BasicMicroUK - Forums

I just got a Roboclaw 2x5 v3 from Pololu. I hooked it up to a bench top power supply and powered it up and checks all the pin voltages against the datasheet, all was good. I connected two motors and an RC receiver, and put the Roboclaw in mode 2. Both motors functioned fine, the bottom of the controller was a little warm but nothing out of the ordinary.

The gear motors I'm using are from Pololu as well and have encoders. When I tried to connect the power for an encoder, the Roboclaw BEC circuit would lose power (4.3v) and the RC RX's error LED would blink (Spektrum, low voltage).

When I disconnected the encoder, everything worked fine again. As I was mulling over the PDF, channel 2 stopped working. If anything other than center pulse is received on channel 2 the error light blinks once then a long pause with status 1 on. According to the PDF that is a driver error.

Between the encoder power issue and the dead channel I'm lost. Is there a master reset button somewhere on this thing?

I should mention that the encoders on both motors check out good. I have them running with 5v on a two channel scope.

I retried everything this morning, and a few new things.

I'm in RC mode #2 and also tried RC mode #1 with a few different options each. I tried using a microcontroller to PWM in RC mode and the same thing.

I also made sure to power up the RC radio first and then let the Roboclaw calibrate.

I noticed you're located about 20 minutes away. Is there any chance I could come there to exchange this?

The 2x5amp roboclaw has a linear power regulator so it can only source 100ma from its 5v rail to outside devices. Your encoders may be drawing more than that which is causing part of your problem. As for the channel not working we'll need to take a look at it. If a mosfet driver failed you should be getting an error message. If a mosfet failed I would expect it to over heat badly. The linear regulator should be a little warm as well but not too hot to touch. It will take some examining to determine why the one channel isn't working since you aren't getting an error and not a significant amount of heat.

I'm posting this so others have the information since I already talked with you yesterday.

Thanks.

I measured my RC receiver and the two encoders and they're 40ma @ 5v.

I mailed it in last week thanks for looking at it. I ordered another one to use in the meantime so take your time. There's no RMA number on the box I sent, it has the original sales invoice from Pololu which has all my info on it.

Thanks.

Yes, I believe we have received it. It should be going back out by Monday.

Thank you. Let me know what you find with the old one if possible.

Hello, I was wondering if you would provide a status on this controller please?

Sorry about the delay on this RMA. It normally doesn't take so long. We had some trouble replacing the driver chip and we just got more 2x5a boards back in stock. We are currently programming and testing them. You replacement should ship out this week.

I haven't touched my roboclaw 2x5 since it came back from RMA until recently. I decided to give it another go the other day.

I'm not getting the error like last time.

Except, in all RC modes, either hooked up to an actual RC receiver, or microcontroller I lose a motor speed control until the RC pulses are centered for a few seconds, then it resumes normal operation until I make rapid changes in pulses they everything goes south.

All modes have been tested. Different power options have been used, everything is grounded, etc.

I am not new to this stuff, there is definitely something going on with this controller again and would like to RMA. I can post some videos if you like.

Could you describe the RC mode problem in a little more detail? Or a video would be good too.

Note that the default RC mode expects to receive some center RC pulses to calibrate its stopped position after power up. If you don't want to use autocalibration(which may be what is causing your problem) then use the MCU option in RC mode. That will disable the autocalibration so that center is always 1.5ms pulses and max and min don't change. If the problem goes away when you use MCU mode then autocalibrate was the problem. If your RC radio sends out garbage pulses on power up that could confuse the autocalibration. If there is noise(eg like you are using an FM RC radio) the noise pulses can also mess up the min and max calibration that is constantly running when autocalibrated is enabled.

Thanks for the response. It would be hard for me to post a video of this now so I'll explain in better detail.

This is using MCU mode, however all RC modes have been tested with a microcontroller and a Spektrum RC rx.

When there is mixing going on with any of the two channels I lose channel two, it comes to a stop. For example:

When the S1 and S2 signals are close to the same everything works. For instance I can go forward, rotate either direction, backwards, all is well with that.

When S1 and S2 signals are more than 100us apart I usually lose control of the S2 motor, until both signals come within 100us of each other, then it will work again. For instance if I move the controller stick all over the place I will lose S2 after a few seconds.

Aside from trying three microcontrollers and one RC radio I have also turned my endpoints down, my original thought was that I was exceeding the endpoint, that didn't seem to help. I have used my scope to ensure the PWM signal is working properly.

It seems as though when the pulses on the two channels are more than 50-100us apart, for instance making a forward left turn (where the right motor is at full speed and the left motor is at half speed) the problem comes into play. As soon as the pulses are close together on both channels, it "resets" the issue until the channels are mixed again.

Have you set the Roboclaw to use MCU mode. This will disable the auto calibrate. Based on your description is sounds like that is the problem.

Also have often are you sending out RC pulses? Roboclaw can handle faster rates but I recommend you stick with the 50hz rate of standard RC signals.

Second line of my last post - "This is using MCU mode, however all RC modes have been tested with a microcontroller and a Spektrum RC rx."

I don't know how else to make this clear to you. It does not work in RC mode except when both channels are receiving the same pulse values.

I am requesting another RMA, if you cannot do that please let me know and I'll take it up with Pololu.

Sorry I just scanned your email and somehow completely missed that second line.

We will take the board back to look at it again of course.

Before you send it back though, does this or something similar(or odd) happen when not using mixed mode?

I assume you are still using the Spectrum radio as well?

Hey it's all good, I really like this controller, and it works fine in packet serial. I just want the option of RC should I need it someday.

It happens in both mixed mode and non-mixed mode. It should be very easy to replicate the problem on your end, you can use a microcontroller or an RC receiver, RC Mode 1&2, RC options 5-8. I have also tested with the onboard BEC as well as external +5v.

Thanks.