Hi TI.
I have made a prototype board with the DRV8814 with the purpose of driving a high-frequency resistive circuit (a circuit that requires a >80kHz bipolar square wave input, +/-36V, 1.5A). I have attached a schematic of the circuit. (I have later realized that DRV8840 is a parallel version of DRV8814). The INA213 and LTC1799 was not mounted.
I power the circuit with 24V and connect an oscilloscope to the output. I have connected DECAY to ground, xVREF to V3P3. When I wake up the device (temporarily putting 3.3V to SLEEP pin) I get a good V3P3 reading, the and the device output is shorted (as expected in slow decay mode.
When i enabled the device (xENABLE to external 3.3V) while the xPHASE was floating, the device output changed to a cyclic 3.9Hz response and the device heated up fast. I am pretty sure I broke the chip, as It now heats up and outputs the same repetitive signal without being enabled. I repeated the whole thing with a new board just to make sure...
The questions are: Why did it break? I have no load (only oscilloscope) and thus should not fry the bridge even with uncontrolled phase?
Can the SLEEP, RESET, DECAY and VREF be supplied from the V3P3? (With the device needing a 3.3V 'wake-up' to start)
Can you suggest an alternative chip that can be switched faster than 100kHz?
Thanks in advance,
//Rune
