then test with just a single driver in place.


UPDATE: All working. turned out it was a combination of three poor connections, one was a poor solder joint on the enable (metered out OK, but once plugged into the Arduino the joint separated) - this was the big issue. The other two issues one on the wiring for the Y and one on the carrier for the Z were easily diagnosed and fixed once enable was resolved.

Thank for your suggestions and help. Now its time to install the belts and let it go...

[DrRob]

UPDATE: All working....

Great.

It'd be good to see some photos of your variant, and please add yourself to the list of Makers on http://www.shapeoko.com/wiki/index.php/ ... rier_board too.

Rob.

[neilhand]

Updated the wiki

An image of the electronics mounted on the back of the machine (need to beef up the heatsinks). So far it has successfully drawn the hello world from the Wiki, and i am in the process of setting up the software workflow before starting any cutting.

Perfboard implementation of DrRob Driver board.

IMG_0168.jpg (23.24 KiB) Viewed 296 times

[edwardrford]

Looks great Neil! Depending on what you're planning to cut, You'll probably want to cover the electronics at some point with either full, 3 sided enclosure, or maybe just a basic 'roof'. That'll prevent debris from shorting something out.

I like the location that you mounted the electronics. Did you double sided tape it or drill & use standoffs?

-Edward

[neilhand]

For the moment it is double sideed tape. A box is one of several things i plan on making once i get everything running - i like the idea of the mill making its own improvements, and saw an article about CNC joinery that i want to try out.

The Mill itself went together very well - i really like your design.