curious defect

[albert_emule]

curious defect

http://s186.photobucket.com/albums/x...SMPSIR2110.jpg

good night colleagues.
this fountain above, already it is mounted and when it is working only what it-42vcc falls for zero when it places a resistor of thread of 10r x 50W
the tension 42vcc, it does not fall,
two joined intentions (-42 42vcc = 84vcc) does not fall, I placed the same resistor above described.
I already withdrew the couples of referring diodes of exit to-42vcc for verrifica-los and the same completed estam.
I exchanged the position of the couples of diodes to see if the defect changes of position, but the defect continued at the same place.
the uníca praiseworthy explanation what I found, was the fact of the feedback of regulation of being withdrawn only of 42vcc.
I imagine that placing the resistor in the point-42vcc, this one consumes load making to tension falling more and more.

[pacificblue]

Originally posted by albert_emule
the uníca praiseworthy explanation what I found, was the fact of the feedback of regulation of being withdrawn only of 42vcc.

Exactly. The controller has no feedback from the negative rail, so how could it react to a voltage sag there? This power supply will work best, when both rails are equally loaded.

[luka]

Originally posted by pacificblue

Exactly. The controller has no feedback from the negative rail, so how could it react to a voltage sag there? This power supply will work best, when both rails are equally loaded.

X2!!

[star882]

The output section looks a lot like that of the power supply for some old computer speakers I had. The supply even says something like "+ and - outputs must be loaded simultaneously".

So it's a design limitation, not a fault. Use two independent switchers for independent regulation.

[Eva]

A well coupled output inductor should provide acceptable cross-regulation.