Strange things afoot in this version of a Silvertone 1474

[SoulFetish]

Two things in particular, but let me back up. A buddy of mine was having issues with his Silvertone 1474, so I had him bring it by so we could bench it. He said there was really low output and signal seemed to "fade".
Anyways, step one: plug in the amp and see if it exhibits any of the problems he is describing and see if I recognize any of them.... is totally what I should have done first, but, I forgot to.
So, I went right into step two and started checking voltages against a couple of schematics we were working from.
It was quickly evident that this version had either been modded or was a different version than any I had seen in the schematics I had available. I wanted to run a few things by you guys, I'm betting you can clear a couple of things up.
First thing is that there is a back bias supply, which seems to move the HT centertap and voltage reference for the AC heater supply to around -27V. I've never seen anyone tie the heaters to a -V before, but that may have just been for wiring convenience. From looking at it, the main purpose seems to be for dropping the B+ to around 300V on the plates. Back biasing isn't all that strange really, I just don't see it all that much.
But I've never seen this grid bias arrangement before. The 6L6 cathodes are tied to a 39Ω resistor in series with two 12V, 12AX7 heaters in series to achieve a grid bias of ≈-27V
...the hell is THAT jive? Now that one is a first for me.

[SoulFetish]

hmmmm... so I don't think there is supposed to be any back biasing in this amp. I think it was wired incorrectly by the person working on this previously.
One of the things I'm concerned about is the series voltage across the 12AX7s fluctuating with signal drive. Anyone try and put any regulation (ie. zener diode, etc.) for the heater string portion of the 6L6 cathode circuit?

[Enzo]

Maybe it is just me, but what is "back bias"?

And can you post or link a schematic, even if it is only close.


Power tube cathode voltage is an average. The signal voltage jumps around, but it averages around the rest position.

[g1]

Here's an article on back bias:
http://www.aikenamps.com/index.php/what-is-back-biasing

[Enzo]

I'll be darned, learn something every day.

[SoulFetish]

Maybe it is just me, but what is "back bias"?

Traditional back biasing is a way of achieving a negative voltage supply by adding a series resistor between a secondary center tap and ground. The pulse current through resistor causes a voltage drop across the resistor and charges a parallel capacitor creating a negative voltage, which can be used as a negative supply or as a way of dropping the B+ (similar in how a zener diode would).
http://aikenamps.com/index.php/what-is-back-biasing

And can you post or link a schematic, even if it is only close.

you bet

[SoulFetish]

Here's an article on back bias:
http://www.aikenamps.com/index.php/what-is-back-biasing

damn, you're quick

[g1]

I was wondering why the circuit description rang a bell, it's come up before. And some of it was about using the cathode voltage for the heaters: http://music-electronics-forum.com/s...ad.php?t=42818


Schematic:

[SoulFetish]

the difference in the amp on my bench is instead of R53 220Ω resistor connected in parallel with the 12AX7 series heaters, it is connected as follows:

[g1]

Well, that would certainly account for the sad face on the rectifier tube.

[SoulFetish]

Well, that would certainly account for the sad face on the rectifier tube.

Well that, and dealing with electrons all day. They're always so negative

[g1]

The 6L6 cathodes are tied to a 39Ω resistor in series with two 12V, 12AX7 heaters in series to achieve a grid bias of ≈-27V

I'm still a bit confused. You have -27V at the grids from the back bias, but you also have around +24V at the cathodes supplying the 12AX heaters in series? So grid to cathode voltage (actual bias) is around -50V ?
Or are there 2 separate things going on, and just by coincidence they are both 27V ?

[SoulFetish]

I'm still a bit confused. You have -27V at the grids from the back bias, but you also have around +24V at the cathodes supplying the 12AX heaters in series? So grid to cathode voltage (actual bias) is around -50V ?
Or are there 2 separate things going on, and just by coincidence they are both 27V ?

No and Yes. The grids are still referenced to 0V, and the 6L6 cathodes are still +24(ish)V.
I think whoever worked on it once upon a time screwed up. Instead of the 12AX7 Heaters being wired in parallel with the 220Ω resistor, they connected the 220Ω resistor/cap between ground and the CT. Know what I'm saying? The 12AX7 heaters are the only thing connecting the 6L6 cathodes circuit. Weird, right?
I think it was a wiring mistake which resulted in a working back bias circuit. The the only thing connected to the -27V are the two 68Ω ballast resistors in the AC heater supply.

[SoulFetish]

Actually, thinking about it now, it could have easily been a wiring mistake. The 39Ω resistor which connects the 12AX7 heaters to the 6L6 cathodes is soldered to the 2 outside terminals of a 3 terminal strip (the center terminal is also an non isolated mounting terminal connected to the chassis). The 220Ω resistor/capacitor is soldered to the center terminal (ground) and then connected to the junction of the B+ centertap/68Ω resistors, which are connected to an unused pin terminal on the rectifier socket. Make sense?
I'll take a picture tomorrow if my description is confusing.

[SoulFetish]

G1, here is a picture of how it was wired:


so I rewired it to the correct cathode arrangement:


Which would have been fine, until I began to look around the chassis some more. When, the caps were replaced, they were all replaced with 450V Electrolytics. When we tested the unloaded voltages last time, the HT was around 425. But, after fixing the back bias, we were going to gain at least 27V on the B+, so those caps will need to be replaced.
I was getting a closer look to see how the connections were made, seeing what kind of space I had available for the replacements, and referencing the schematic, I saw that there were only two wires coming off the cathode of the Rectifier Tube. One of them was the 5V heater tap and the other was one of the choke leads. They had completely removed the 16µF reservoir capacitor, essentially turning it into a choke input supply with a choke terribly suited for it! So, I had to rip up the work I just did to make room for a reservoir capacitor.
Now, it seems like too much of a coincidence to make two mistakes rewiring the amp which results in two of the most common methods for dropping the B+.
I could guess all day what was going on. Either of these thing changes would be fine if executed properly. But, I'm glad he brought the amp over. I don't like the way the mains wiring was done on this at all (notice that someone replaced the polarity cap and wired a mallory pvc type capacitor from line to earth!!). It's going to be completely redone:


By the way, this is why we call it the "death cap". This is one of the things that happens when they fail