EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: presse55 on October 03, 2018, 08:18:23 pm
-
Hello folks!
I've recently gotten a Rigol DS1102E with the ambitious goal of troubleshooting this A4988/Stepper which turns verv unevenly as described in this post (https://www.eevblog.com/forum/testgear/using-an-oscilloscope-to-view-current-waveform-through-winding/msg1852160/#msg1852160).
Recaping the stepper’s characteristics:
- Stepper rated voltage: 12V
- Stepper rated current: 0.33A
- Supply voltage: 9VDC battery, circuit is floating, ie not ground referenced
- L=48mH
- R=32.6ohms
- f= 10Hz - 100Hz (ie 10 to 100 full cycles of 16uSteps per full step per second)
Using the 2ohm shunt resistor, see attached waveform, with 250mA/div.
At first glance, there's a lot of spikes there ::)
Anybody care to help me steer in the right direction.
Am I misusing or expecting more from the scope than it can deliver?
Thanx in advance for helping this noob totally new to oscopes!
Ron
Montreal
-
I'd be a little concerned that under-volting the stepper by 25% will have a large impact on the available torque. A 12V 0.33A coil should have a resistance of 36 ohms. At 9V that gives you a max of 0.25A. Which is nearly half the power. Over-volting is more common, which works ok with drivers like the A4998 because they limit the current.
Micro-stepping also reduces available torque. So one thing thing to try would be to turn off micro-stepping, just to see if the behavior goes away. If you don't actually NEED that much resolution for your task, it may be better not to use it. If your minimum move at 16 microsteps is 75 steps, perhaps 5 full steps would be just as good. Or 9 half-steps. You get the idea.
But before that, I'd monitor the step input signal to the A4998 to see what it's being told to do. It's possible you're chasing a software behavior and not a hardware problem. Wouldn't surprise me at all if you're moving 75 microsteps in a fraction of a second, then pausing for the rest of the second. That would be audible.
-
Thank you Nusa.
I've run this stepper on 18V, 12V, and 9V supplies, and you are right that at 9V the motor will be underpowered.
At the moment, I'm more focusing on finding the cause of the uneven stepping.
Running with these three supplies didn'nt change anything with respect to uneven stepping.
You raise a point which I'm also suspicious about, ie software.
As it turns out, the pulse signals are controlled by one of the timers from an Arduino.
I checked, and the pulses are 'well behaved', evenly spaced, and timed as intended.
I'll post a trace showing pulses & waveform.
Thanks again!
Ron
-
What does this mean?
f= 10Hz - 100Hz (ie 10 to 100 full cycles of 16uSteps per full step per second)
To what value did you set the driver current limit?
9V is useless. Microstepping won't work properly if the driver current limit is not very carefully adjusted, and that is very difficult to do with a high impedance motor and such a low voltage motor power supply.
Furthermore, none of these drivers works well with high impedance motors, because the decay time is usually not properly chosen. See this informative post http://cabristor.blogspot.com/2015/02/drv8825-missing-steps.html (http://cabristor.blogspot.com/2015/02/drv8825-missing-steps.html)
-
I'm testing the motor at low speeds - between 10 and 100 steps per second.
Since I'm setup at 16uSteps per step, this means between 160 and 1600 uSteps per second.
I've gone back to testing with an 18V supply, and checked to make sure the current limit was at 330 mA.
Thankx for the DRV8825 article; let's see what I can make of it!
Ron