I am having a hard time getting Closed Loop Speed Mode running. I have my motors geared down to 22 RPM with the front and rear wheels chained. I have an external encoder attached to the second wheel shaft. It is a US Digital H15-1250-I-S.

Connected with RoboRun+ I have everything but Serial and the Encoders disabled. When I run the motors that start to spin up to speed but will just stop ever couple of seconds and then just start again.

Any Ideas what could be causing this?

Thanks

First make sure that you see speed reported in Roborun while in open loop. Then verify that speed is positive when the slider is moved in the positive direction.

Then set the PID to have moslty a I gain (eg. P=1 I=2 D=0).

Close the loop and it should work.

Here's an update. I've attached the profile to this post.

The setup is the same, with motors on each side, chained up to encoders:

Encoders: US Digital S5
usdigital.com/assets/datasheets/S5_datas...k=634792491037187977

Motors: 24V 252RPM

In open loop they run fine. The left motor is motor 1 and the right is motor 2. Due to the motors being on reversed sides I have to drive the right to a negative value to get it to move in the correct direction.

The encoders being used have 500 pulses per revolution and the motor controller is set accordingly. Encoder 1 is set to feedback to motor 1 and encoder 2 is set to feedback 2.

It's set in Closed Loop Speed Mode with the PID set to 1,2,0 for each motor. When applying power to the motors in the Roborun software (run tab) the result is an oscillation in both speed and direction and traveling no faster than 20 RPMs in either direction. The loop error detection is 1s @ 50% Error.

Now, when I swap the polarity with one of the motors, it seems to work better but the motors max out and will not stop unless I hit the emergency stop button. They also do not react to a power less than about 250 and both motors activate when one of the sliders are moved, it doesn't matter which I move.

If I tweak my setup by having both encoders set to Feedback(1,2) then motor1 seems to function properly with the motor control slider but the second motor control slider has the same effect in the previous situation (both motors go full bore and will not cease until I hit the emergency stop button).

EDIT:// Flipping A and B on the encoders seems to have fixed it.


Any advice

Once I was able to get the single SDC2130 running smoothly I moved to using two. On the second I loaded it with the working profile of the first but I run into periodic jitter. Every couple seconds power seems to get cut abruptly and the wheels jerk. To verify that the encoders and wheels weren't the problem I used the working SDC2130 in both situations and it worked.

I then moved to replace the non-working SDC with a fresh one and the problem persisted.

Any thoughts on what the problem could be?

Encoders: H15-1250-I-S

What is the firmware revision and date on each unit?

Report the controller responses to ~zpac sent from the console?

Roborun is reporting that both have the v1.2 RCB100 10/19/2011 firmware

Jittering controller:
ZPAC=309:308

Working controller:
ZPAC=318:318


I can't find a reference to zpac in the manual, what does this mean?

So after a good amount of time messing around and trying to tweak the system to smooth everything out I think I am onto something. I put the controller back in Open Loop mode and observed the following.

We have a 1250 PPR Encoder (5000 CPR) on a 122 RPM Motor.

The system starts out running pretty good and then starts jittering.

I pulled up the Speed (RPM) on the graph and noticed the same time it starts jittering the speed value jumps to around 13k. Clearing the encoders and driving slow the RPM is correct until it hits around 32767 encoder counts. I stopped it there and was able rotate the wheel back and forth across that count and see the speed value jump.

It seems like the math is using a 16 bit number while the encoder values are using 32 bit.

I attached my profile settings and had one question. What value should be set for Position Turns Min To Max.

Thanks
David