HobbyCNC microstepping problem

Z

ZeptoBit

Registered
Registered
Joined
Dec 18, 2014
Messages
20
I'm setting up my Taig mill using a HobbyCNC pro chopper driver and 23HS30-3006D steppers, 39V 11A PSU, LinuxCNC.

At first I thought it ran fine using 1/4 step, but then I checked the motion with a dial indicator and noticed that the motion was not smooth. I know the size of the microsteps will never be exactly the same, but this is much larger.

I did some testting and got these results:
Full step: move
1/2 step: move - stop
1/4 step: move - move - stop - stop
1/8 step: move - stop*7
1/16 step: move - stop*15

So regardless of the numbers of microstep set the stepper moves only once or twice per full step, resulting in correct motion only for full stepping. The result was the same for all 3 axis. For each microstep where there is no motion I can still hear the sound of the stepper changing. Current limit is set to 3A and I have disabled idle current reduction.

I don't think this can be caused by wirering the steppers incorrectly, as full stepping should not work either in that case.
I don't think it can be the motors, as I don't see how a faulty motor could produce these symptoms, and it in any case it should affect only one axis.
I don't think it's missing step pulses from the PC, as the pitch of the sound from the microstepper changes for each microstep even when there is no motion.
A faulty driver chip would be my first guess if it was only one axis, but all 3 seems very unlikely.

Anyone know what might be wrong?


Emil
 
Does varying the feed rate have any effect ?
Are all your mechanical connections ok ?
Are the gibs too tight ?
I'd be looking at the pulse length to begin with, then I'd look at the speed\acceleration. Being a Mach 3 user this is what I'd check for.

If the board has been built from a kit I'd also check the construction. The board appears to be rated at 36V max not 39V (or is that a typo).

It may also be an idea testing a stepper whilst not connected to the machine, issue movement commands that should give you a full revolution,half or quarter with the required movement commands.

My apologies if you have tried any or all of these suggestions.
 
bladehunter said:
Does varying the feed rate have any effect ?
Are all your mechanical connections ok ?
Are the gibs too tight ?
I'd be looking at the pulse length to begin with, then I'd look at the speed\acceleration. Being a Mach 3 user this is what I'd check for.

If the board has been built from a kit I'd also check the construction. The board appears to be rated at 36V max not 39V (or is that a typo).

It may also be an idea testing a stepper whilst not connected to the machine, issue movement commands that should give you a full revolution,half or quarter with the required movement commands.

My apologies if you have tried any or all of these suggestions.
Click to expand...


Thank you for your reply.

I don't think it's a mechanical problem, because then I would expect it to move erratically or stall, not move in a completely repeatable pattern like described in my first post. My descripton of the result may have been unclear/misleading: When I wrote "stop" I meant that it didn't move on that microstep, not that it stalled.

The board it rated for 42V. My power supply is nominally 36V, but adjustable +/-10%. I set it to 39V.

The LinuxCNC stepper wizard set the pulse width automatically when I selected the board I'm using. I will check if the values used are correct.

You could be right about it being a construction error. I built this board many years ago and used it on another machine, so I assumed it to be working, but this problem is subtle enough that I may have missed it.

When looking at the steps I just used a dial indicator and jogging one step/half step/micro step at a time. So effects of acceleration and feed rates have been removed from the test.

The motors does give the expected movement when connected to the mill, as long as I use full stepping or using micro steps if the motion before reversing direction is a number of microsteps which adds up to a full step.


Emil
 
I'm a complete noob to LinuxCNC but I will ask, are your steps per revolution matching your microstepping settings on the driver? If you have your steps per revolution set at 200 and your board is set to 1/16 microstepping a 1.8* motor (3200 steps per revolution) your going to have problems.
 
bladehunter said:
Sounds like you maybe describing a backlash issue.
Click to expand...

There is some backlash on the X and Y axis, ~40-50µm which I have not tried to do anything about yet.

But this is different.
For example if I jog the machine in one direction one 1/4 step at a time it will be like this:
1: Moves 1/4 step.
2: Moves 1/4 step.
3: No motion (but stepper sound changes)
4: No motion (but stepper sound changes)
Then this patter repeats consistently as I keep joging the mill.


Emil
 
MarkStephen said:
I'm a complete noob to LinuxCNC but I will ask, are your steps per revolution matching your microstepping settings on the driver? If you have your steps per revolution set at 200 and your board is set to 1/16 microstepping a 1.8* motor (3200 steps per revolution) your going to have problems.
Click to expand...

I have set the step size correctly. If this setting was wrong it would be a problem as the machine would move slower/shorter or farther/faster than expected. (And in the last case, possibly stall.)


Emil
 
Assuming the chips are OK, it is starting to sound like a wiring issue. I could not find good documentation on the board, but the key seems to be that the White and Yellow wires are connected to their respective common. I haven't worked with a uni-polar 6 wire stepper, but I know the phase relationships must be correct. Just by logical inference, miss-wiring might exhibit the symptoms you describe.
 
JimDawson said:
Assuming the chips are OK, it is starting to sound like a wiring issue. I could not find good documentation on the board, but the key seems to be that the White and Yellow wires are connected to their respective common. I haven't worked with a uni-polar 6 wire stepper, but I know the phase relationships must be correct. Just by logical inference, miss-wiring might exhibit the symptoms you describe.
Click to expand...

The two common wires are not connected together in the stepper. There's seperate terminals on the board for the two common wires but they are connected together on the board.

I have checked the wirering (a few times) and I think it is correct. I don't think the stepper would run correctly when full stepping either if the wirering was wrong?

I will check my wirering again, maybe I missed something.


Emil
 
ZeptoBit said:
There is some backlash on the X and Y axis, ~40-50µm which I have not tried to do anything about yet.

But this is different.
For example if I jog the machine in one direction one 1/4 step at a time it will be like this:
1: Moves 1/4 step.
2: Moves 1/4 step.
3: No motion (but stepper sound changes)
4: No motion (but stepper sound changes)
Then this patter repeats consistently as I keep joging the mill.


Emil
Click to expand...

Damn what class ballscrews are you using ?
 
You must log in or register to reply here.
Back
Top