REL Stadium III plate amplifier filter cap orientation

[matski]

Afternoon all,

I am not very well at the moment but am trying to keep myself busy. I’ve got a lovely REL Stadium III with low output, buzzing etc. Decided to pull the (mint) plate amplifier but did not note the polarity of the main caps before I unbolted them. The caps are marked clearly. Having been caught out in the past, can someone clarify if both negative poles couple in the centre? This would be my guess.

Kind regards

Mat

 

[adason]

Should be easy when you follow what comes out of diode bridge.
Upper left cut corner on bridge is positive, just trace it to positive of one cap.
Opposite bottom right corner is negative, goes to negative of other cap.

 

[rayma]

If you're not sure, you should post the other side of the board, where the traces are.

 

[matski]

Many thanks for your replies, I entirely forgot to add the underside. Limited mobility doesn’t help.

Kind regards

Mat

 

[AcoustatAnswerMan]

Odd for a manufacturer of a silkscreened PCB not to take advantage of said silkscreen and mark the intended polarities of the large capacitors. That's just asking for trouble. Assuming this amplifier has a biploar power supply (both + and - supplies), then one cap would have its negative at the common center point, and the other cap would have its positive at the common center point. Determining which is which can only be done by following the traces from the bridge rectifier, as outlined above.

 

[matski]

I agree it’s odd, and why I didn’t bother to mark as I assumed they’d be there. What you describe, combined with the previous rectifier note, makes sense. Interestingly, the pcb is labelled 300w MOSFET amplifier so is the same as used in the later Stentor, but a smaller toroidal is used. Attached is the transformer and my interpretation of the polarity.
Kind regards

Mat

 

[bullittstang]

I would place caps in same orientation you have notated

 

[matski]

I’ve reassembled the amplifier. All resistors, diodes, and caps (including the main ones) test well. During re-flowing all of the joints, I found this. Not sure how REL used to charge so much only to have a QC process so sloppy. At the moment I’m running the sub from a 500w class D external unit and it works exceptionally well, and I may leave it this way, but also have too many units!