FPGA PWM motor control "disabled by drive fault"

[Sam Jackson]

Hi,

This summer i'm working on a ballbot self balancing robot-sort of like an inverted pendulum but on a ball (childs bowling ball) instead of a cart. I have the labview program setup and everything in place pretty much but my 9505 driver keeps on cutting out. The motors i have can draw 16 amps and 24volts max, however i know that the drivers max current is 12 amps (spike). I am therefore limiting the voltage to 10v with the power supply. Also the power supply apparently only goes up to 5amps.

However having said this when i run the motor, (i'm using the inputs from an IMU to produce a pwm signal) the 9505 driver cuts out if i do a few quick jerky movements and change the voltage form +10 to -10. I don't see why this is causing it to cut out and say "disabled by driver fault". I have the FPGA PWM motor control program showing the current sense on the front panel so i can see what is going on. The current sometimes peaks at 2000 (2000 x 0.0062 = 12.4 amps (with 0.0062 conversion factor) which is over the 12 amp spike limitation. However sometimes the current won't even hit the 12amp spike before it cuts out, sometimes it's around 7amps? I'm not really sure why this happens, i could understand if there was always a spike at a current over 12amps which caused it to stop but not below?

Just for reference i'm using a cRio realtime controller with a 9205 analogue input attached to a 5 degree of freedom inertial measuring unit. The measured inputs then go through a control loop implementing kalman and lqr techniques to produce a desired motor voltage to invert the pendulum. This voltage is turned into a pwm motor signal through an FPGA read/write and then this is passed through the Ni 9505 driver to power the motor. This works great without the ball in place-just running the motors with no load, but when the ball is in place the extra load causes issues.

Any help on the matter would be greatly appreciated.

Sam

[Daniel_Thomas_NIUK]

Hi Sam,

It sounds like you are hitting the 12amp limit, however the application is cutting out before you have chance to read it, hence the apparent 7amp cut out. With the 'back' torque being produced by the extra load of the ball it sounds perfectly reasonable that the current draw could be this high.

You really need to reduce the current draw or voltage to avoid this spiking effect.

Kind regards,

[Sam Jackson]

Ok i've done some more testing today to see what happens at very low voltages etc. I have saturated my control part of the program to 6volts which is the lowest it can go. This however still causes the driver to fail. It seems very odd, sometimes the spike is at 3 amps then next time its at 4.6 amps. Sometimes i'm not even accelerating the imu quickly and it stops and also quite a lot of the time when i run the realtime application the driver will be disabled when started. The motor will make a little sort of bump noise like it has moved a tiny amount and then it cuts out straight away.

I have tried the driver/motor at 10volts and it does indeed spike at over 12amps when it cuts out but likewise it seems very random and will often cut out when i don't even accelerate the imu quickly.

I can't understand what's happening as there doesn't seem to be any pattern, only that it keeps on cutting out.

Thanks for you help so far, 

Sam

[Sam Jackson]

Just tried it with no ball in and it cuts out with a 6volt supply too. It seems to cut out at very low currents around 1-2amps and im moving the imu very slowly. Maybe there is a fault in the driver?

[Daniel_Thomas_NIUK]

Hi Sam,

I can't seem to find anything on our database that would suggest that the 9505 couldn't deal with the figures that we are discussing here. The error message  "disabled by driver fault" would suggest however that there could well be a fault with the driver. Have you tried contacting the driver manufacturer?

Kind regards,

[Sam Jackson]

More tests have been done with just a standard pwm signal that i vary with a slider on the front panel. When the power supply is at 10v I can vary it quickly enough to cause the current to spike at 12ish amps and it then fails. Is there a way to limit the current on the 9505 driver in the actual program so that if say it wants to draw 14amps or something i can limit it down to 11 or something?

Also not too sure how the motors are actually drawing 12 amps at any stage since the power supply is 1-20v and 0-5amps. So the max that could be drawn is surely only 5 amps aswell? Or can the power supply also spike as well? The power supply i'm using along with the manual are in the links below:

http://www.maplin.co.uk/1-20v-0-5a-slim-bench-power-supply-with-lcd-screen-219129

http://www.maplin.co.uk/media/pdfs/N93CX%20manual.pdf

Thanks again for your help!

Sam 

[Daniel_Thomas_NIUK]

Hi Sam

Just a quick question in the meantime. How many motors are used to run this machine and do you have a link to the specs on them?

Kind regards

[Sam Jackson]

There are two motors, but currently (no pun intended) i'm just testing one in one plane to not overcomplicate things for the time being. The controll is exactly the same for the other motor i just need to double it up once i get the one motor working correctly. Below is a link to the datasheet for the motor, it is the 24v DC one:

http://img.ebmpapst.com/products/datasheets/BCI5260C00-615509.pdf

Regards,

Sam

[Daniel_Thomas_NIUK]

Hi Sam,

We are taking all this information on board here in the office. However as you have a direct service request open with Mike. It would be far more efficient if you could proceed with your enquiry through him.

We will update this thread when we find a solution so that other users can benefit from the information found.

Thank you for your understanding.

Kind regards,

[IanC_]

Hi Sam,

Might your drive fault be being caused by an undervoltage error?

If you're trying to pull more current than your power supply can provide, that will drag down the supply voltage.  Once the supply voltage drops below ~6 V it will trigger a drive fault on the 9505.

Regards,

Ian