We are struggling with a design containing an ADP2325 buck converter that will be used to generate a 1V for an FPGA.
(It would be the first time using this part)
Here are some of the relevant values:
- 5V input
- 1V output (Parallel Buck config)
- Osc frequencey (tried 300KHz and 1MHz)
- SCFG internal clock
- Inductors (Sumida CDRH between 2.2uH and 10 uH tried)
- Cout 2x100uF
- Comp C 820pF
- Comp R 47K Ohm
- FET FDS8880
So for an initial test we disconnected the 1V from the FPGA and built used a dummy load of about 3 Ohm to verify
if the output voltage looks in the correct range (FPGAs are expensive). The parallel config was chosen since the FPGA might
end up using a lot of power.
Sadly things aren't working as advertised. We see 2 "modes" currently depending on the selection of the dummy load, chosen inductors
and Osc frequency selection.
- Mode one looks like the voltage has a massive ripple. It looks like a very short duty cycle with a long discharge time. There does not seem to be any filtering of the output voltage. We tried to change the load (more current), the inductor (higher value) and the caps (more) to see if we can filter it out but to no avail. It looked like the ripple current was so high that the bench supply had trouble supplying the ripple. The output is however avg of 1V but the inductor "sings" quite a bit.
- The second mode looks like the ADP does not want to do anything at all. The output voltage is stuck at around 3V, there does not seem to be any sync out being generated. The gate voltages of the FETs are all constant. The feedback voltage is about 1.8V but the power good is high for some reason.
It feels like we are missing something basic. How hard is it to break one of these converters? We have tested 2 boards with no success.
Any suggestions/ideas on how to start debugging this would be much appreciated.