[Dougal]

I've got two steppers which are running rough. Here's a vid.

http://users.actrix.co.nz/dougal.ell...er%20catch.wmv

This is the worst, the second isn't so bad and the third one runs perfectly smooth. They are all 280 oz-in 8 wire dual shaft nema 23 motors. I've pulled the end caps off this one a couple of times and can find no sign of internal conflict.
The motors are new, bearings are good and they spin smoothly by hand. But when running they're rough. This one catches every 2-3 revolutions and in the same spot forwards, a different spot backwards.

Has anyone had a similar issue?

[15mgtar]

I have exactly the same issue with my smaller motor and I don't even have a clue. Anybody help on this?

[lasersafe1]

The magnetic forces are so strong when power is applied that the slightest misalignment internally will cause problems. You will not notice it when power is not applied.

The manufacturers will tell you to NEVER pull the end caps off. The alignments and screw torques are critical for proper operation.

Start putting all the ones that click in a single location. When you get enough of them you can make a boat anchor.

[samco]

Are you sure it isn't you computer outputting a crappy pulse train?

sam

[Dougal]

Are you sure it isn't you computer outputting a crappy pulse train?

sam

No I'm not. Is there any way I can check?

The magnetic forces are so strong when power is applied that the slightest misalignment internally will cause problems. You will not notice it when power is not applied.

The manufacturers will tell you to NEVER pull the end caps off. The alignments and screw torques are critical for proper operation.

Start putting all the ones that click in a single location. When you get enough of them you can make a boat anchor.

Wouldn't misalignment make them tick in the same place every revolution? I've got every 2-3 revolutions and it's in a different place when it rotates the other way.
The ticking is in all of them to some degree (now I've had a closer look), but it's barely noticable on one and terrible on the one in the video above. The third lies between the other two.

So ultimately I'm looking at a motor problem (if I can rule out any pulse quality issues) which will most likely be solved with new motors. In that case I've got nothing to lose with experimenting with these.
I'll pull the endcaps again on the worst one and check the bearing fit. I'll try and spin the rotor in my lathe and check how true it runs. I'm not sure if the 3 jaw chuck I have will hold the shaft well though.

[Spiral_72]

Swap the steppers around. It might not prove anything, but it's a free check and only takes ten minutes or so. At least it would rule out the drivers.

You can check the output from your computer with an oscilliscope (SP?) It should be a nice square wave.

I've also heard of parallel ports that don't bring the output up to 5V, sometimes as low as 3V. It may cause problems with your driver. Check with a meter.

Last, check you have a true 5V on each of your driver IC's.

Just an idea.

[Dougal]

Swap the steppers around. It might not prove anything, but it's a free check and only takes ten minutes or so. At least it would rule out the drivers.

You can check the output from your computer with an oscilliscope (SP?) It should be a nice square wave.

I've also heard of parallel ports that don't bring the output up to 5V, sometimes as low as 3V. It may cause problems with your driver. Check with a meter.

Last, check you have a true 5V on each of your driver IC's.

Just an idea.

I've already swapped that stepper onto different axis (each runs it's own power supply) and different driver cards. The worst motor remained the worst motor.

The breakout board is isolated and has a 5v power supply. So I'd expect a 5v signal to the stepper drivers (gecko G250's). I'll check what I can with my digital multimeter but an oscilloscope is out of the question.

Thanks for the help, I still haven't opened the motors again.

[Pandinus]

Where are your step / dir pulses coming from ???? older with windows on it??? low end control software ??? the glitches look to be every second.... OS timing issues?? software DRO update timing??? Turning of real time DRO or onscreen graphical tool position update may help some. Even better - set up a 555 timer for your step pulses and see if the hiccups stop...

[Dougal]

Where are your step / dir pulses coming from ???? older with windows on it??? low end control software ??? the glitches look to be every second.... OS timing issues?? software DRO update timing??? Turning of real time DRO or onscreen graphical tool position update may help some. Even better - set up a 555 timer for your step pulses and see if the hiccups stop...

Good spotting with the 1 second timing. I'll run the same test at a different speed to see if the 1 second tick remains or not.
The control is mach3 running on xp via a dell 600m laptop. All my computers here are laptops, I don't have a desktop to try unfortunately.

[2muchstuff]

I've also heard of parallel ports that don't bring the output up to 5V, sometimes as low as 3V. It may cause problems with your driver. Check with a meter.

I've heard of this also.

[lasersafe1]

When I commented earlier I had assumed that the driver (and windows) had been proven to be fine by the fact that one motor did not misbehave. If not, then by all means you want to put your good motor on all drive channels and make sure it is still good.

Could it be the motor and still only occur every third rotation? Hmmm. Odd. One would assume the step sequence would always make the same magnetic force at that point during every rotation, but sometimes a wave drive sequence might bypass a certain point for 3 rotations. Who knows?

You cant beat the swap out method for finding a fault. Just keep conditions the same during all swaps. i.e. presumed good motor to presumed bad location and vice versa keeping all channels loaded with a motor.

[lasersafe1]

whoa... Just saw the fact that you are running Mach 3 on a laptop. This typically doesn't work because the parallel port is polled, and yes, polling could be your problem.

You can find lots of info about this on the Mach 3 website. You need to pay attention to the time in interrupt number. If it is large, you're screwed.

A good P4 desktop is only $69 from Geeks.com

[lasersafe1]

More info on "lost steps" Mach 3 and laptops.

http://www.cnczone.com/forums/showthread.php?t=53254

[Spiral_72]

Granted I didn't pay it much attention because it didn't apply to me, but.....

I did read something about running a USB-parallel port adapter. I know there was something special to get it to work. I THINK there was a driver from Mach3

[Dougal]

More info on "lost steps" Mach 3 and laptops.

http://www.cnczone.com/forums/showthread.php?t=53254

Interesting, thanks.
Does anyone reading this thread know how to find and run the mach3 driver test? I've asked the same question in that thread.

The multimeter shows the bob is getting 4.98v and output on the direction signals to the stepper cards are 4v when it's idle. I haven't done a running test, I'll try it later.

Interestingly enough, mach3 appears to have disabled all the power management in that laptop by itself. Machine doesn't power itself down or even scale down the processor speed at idle. It's running the cooling fan all the time, it doesn't use the fan at all if it's below 800MHz, clock speed is 1.4 MHz.

Funnily enough this setup is a stop-gap until I manage to get my USB CNC-Brain working.

[Dougal]

Problem solved.:rainfro:

Thanks to Pminmo for his webpage with the easy steps to happiness with a laptop running mach3. I followed his instructions here: http://pminmo.com/PMinMOwiki/index.php5?title=Laptops

The rightmark utility solved the problem for me straight away. I now have smooth and sweet motion from two of my steppers (one of those is the ticker shown in the video up top), the third is smooth but noisier and is probably an alignment issue.