Using VPROPI on DRV8801

Hi Raphael,

When the H bridge is disabled, there should not be any current. There might be some while the current decays, shortly after disablement. But once current reaches zero, the outputs are tri stated, so there will be no current flowing through the FETs afterwards.

This also means there is no current flowing through the SENSE resistor, so the VPROPI output is to be completely ignored. VPROPI only makes sense when there is current through the SENSE resistor. Do note that during slow decay, there is no current flow through the SENSE resistor, so VPROPI cannot offer information either.

Hope the info helps. Best regards,

Jose Quinones

Hi Jose,

Thanks for the promp reply.

I understand all this from the datasheet and the app notes, but it doesn't quite answer my question which has to do with a fault detection. I need to detect a fault condition when the H-bridge fails such that the motor is energized even if the ENABLE pin is low.  Not knowing all the internals of the driver, I assume that in my fault scenario there will be current flow through the SENSE resistor. If VPROPI amplifier is independent of the H-Bridge operation and  ENABLE pin state, I would think there will be current indication on VPROPI pin.  Also, there is an offset voltage at VPROPI pin at zero current that needs to be included in the math when the driver is active.  With a signinficant idle drift of VPROPI signal the current measurement acuracy degrades.

Thanks,

Raphael

Hi Raphael,

I think I understand the problem, but what I am having a hard time is accepting the possibility of the motor being energized through the device's FETs when the H bridge is disabled. But let's assume it does happen. In this case you should see VPROPI output being 5X the winding current. As you stated, the VPROPI output is independent of the H bridge. What it is not independent of is the SENSE node. If there is a voltage here, you will get 5X at the VPROPI output.

I am not aware of an offset on the output while there is a current present, but I will check with the design team. When there is no current (H Bridge is tri stated), I would imagine the VPROPI will drift as the amplifier's inputs are floating.

Best regards,

Jose Quinones

Hi Jose,

In our application unless I can quantify the likelyhood of the of a failure, I need to mitigate all single point failures. As unlikely as it may sound, the failure I need to mitigate here is the H-bridge control failure.  Whether it's caused by FETS, control logic, or something else is immaterial.  The bottom line is:  does the VPROPI output produce 5X V_sense regardless of the ENABLE pin state?  If so, is the drift an expected behavior, and what is the cause?  If the amplifier is independent of the ENABLE signal, then it's input is not really floating because R_sense provides pretty strong bias to GND.

Thanks,

Raphael

Hi Rapahel,

To answer your question, yes, the VPROPI output operates regardless of what the ENABLE input is doing or whether there is current or not on the H Bridge. It is a separate block and it works standalone.

As you specified, the SENSE resistor is basically grounded when there is no current. In this case, the amplifier is an undefined state. Design tells me there is about +/- 10 mV of bias in the inputs, so you can see as much as 50 mV bias at the output.

Hope this answers the question, but do let me know if there is anything else I can help you with. Best regards,

Jose Quinones

Hi Jose,

Thanks for the support.  Your last response regarding VPROPI operation answered my question.  As far as the drift, please disregard that. The drift I see is not related to the driver, and it has been fixed. 

Thanks again,

Raphael