We are having an issue where, cyclically, the regulator will stop switching every 150ms for 1.5-4ms, resetting all our electronics. We do not know what is precipitating the pauses in switching.
Our input voltage ranges from 7V down to 3.9V, our output voltage is 3.3V.
As the input voltage reaches 4.15V (roughly, varies a little), the regulator begins a cyclical pause in operation. Every 156ms, switching will halt completely for at least 1.68ms, with additional smaller pauses (presumably as the regulator resets). This resets our electronics.
After this halting behavior, the regulator resumes normal operation, until about 156ms have passed and then it pauses operation again.
The cyclical resets do not stop until the input voltage is raised again to roughly 4.5V.
During all of this, the ‘Power Good’ output will indicate correctly, as shown in this image with the green trace.
Increasing the value of C_out has little to no impact.
Switching C_out to a higher ESR tantalum makes the voltage drop on the output by just about 100mV or so. Problem remains.
Decreasing the value of C_out to 10uF has some small impact to the oscillations that are sometimes present, but the problem remains.
Replacing our implementation with an output from a TPS629210EVM yields the same results, but worse. The resetting behavior starts as early as approaching 5V, instead of 4.15V
The first pause in switching lasts about 1.685ms, then it alternates between the two following states, one with 2 breaks in the switching, and one with 3 breaks in the switching.
Sometimes the 2nd pause has an pulse train.
Sometimes the 3rd break is longer.
Since this behavior is being replicated on both our implementation and the evaluation board, we are thinking this could be some behavior of the regulator we do not understand, or are misinterpreting. Perhaps some aspect of the DCS-Control behavior.
We would greatly appreciate any assistance you can provide us with this.
Thanks for your time.