[BradJudy]

My PPA has been in storage for a while and I pulled it out to use. The epsilon 12 in it was oscillating on the relay, switching it on and off. I remembered this being an adjustment issue, so I popped it open and started working the trimpot. Unfortunately, the only stable state I can get it into is a permanently open relay. I went back through the setup doc and walked through the steps again, checking my voltages. When I get it adjusted so that U1 pin 6 and the amp output ground voltage is about zero, it starts to oscillate again. I can tweak VR1 until the relay is properly closed, but then after I power it off and back on, the relay stays open again.

Any ideas?

For reference, this is on a PPA v1 with Glassman buffers (checked DC offset on outputs and they are good) and a STEPS power supply at 24VDC. The pin 4/7 voltage measurement is 21.6VDC for both U1 and U2. The U1 6/7 measurement is 10.8VDC. One note is that U1 was not seated properly when I opened it to start tweaking, so I seated it. I don't know if another part could have been damaged by the fact that U1 wasn't seated properly at first.

 

[BradJudy]

Bump - any ideas?

 

[MoodySteve]

First, if you have some spare ICs, swap them in.

Does it start to oscillate when you're *really* close to zero? You can get away with a few mV of offset.

If it starts oscillating when you hit 0mV offset, I'm guessing it oscillates with an amplitude around 30mV? If so, one possible workaround is to set the offset point at the midpoint of the oscillation...in this case, +15mV or -15mV.

If that doesn't work, you can lower the sensitivity of the circuit by reducing the value of R9 (I believe). If you do this, be sure to increase the value of C5 accordingly to keep the corner frequency as close to DC as possible.

 

[BradJudy]

Thanks for the tips. Since I don't have extra IC's, I just fiddled with the trimpot for another half hour tonight. I was finally able to get it to settle down (at about -19mV, in line with your suggestion) and we'll see how well it works. I don't remember it being nearly this finicky when I first set it up.

 

[dBel84]

I have not had much luck with the e12. I have modded the OPA551 to see the amp ground and not the virtual ground. This issue was discussed quite extensively on the Headwize thread but mostly, this is not a problem..dB

 

[Pars]

I think you mean modding the comparator opamp (TL082?) to see the amp ground on pin 3, rather than the virtual ground. If you watch the virtual ground between relay states, you will see that it moves ~20mV or so (from what I recall). On mine, I cut pin 3 away from the virtual ground on the e12 board and ran an amp ground to it.

 

[dBel84]

yes Pars, you are quite right, the TL08x . The virtual ground was left intact (wouldn't be work at all otherwise

) ..dB