High voltage Super Reverb

[Tomislaw]

Hello,

I've got an interesting case of excess voltage in power supply on the bench right now - it's my pal's mistreated 1972 Super Reverb AA270. It's got a whole bunch of problems but the most interesting one is that it the voltage is way too high. When I first turned it on, the Chinese 6L6WGS were dissipating 26 watts with 540V on the plates! I had to modify the bias supply in the first place to get enough negative voltage to stop them from everheating, but they should probably be replaced with proper 30W 6L6GC's as well. Rectifier tube was GZ34 even though the tube chart calls for 5U4GB. I put one in and we are now at 500V which is probably still too much for those cheap tubes. A correct rectifer does help the voltage problem a bit, but there's 7,2V AC on heaters under full load and the voltage selector is already in the 240V position. I assume this gives away that the PT could have developed a small short in the primary winding and needs to be replaced. The owner is hesitant, though, as he's worried that the unique - as he say - sound of his amp could change.

Are silverface Fenders notorious for the voltage creep up problems? I know the network mains voltage is now higher than 50 years ago, but in case of this amp the difference is too big anyway, and it's not even running on the 230V selector setting either. It's obvious that voltages in old amps are usually higher than specified in the schematic, but I'm sure we're not talking about 540V on 6L6s' plates. 

On the side note, only the filter caps were replaced in this amp at some point in the past and probably the tech was aware of the higher voltage as they are now all doubled in series, not only the first one. No resistors across the converted ones, unfortunately. It's definitely the AA270 model judging by the bias supply arrangement and a few other details, but values of the resistors in the supply doesn't match the schematic which calls for 1K and 4,7K after the choke and I have 2,2K and 10K instead.
All other caps including filter caps in the bias supply are still original from 1972. I could observe the bias voltage go rapidly up or even disappear as I was lightly tapping the bias board - frightening, so I had to replace the associated caps immediately. The 12AT7 driver tube turns off and on by itself when it feels like it, which is always fun.

Cheers,
Tomi

[K O'Connor]

Hi Tomi

If all the voltages are high by the same factor then the amp should stay biased reasonably well, except for the detail that the bias filter caps may not be rated properly. However, since someone has been inside the amp previously and boosted all the plate caps, you would expect them to change the bias filters, too.

In any case, this is not normal. I'm not sure if a few shorted turns on the primary would do this (by changing the effective turns ratio from primary to all secondaries), or if such shorted turns would just blow the mains fuse? if this is the fault, then the former is the result and really the PT should be changed. If a heating event could cause this then the insulation on the coil wires and between the windings is suspect. The previous tech should have done this.

Series filter caps without sharing resistors is another disaster waiting to happen.

Bandaids such as a small PT to reduce the incoming voltage only put off the inevitable.

I had a similar seeming incident decades ago. An amp a customer brought over had all too-gihh voltages inside. It turned out my meter had a burnt resistor inside, which once replaced restored the universe to its previous balance.

Have fun

[makinrose]

I agree with Kevin here. It's definitely time change the power transformer. Hammond now makes an excellent replacement that will drop in.  I've seen the same thing on Silverface Vibrochamp that had really high plate voltages.  It's a weird way for a PT to fail but it does happen.

[Tomislaw]

The owner agreed to replace the transformer and today I installed a brand new Hammond 290DEX. We're back to healthy heater voltages (6,35V under load) now and with a GZ34 (which I had to substitute for the 5U4GB anyway) I'm getting 437V on the plates, which is the exact expected voltage according to the original schematic. There's 30 volts more on the driver node and further down the supply chain, but I'm going to leave it as it is. I installed 1K/4,7K dropper resistors as on the schematic, replacing the 2,2K/10K ones that I found in the amp so that's probably why I'm seeing more volts right now. 

There's some more restoration work left to do as it's definitely been a long time since this amp was thoroghly serviced for the last time, but it already sounds OK. I installed all new electrolytics and I could use single caps after the choke as the voltages are now within their specs. Interesting that the amp has the ground switch installed, but it was only used as soldering lugs for the power cord. The slider lug has no solder on it, so the death cap wasn't installed. I'm going to use it as a NFB loop selector. 

I think the original transformer could have been damaged by occasionally shorting indicator light connection in the 6,3V heater circuit. Someone has snipped off original artificial center tap resistors and soldered new ones leaving old legs sticking out very close to soldering lugs. There was also another weird problem as the amp wouldn't turn on every time the power switch was flipped on and I thought it will have to be replaced, but once I cleaned up the indicator wiring the amp will now turn on every time no problem. 

I was told the amp was on someone's bench 4 years ago and the voltages were fine then. 

Tomi

[makinrose]

The owner agreed to replace the transformer and today I installed a brand new Hammond 290DEX. We're back to healthy heater voltages (6,35V under load) now and with a GZ34 (which I had to substitute for the 5U4GB anyway) I'm getting 437V on the plates, which is the exact expected voltage according to the original schematic. There's 30 volts more on the driver node and further down the supply chain, but I'm going to leave it as it is. I installed 1K/4,7K dropper resistors as on the schematic, replacing the 2,2K/10K ones that I found in the amp so that's probably why I'm seeing more volts right now. 

There's some more restoration work left to do as it's definitely been a long time since this amp was thoroghly serviced for the last time, but it already sounds OK. I installed all new electrolytics and I could use single caps after the choke as the voltages are now within their specs. Interesting that the amp has the ground switch installed, but it was only used as soldering lugs for the power cord. The slider lug has no solder on it, so the death cap wasn't installed. I'm going to use it as a NFB loop selector. 

I think the original transformer could have been damaged by occasionally shorting indicator light connection in the 6,3V heater circuit. Someone has snipped off original artificial center tap resistors and soldered new ones leaving old legs sticking out very close to soldering lugs. There was also another weird problem as the amp wouldn't turn on every time the power switch was flipped on and I thought it will have to be replaced, but once I cleaned up the indicator wiring the amp will now turn on every time no problem. 

I was told the amp was on someone's bench 4 years ago and the voltages were fine then. 

Tomi

I'm glad you got up and running! The NFB switch is good idea.  When I do those I often use a rotary switch or a pot which gives you lot of possibilities but for some players who like it simple a switch works better.