Hibernate and small power spikes

A Panda II I’m running with a shield that contains a a quad OP amp IC (LM324AN) running off the 5 volt Panda supply. When the Panda goes in and out of hibernation there is a small ~50 mv spike in the supply voltage. Unfortunately the amplifier as configured uses an voltage offset of ~1.6 volts that also sees the spike at the input and thus also gets amplified. The offset is needed to get a full 0-3.2 volt swing for the Panda A/D. After amoplification there are full scale voltage spikes that can present problems in analyzing the digitized signal (see pic).

I’ve tried adding 320 uf of filtering on the 5 volt rail, but with little effect. I can add another regulator to power the amp off the 12 volt supply and bypass the Panda power circuits, but don’t want to do that unless absolutely necessary. I suppose I could also try another amplifier circuit or forgo the offset by tying to ground and use only half the signal, but not an elegant solution either.

I doubt there’s a simple solution, but I thought this might be worth mentioning in case it helps someone else. On the other hand if someone has a idea I’m all ears (small Yoda joke).

I am not sure you can solve this.

I agree as it is really a problem with the amplifier and not the Panda. However, I may have tried the big capacitor in the wrong place. I rewired it to filter the 1.6 reference voltage that’s at the Op Amp input and the spike is much smaller - maybe tolerable. If it works well enough I’ll followup with a closing post.

By putting the filter on the offset voltage input at the first stage of the Op Amp instead of on the 5 volt supply, the spike was reduced to 100 mv at the output which should be tolerable. 220 uf was enough. It takes a few seconds to achieve offset at startup, but other than that I can’t think of any downside. Thanks for getting me thinking, Gus.

Gus has amazing powers. :wink:

I didn’t give any good answer so I didn’t really help much :slight_smile: But I am glad you got it to work.

Actually Gus, your answer did help. It made me recognize that as long as that small glitch was on the Op Amp input, I was sunk. Realizing that,m I refocused on the input and offset voltage rather than the 5 volt supply. Your answer was just what was needed. Thanks.