Problem with custom PCB of the DRV8301-69M-Kit

Hi Lukas,

First impressions based on what I see and what I might recommend investigating:
1) you mention that "the connection to the MC was cut" - I assume this means that the GUI/CCS disconnects?  If so, then EMI is likely causing issues for this.  I might recommend putting an EMI bead on the USB cable.  I might also recommend using an isolated emulator to connect to the C2000 device - if you want to debug via JTAG while the motor is running.  Or add in something like the TMDSADP1414-ISO to improve the debugger connection.
2) Your trace/component clearances in certain places - like between the "VCC" and "GND" connectors - seem very tight.  As you're connecting things up, make completely sure that you're not creating shorts.
3) I do not see any major bus capacitance on the DCbus.  Note that if you are applying torque to slow down the motor, the DCbus voltage will rise.  Vbus on the board may then get sufficiently high to damage the inverter circuitry.  Bus capacitance helps this - to some extent.  Another factor in this involves the power supply that is generating the DCbus. I might recommend monitoring the bus voltage during operation.
4) Your decoupling/etc caps for the DRV chip seem to be a good distance from the DRV chip - this can reduce performance to some extent, but I doubt it's the root cause of your issue


Thank you,
Brett

Hello Brett,

thank you for your answer. We've now added a bus capacitance and managed to establish a quite reliable USB connection to the board. However, in few cases we still get disconnected, but mainly because of high current consumption during identify and the resulting undervoltage. Now we're trying to fix this issue by adding a higher capacitance to the intermediate circuit. We're using using an external custom-made programmer with isolated connectors and four of the ISO-7221 IC.

Unfortunately, we're now facing yet another problem. Our modified example program which works quite well on the reference design but not on our self-build board. We think that this problem might be due to the different MOSFETs that we used, although the R_ds value is almost the same (SUM110N06-3m4L: R_ds_typical = 2.8mOhms at Vgs = 10V vs. our MOSFETS IRLB4030 R_ds_typical = 3,4 mOhms at Vgs = 10V).

Do we have to change the Rds values in the DRV or are they identified automatically? Is is also important that the gate threshold voltage has to be exactly the same as the original MOSFETs?

Another issue that has occured is that the motor makes a high-frequence noise with our self-made board and the same project that worked well on the reference design. Do you know what could be the reason for that?

Thank you and best regards,

Lukas.

Lukas,

It is my understanding that re-running Motor ID on the new board will resolve any minor hardware changes to the overall design.  Your change to a different set of FETs shouldn't be a problem.

Your high-frequency noise likely either comes from [A] something wrong with configuration (something like motor ID) or [B] some kind of problem in layout.


Thank you,
Brett

Hi Lukas,

The gate voltage and Vds trip point should not have a large effect.

The main concern would be if their were big differences in the gate charge of the MOSFETs. This could lead to much faster switching speeds even when using the same gate drive setup. Faster switching speeds can lead to more noise in the system and over/undershoot voltages.

This would not explain the high-frequency noise though.

I would recommend take a look at some of the MOSFET gate and switch node voltages and comparing between the two designs.