No response from motor

I am trying to set up a 2 DOF simulator using Pololu 12v12 cards.

I haven't had much luck in the past using potentiometers. I don't have a good sense about how to limit the range of motion. Somehow the motor arm always swung too far and ripped the pots apart or popped the glue holding pot to the motor hub.

To get around this I made Hall sensors. Since there is no physical connections to ruin, I can tweak the Hall sensor manually to limit the range. I actually moved the sensor head within the magnetic field to get the range right. This worked well for the first motor-Pololu pair. I can control the motor via mouse through the Pololu software or X-Sim.

As I was setting up the second motor, it freewheeled – it circled round and round – before I could shut it down. This happened several times. I assumed the sensor head and magnet were poorly aligned and I moved it about. No help. I thought I had reversed output and voltage-in. I double checked and this was not the case.

I checked the Hall sensor under Windows joystick calibration. It was fine and outputting signal.

I deleted the axis profile in X-Sim and started again. No movement under X-Sim or Pololu set up tool.

Any ideas?

I use pots successfully here.
On a seat mover running Jrk12v12.


Also make sure you are setting the min max in Jrk Configuration Utility
As in Feedback - Reset to full range - Apply settings to device.
Learn - set Min point for lever - OK - set Max point for lever - OK - Apply settings to device.
Now before testing this new setting make sure your PID is set too.
Then on the Motor tab - Detect Motor Direction, checks the direction of motor against the direction of the pot values if all correct it will jerk slightly off centre and exit, if not it will inform you that it inverted motor direction and you need to press Apply settings to device.
And another one to check , check you have it wired correctly.

Once all correct test in the Input tab for motor movements, if you click in the scroll bar , you can use the arrow keys to move finely or page up/down to step or home/end to go to min and max, now you can test and tune the pid if needed.

Then they are ready for Sim Tools, much better than the dreaded Xsim (excuse me swearing) lol

Eaorobbie,

Thanks for the speedy feedback.

Updated the Pololu 12v12 firmware to v1.4.

Under the FEEDBACK tab I "reset to full range" and "applied settings to the device."
I clicked "Learn."
I moved the lever Forward, clicked OK. Moved the lever backward, clicked OK. "Applied settings to device."

I got lost on the PID part. (Actually, I don't know what this is.)
Currently all the coefficients are set at 2.
PID Period = 25
Integral limit = 1000
Feedback dead zone = 2

As these settings got a signal to the motor before, I didn't tampered with them.

Under the MOTOR tab, I tried running the Detect Motor Direction utility. The pop up window read: output not centered. Center the output, then click OK. I clicked and the same window returned.

How do I center the output? I thought returning the lever arm roughly between its full forward and full backward range would do the trick.

Still stumped. Ideas?

Another question, under Manage Devices in Windows 7 the Pololu board has two COM addresses. One for Command Port and one for a TTL port. Which do I assign to the axis in SimTools? Does Command Port or TTL Port take precedence or does it matter?

Cheers,

Flying Colander

Eaorobbie and other Novices,

Thanks for your check list. I found the problem.

My home made sensor head was rotated too far. Once I had the sensor, which was mounted on the simulator across the room, hooked up to Windows joystick calibration I could see that the sensor read was pegged at the limit of usable readings. Once I centered the sensor head, all fell into place.

Again, thanks.

Any ideas about Command Port vs TTL Port?

Flying Colander

Ok cool, nice to see its working in the Jrk Configuration Utility.

Now Pid is a little technical to setup, a bit a trail and error depending on what motors you are using , I would start at P = 2/2 and I = 0/2 and D = 1/1.
Now test with the slider, if it seems to react slowly. Increase P until the motor begins to occilate then half the P value that should be a good base for P, now test it by setting a value in the Input tab and "Set Target". If it overshoots the target position and returns to the target, increase the D value, beware going to high in D will also make the motor occilate back and forth. They key is to find where it moves the best.
And all motors are different.

Ok now we need to communicate with the Jrks with the Command Port.
So when setting up SimTools , a typical Jrks setting are as follows -
Interface Settings - Interface 1 .
Comport - Command Port of the Jrk for Axis One
BitsPerSec - is the default baud rate of the JRK - 9600
Data Bits - 8.
Paraity - None.
Stop Bits - 1.
Output - Bit Range = 12 JRK.
Output Type - Binary.
Startup - Output - <223><63> @ 25ms. as you have set it to 25ms in the JRK Util.
Interface - Output - <Axis1> @ 25ms (which refers to the PID Period in JRK)
ShutDown - Output - <255> @ 25ms.
Save.

Ok this all can be set automaticly too by selecting -
Output Type - Ser.
Presets - Jrk 12bit.
Set the Comport to the command port and set the BitsPerSec to 9600. and Save.
Just beware when setting the one for Axis 2 make sure you enter Axis2 into the Interface Output or both Interfaces will only move on Axis1 commands from SimTools.

Just a follow up. Motor is working fine. Thanks for the step by step instructions. Very kind of you to take the time, Eaorobbie.

A reminder to other novices setting up their motors. My problems were accumulative. A series of mistakes adding up.

1. My homemade hall sensors worked well on the bench but their angle of rotation was quite narrow compared to an old fashioned rotary pot. So I changed back to pots.

2. The epoxy I was using to cement the pot arm into the motor cracked repeatedly. The best way to mate the pot arm to the through-bolt in the motor is to fasten it with a screw. See Eaorobbie's picture above. He's drilled out the through-bolt, slid in the pot arm and run a linchpin through both.



3. Don't forget the same trick for the through-bolt and motor. I forgot to do this. The motor rotated but the through-bolt was just clattering about. There's a tiny 4mm hole in the axle sleeve for this. Make sure you have a strong linchpin for this. I was just using a nail which sheared off due to the torque.

Thats the right idea, myself I turn the bolt the other way around so I can weld the lever to the back of the bolt head and drill out the other end to mount the pot through, just beware the torque of change of direction these motor will produce might undo the nuts.
Should not be too long now and we should see ya sim moving, cool, nice work mate.