Power Supply with Active Rectifier, RF Filter and Super-Regulator

Is OPA828 soldered in orientation as shown at post #103?
Is there some 0.6V across R3?
OPA828 orientation is correct. There is 0.2V on R3, so B-E is closed.
 

Attachments

  • IMG_8900.jpeg
    IMG_8900.jpeg
    810.4 KB · Views: 124
  • IMG_8901.jpeg
    IMG_8901.jpeg
    807.5 KB · Views: 132
I would touch up on the soldering there, with flux added, it looks very sloppy and can easily be cause of your issues. Several pcb/trace damage but i'd not account it to the issue at hand. HF filter seems to have melted a bit, but not enough to eat trough as what can be seen.
 
There is 0.2V on R3, so B-E is closed.
Soldering doesn’t look very nice and there could be a problem at some point. But, we have to see first why there isn’t expected voltage at the base of Q7.

Two biasing diodes (D4 + D5) should provide about -1.27V at the Q6 base, compared to the input rail voltage. Check D4 & D5 diodes. If green LED is lit, then biasing current is flowing but one 1N4148 diode may be burned or in short circuit.
 
Last edited:
Soldering doesn’t look very nice and there could be a problem at some point. But, we have to see first why there isn’t expected voltage at the base of Q7.

Two biasing diodes (D4 + D5) should provide about -1.27V at the Q6 base, compared to the input rail voltage. Check D4 & D5 diodes. If green LED is lit, then biasing current is flowing but one 1N4148 diode may be burned or in short circuit.
There is only 0.4 V voltage drop on each diode (D4,D5) Q6 B-E is only 0.59V
 
OK. That raises a question if green LED is really lit or do you see some light reflection?
0.4V over each diode, means that only very small current (uA range) is flowing.

Possible cause is wrong orientation of the green LED. Check post #103.
 
Measurements are weird. CCS diode D6 would pull close to 1.5 mA through D4 & D5 and very small current through BE junctions of Q6 & Q7. In that case, there should be > 0.6V over each D4/D5 diode. O.8 V is impossible if diodes are OK.

If one diode would be burned to short circuit, then it could be close to 0.8 V over one diode.

If one diode is burned to open circuit, then green LED would be still lit, as required current would be pulled through BE junctions of Q6/Q7. However, that would result with high output currents at Q6/Q7 collectors (>2V over R3/R5), but regulator would power up and work.

Did you actually measure each diode voltage for D4/D5 or just divided voltage at the Q6 base?
 
Measurements are weird. CCS diode D6 would pull close to 1.5 mA through D4 & D5 and very small current through BE junctions of Q6 & Q7. In that case, there should be > 0.6V over each D4/D5 diode. O.8 V is impossible if diodes are OK.

If one diode would be burned to short circuit, then it could be close to 0.8 V over one diode.

If one diode is burned to open circuit, then green LED would be still lit, as required current would be pulled through BE junctions of Q6/Q7. However, that would result with high output currents at Q6/Q7 collectors (>2V over R3/R5), but regulator would power up and work.

Did you actually measure each diode voltage for D4/D5 or just divided voltage at the Q6 base?
I measured again 0.4 V on each. I measured 0.8 V from V+ to Q6 B. I can replace them.
 
Dzoli, besides improving skills (use a magnifying glass/lamp combination) please learn to first cut lead wires and then solder them. That way you will not put mechanical force on solder joints and the cut surface will also be soldered.

Get used to it and get reliability as reward.
 
Last edited:
  • Like
Reactions: dzoli
@dzoli
measurements point to a problematic component in the D4, D5, D6, D7 string. Which one is hard to say as measured voltages are in contradiction with fact that green LED is lit. Anyway, replacing those parts one by one will reveal which one was the culprit.

What is voltage at green LED terminals? It should be about 1.9 V.