Single transformer - multiple bridges!

[DigitalJunkie]

Originally Posted by UV101

What I have is this


And I'm wondering what will happen if I move the this


I am already dealing with multiple grounds so I've no issue there. Obviously I could got for a transformer with multiple secondary windings and keep that side as per the 1st diagram and the mains side as per the second, but I wonder how using an off the shelf large va transformer would effect things.

I treat each supply as being in its own noise domain but I wonder if going all the way back to the secondary windings is really necessary. I'm not just looking for an it will work or wont, I'm interested in opinions on the quality of the supply.....

I appreciate the comments so far, thanks ;-)

The first/top picture will work. Each secondary/bridge rectifier/cap is isolated from the others,except for the common ground.

The second/bottom pic will release magic smoke/blow fuses. Simplify it by looking at 2x bridges/caps at a time. If you trace the current flow,you will notice that one part, or the other,of the bridge rectifiers is shorted every cycle. It would work if the grounds were isolated,but they aren't.

[UV101]

Originally Posted by DigitalJunkie

The second/bottom pic will release magic smoke/blow fuses. Simplify it by looking at 2x bridges/caps at a time. If you trace the current flow,you will notice that one part, or the other,of the bridge rectifiers is shorted every cycle. It would work if the grounds were isolated,but they aren't.

Can you explain, I'm struggling to follow.

[DF96]

Grounding is the problem. You say you are dealing with multiple grounds. At present you have separate routes for charging pulses. In the new arrangement you have not, so you have lost control. With lots of diodes in parallel you have no control over where the currents go. You will be injecting pulses into your ground.

[Emerson Prado]

I can't follow either. In my perception, this is what happens:
1.When the transformer top lead is positive, all right-top diodes will be open, while right-bottom ones will all connect transformer bottom lead to ground. In the next half-cycle, the opposite happens. Summing up: all bridges will have the right top/bottom diodes alternatively paralelled/open.
2.When the transformer top lead is positive, all left-bottom diodes will be open, while left-top ones will all connect transformer top lead to their respective cap positive lead. In the next half-cycle, the opposite happens.
3.Between each cap + lead and ground there will be a load (besides the cap itself).
4.The charging current for each cap goes to the same node as right-side diodes anodes, so being distributed among these.
The only issue I can see is right-side diodes get paralelled, making their currents strongly unmatched (due to the different Vf's), possibly causing some to take too much. One could think the common ground would act like a short circuit, but I can hardly see this, since this doesn't connect to other supply nodes, so there's no current path (except for RF, but pointless here).
Did I miss something?
Best regards,
Emerson