Tool/software:
Hi Forum,
i've used the DRV8840 in a recent design and cannot get it to work. I've tested two different boards, both were manufactured by a professional company so soldering is not an issue.
I've used two different BDC motors as well as a ~2mH Inductor (magnet wire from the lab as a simple coil).
Measuring the voltage over Rsens I get the following behaviour:
zoomed in:
The yellow shadow is due to persistent turned on,.. the Voltage is all over the place and far from the set max current. Also, the current seem to drop down back to zero before the next cycle starts,... although the internal frequency of 50kHz can be measured.
I've tied the nSLEEP to 3.3V (external supply on the board). All other Pins are tied to 3V3 from the DRV LDO Out,... either via 10k or directly (checked both options,.. not working at all).
The schematic is the one from the board and has some features which i've removed for testing:
* the Diode in parallel to the motor is not placed
* the nSLEEP is no longer tied to the res array, but to the ext. 3V3 rail.
* the EN pin is not actuatet with a PWM but with a constant 3V3 signal.
* All Ixx pins are tied to the "3V3_driver" rail
* reference voltage, 3V3_driver has been measured correctly
And the last thing: the layout. There are many vias and thin traces, but the motor was designed to run at 300mA,.. thus the layout is not that high power.
I have no clue at all on what is goin on here,.. did you have any sugestions?
Thank you and kind regards,
Tony