Keantoken's CFP cap multiplier

Here is another odd thing. The total voltages of the positive and neg rail sum to the proper total amount. It is as if their is an imbalance and ground point has shifted towards the positive rail more than the negative rail. I do not create a star ground point until the main board. That is where they are tied together. Is there something that could be causing an imbalance and shift theis mid point to favor one side over another.
 
Q1 was replaced and offset improved slightly. I have dual secondaries running through dual graetz, and each rail is floating until it is joined at the main board. It is a dual mono setup. Both multipliers have equal input voltages. The one channel is equal post multiplier. The other shows an imbalance.
 
Pic
 

Attachments

  • image.jpg
    image.jpg
    679.3 KB · Views: 335
I will measure tonight. What are you interested in. Oddly enough, I changed the BC337 and the rails got a little closer. I am swondering if there are transistor parameters that are saffecting the outcome. The small BJt's came on a tape and I have always found that these tend to measure failry close when traced, but perhaps not. Is it reasonable to think a significant parameter difference could cause the change. My rails now only show a difference of about 1V, but still show an imbalance. I guess i could replace the diodes and see if that resolves the remaining difference. I am more interested in knowing what went wrong or what changed?
 
The ultimate test of whether it's working is whether the ripple is below 1mV. I'm not sure either one is working correctly. Oscillation seems like a potential culprit here, and if so, then it could be counterfeit transistors (where did you get them?) or perhaps the long trace from the first lytic to the KM input. I haven't had that problem yet though.