We're using the DRV595 eval board to drive the TEC of a sensor. That TEC requires +8.5V. When this voltage is applied to the chip and the chip is turned on, the voltage amplitude will change depending on the differential input as expected. However the outputs do not seem to be switching and the load current is not acting as expected. It's as if the H-bridge is turned on in only one direction, which becomes very inefficient. When the two inputs are similar, the output will oscillate and appear sinusoidal.
Are we using the DRV595 correctly? Is there a way around this problem?