Need advice on re-capping a Scott 272 integrated amp

[drew]

I picked up an HH Scott model 272 stereo integrated amp from the early 60's.  4 x EL34 output tubes, dual 5U4GB rectifier tubes.  Bias voltage and preamp heaters come from a separate winding on the PT, rectified by a selenium bridge which I have replaced with a silicon unit.  The amp was inoperable when I got it, with a blown fuse and several bad tubes.

I tried to reform the power supply electrolytics, hoping to at least be able to power the amp up and see if it produces sound, but they don't seem to want to cooperate.  The electrolytics are in multisection cans as follows:

C1: 4 x 20uf @ 475v
C2, C3: 2 x 20uf @ 450v, 2 x 25uf @ 25v
C4: 3 x 4uf @ 250v, 1 25uf @ 25v
C5, C6: 75v sections for the bias voltage.  My reforming efforts did not include them.  Ignore for now.  Supposedly they're almost always OK.

The 25/25 sections can be replaced by individual separates at the appropriate tube sockets (although it's going to be like trying to build a ship in a bottle that already has a fleet of ships in it; see gut shots).  Also somewhat hopeful that C4 will turn out to be OK once C1-3 are dealt with.

Question: should I up the capacitance of any/all of the 8 20uf sections?  If so, which, how much, should any other changes be made, etc.  If I'm reading the attached schem correctly, there are initially 3 sections coming off the rectifiers (C1A, C1B, C1C [latter two drawn at the OT center taps], then a series of resistor/capacitor combinations.  This amp does not have chokes.

[drew]

Guts

[eleventeen]

Kind of interesting that the 5U4 filaments are in series. I would stay with existing values; I am not sure you can get 4 * 40 ufd (which would be the next higher can-based value) in one single can @ 475 volts. If you start raising the cap ufds, it will be harder and harder to find those values in a single can, and then that starts to force you to use single external ones. Which is OK, as far as I am concerned. I am not a stickler for pure originality, others are.


You're definitely going to have a real estate challenge. I myself am fond of using a few radial-mount (eg; PC-mount) single caps with a dab of silicone to the chassis in these situations...one, because they are cheap. Two, because they are small. You could have a > $100 cap bill replacing all of those electrolytics. 

[drew]

Question: should I up the capacitance of any/all of the 8 20uf sections?  If so, which, how much, should any other changes be made, etc.  If I'm reading the attached schem correctly, there are initially 3 sections coming off the rectifiers (C1A, C1B, C1C [latter two drawn at the OT center taps], then a series of resistor/capacitor combinations.  This amp does not have chokes.

Comparing my amp to the schematic, the C1B and C1C sections actually attach to the OT side of the 80ohm 10 watt resistors, not the rectifier side.  So, either the schematic (which is a third party product, not an official Scott document) is drawn wrong (see attachment 1), or my amp is wired differently.  But the way my amp actually is (see attachment 2), actually makes more sense, doesn't it?  Just one 20uf section straight off the rectifier tubes, then it splits into various R/C circuits. 

Questions:

1.  The first cap connected to a 5U4GB is supposed to be no bigger than 40uf, I believe.  When an amp uses 2 5U4GBs in series, what's the rule?  Double? Half? Something involving pi, differential equations, and/or using both fingers and toes to count?

2.  Since the 20uf sections connected to the output transformers' center taps are actually separated from the initial 20uf section with 80 ohm resistors, can I make them substantially bigger without worrying about blowing up the rectifier tubes?  CE makes a 80/40/30/20 @ 525v cap can; if I got two of them, and connected the 80uf sections to the OTs, would the amp be happy?

[eleventeen]

1.  The first cap connected to a 5U4GB is supposed to be no bigger than 40uf, I believe.  When an amp uses 2 5U4GBs in series, what's the rule?  Double? Half? Something involving pi, differential equations, and/or using both fingers and toes to count?


Actually, those 5U4 are is *parallel*, not series. The idea of having a "not too big" cap as the very first cap the rectifier "sees" (the one connected directly to the 5U4 heater) is simply to prevent inrush current charging a dead-discharged cap from possibly causing the 5U4 (or other rect tube) from arcing over. If you get your several-value "4 caps in a can" as you mentioned, put the small value (20 ufd) closest to the 5U4 with the 80 ohm resistors separating the other sections and you can probably stop thinking about it.




2.  Since the 20uf sections connected to the output transformers' center taps are actually separated from the initial 20uf section with 80 ohm resistors, can I make them substantially bigger without worrying about blowing up the rectifier tubes?  CE makes a 80/40/30/20 @ 525v cap can; if I got two of them, and connected the 80uf sections to the OTs, would the amp be happy?


Pretty much. The purpose of the exercise is what I explained above. When fully discharged, those big electrolytics appear as a near-short ckt to the rectifier tube which, yes, is supplying "DC" but at the heater is highly unfiltered DC aka "pulsating DC" which is more like AC.... which we know that caps will pass....hence will look like a short when the thing is first turned on. 

[drew]

Thanks for the  explanation.