This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPA3255: TPA3255 Low impedance

Part Number: TPA3255
Other Parts Discussed in Thread: DRV5825P, TAS5548

Not sure if this is the best form to ask this, so please direct me to another form if you think it is better suited for this question.

We have a project that requires using a high current amplifier (~20A pk-pk) (or high current inverter) with high bandwidth (100kHz) and low distortion to drive capacitive loads with ESRs in the range of 0.01-0.5 ohms. 

Given the bandwidth and distortion requirements we thought audio amplifiers would be a good place to start, and given the high current requirements we thought Class D amplifiers would be a good option to avoid the complication of a high power linear amplifier. 

The TPA3255 is pretty close to fitting the bill and has some nice features (built in OC/OT protection, built-in feedback, analog reference), however the application we are using it for is certainly non-standard (low impedance loading, capacitive loading, frequency is a bit out of range). We've done some testing with the EVM and it seems to be working pretty well in the application, we are able to get the 17 amps at a wide bandwidth across low impedance loads in PBTL mode and the heat sink remains fairly cool with a bit of airflow. 

That said we have had some amplifiers blow up on us. The blow ups seem to be consistently happening on the output pins (and it definitely smells like blow fets when it happens), so we think we might be pushing the internal switches a bit too hard with the low impedance load. Is this a valid explanation?

We are not sure if the TPA3255 will be suitable for the final design or whether we need to take a different approach.

Are there any chips from TI that might be better suited here? (Ideally with the nice features mentioned above (built in OC/OT protection, built-in feedback, analog reference) that we could use as an alternate? We're currently exploring H-bridge driver with external switches, but this would complicate the design somewhat, given that we would also need to design the feedback and generate the PWM for unipolar modulation. 

Thanks.