DRV8214: Current mirror output

Hey Tom,

Thank you for bringing 1. to our attention, I've notified our systems team of this.  The correct value to use is the one in the Electrical Characteristics table, 244 uA/A.  

Tom Moyer said:

2 - There seems to be a ~60mV shift in the current mirror output under certain conditions.  We can reliably duplicate the error by coming out of sleep mode when the motor voltage is turned off.  Coming out of sleep with the motor voltage on reliably functions as expected.  We are running motor voltage at 3.3V and using an IPROPI resistor value of 4.02K with a gain setting of 244uA/A).

Is this on an EVM or your own board?  If the EVM, can you post a picture of the top of the board?  I want to make sure it is MD069A or B, not E1 or E2 with prototype silicon.  

Which control mode are you in - PH/EN or PWM?  Is the shift visible in both forward and reverse, and does toggling nSLEEP to low and back to high make the issue go away? 

Tom Moyer said:

coming out of sleep mode when the motor voltage is turned off

Just to clarify, does this mean VM=0V (off), set nSLEEP=HIGH, then turn on VM=3.3V, set device settings, start running a motor?  Trying to get enough information that I can try to replicate it here on an EVM. 

Best,

Jacob

This is on our own board.  We are operating in PH/EN mode.  I believe it is the same in both directions, but can't be sure I'm correctly remembering all the combinations we tested.

VM=0V (off), set nSLEEP=HIGH, then turn on VM=3.3V, set device settings produces the issue, toggling nSLEEP clears it.  Our original design had the same signal controlling nSLEEP and the VM regulator EN, so they weren't independent, but as long as we brought nSLEEP back high before VM dropped below about 2V, everything was good.  We have modified our original board to separate the control signals to be sure we were identifying the correct issue.

Hey Tom,

Thank you, sounds good I will work on recreating this in the next couple of days.  If you could try to recreate it on an EVM if you have one that would be helpful, but it's okay if not.  

Thanks

Jacob,

We were able to recreate the error on the EVM using the PC software controls.  We used a 100mA resistive load for simplicity.  If the chip was brought out of sleep when Vmot (Vsup) was present, the IPROPI output was as expected (36mV).  If the chip was brought out of sleep when Vmot was not present (0v), the IPROPI output was low (20mV).  We also found another error in that the current report by the software was 50% higher than the value shown by the IPROPI output.

Any progress in duplicating this error or understanding what is causing it?