What Does This Resistor Do? + 62 Year Old Mistake

[rp]

Still playing around w/ my 5C8. I was googling 5D8 and this image showed up of another late 5C8 with the paralleled 6SC7 instead of the 6J5 in the 2nd gain stage. This one has an odd 56K resistor sitting in the PI socket, looks plate to cathode or plate to plate, but that can't be. Any idea what's up? Maybe it's some later mod but it's a 10% silver band resistor so it might be original and the solder looks untouched. Maybe it just fell in there and isn't attached to anything

The mistake is the V3 cathode resistor which is 1500 like the 6J5 and not ~2500 like it should be for a 6SC7 and as it is in the pict of a different transition 5C8. 1500 will give it even more gain, right? Having messed a lot with this circuit that paralleled 6SC7 is already a lot of raunch for 1954, weird they'd try and up the gain so much. This thing was intended to have a guitar or two plus a mic and bass all plugged in!

[tictac]

Leo was known to always be tweaking amps here and there. I've seen a few examples of stock Fender amp with parts values other than what shows up on schematics. (A Blonde Showman with 82k plate loads for instance)....

You could be witnessing a Leo tweak....

TT

[martin manning]

This one has an odd 56K resistor sitting in the PI socket, looks plate to cathode or plate to plate, but that can't be. Any idea what's up? Maybe it's some later mod but it's a 10% silver band resistor so it might be original and the solder looks untouched. Maybe it just fell in there and isn't attached to anything.

Looks like the 56k goes from pin 2 to pin 4, the plate of the non-inverting triode to the grid of the inverting (input) triode. That would be positive feedback?!

[martin manning]

The mistake is the V3 cathode resistor which is 1500 like the 6J5 and not ~2500 like it should be for a 6SC7 and as it is in the pict of a different transition 5C8. 1500 will give it even more gain, right? Having messed a lot with this circuit that paralleled 6SC7 is already a lot of raunch for 1954, weird they'd try and up the gain so much. This thing was intended to have a guitar or two plus a mic and bass all plugged in!

Parallel 6SC7 triodes with 100k plate and 1k5 cathode resistors actually looks pretty good from a bias point of view (more centered than the 6J7), but it has 70 vs. 20 (3.5x) mu. The ra (paralleled) is 26k vs. the 6J7's 7k, so in the end you'd get about 3x the gain. Gain (bypassed cathode) = mu*Ra/(Ra+ra).

[Structo]

Here are a couple schematics that are clearer than the old hand drawn ones.

[rp]

Looks like the 56k goes from pin 2 to pin 4, the plate of the non-inverting triode to the grid of the inverting (input) triode. That would be positive feedback?!

Maybe that resistor does wonderful sonic things but I think I'll pretended it accidentally just fell into the tube socket. One day when I'm really bored...

Parallel 6SC7 triodes with 100k plate and 1k5 cathode resistors actually looks pretty good from a bias point of view (more centered than the 6J7), but it has 70 vs. 20 (3.5x) mu. The ra (paralleled) is 26k vs. the 6J7's 7k, so in the end you'd get about 3x the gain. Gain (bypassed cathode) = mu*Ra/(Ra+ra).

Did they parallel this 6SC7 primarily for the lower output impedance to drive the tone stack better? Overall gain is moot because of the 100K (vs 220K) correct? From what I read you get lower internal noise too paralleling triodes though for a 6SC7 that's probably a lost cause.

I tried a 4700R across the 2200R in there just to see/hear. Not a big sonic difference but actually sounded a bit clearer seeing ~1.5K at the cathode. If these are my measurements what is more text book from looking at the curves?

2.2K: 175V / 2.3V versus 1.5K: 150V / 1.8V

One other question, why does V1 (in my amp grid leak biased like the schematic) only have 65V at the plate (correct and the same as indicated in the layout) when it is fed off the same node as V2 and V3? Is the low plate voltage a peculiarity of grid leak biasing?

Thanks Tom, I keep forgetting someone cleaned up all the Fender schematics, I have all the old ones downloaded.

[martin manning]

Did they parallel this 6SC7 primarily for the lower output impedance to drive the tone stack better?

The output impedance of the 6J7 will be lower due to the 7k7 ra, so I don't think that would be.

Overall gain is moot because of the 100K (vs 220K) correct?

I think we are comparing a 6J5 (mu=20, ra=7k7) with a 100k load vs. a paralleled 6SC7 (mu=70, ra=56k single, 28k parallel) with a 100k load, right? If you calculate gain per the equation above it's dominated by the mu, and the result is it's 3x higher for the paralleled 6SC7.

I tried a 4700R across the 2200R in there just to see/hear. Not a big sonic difference but actually sounded a bit clearer seeing ~1.5K at the cathode. If these are my measurements what is more text book from looking at the curves?
2.2K: 175V / 2.3V versus 1.5K: 150V / 1.8V

Do you have a 6J5 in there now, or a paralleled 6SC7?

One other question, why does V1 (in my amp grid leak biased like the schematic) only have 65V at the plate (correct and the same as indicated in the layout) when it is fed off the same node as V2 and V3? Is the low plate voltage a peculiarity of grid leak biasing?

I don't have any 6SC7's, but grid leak biasing is pretty hot based on my experiments with a 12AX7. That would certainly drag the plate voltage down.

[rp]

rp wrote:
I tried a 4700R across the 2200R in there just to see/hear. Not a big sonic difference but actually sounded a bit clearer seeing ~1.5K at the cathode. If these are my measurements what is more text book from looking at the curves?
2.2K: 175V / 2.3V versus 1.5K: 150V / 1.8V

Do you have a 6J5 in there now, or a paralleled 6SC7?

6SC7. And good morning. I just had lunch, pasta again

[martin manning]

rp wrote:
I tried a 4700R across the 2200R in there just to see/hear. Not a big sonic difference but actually sounded a bit clearer seeing ~1.5K at the cathode. If these are my measurements what is more text book from looking at the curves?
2.2K: 175V / 2.3V versus 1.5K: 150V / 1.8V

Do you have a 6J5 in there now, or a paralleled 6SC7?

6SC7. And good morning. I just had lunch, pasta again.

I'm still sipping coffee, but pasta with clam sauce is a favorite in my house!

Here is the load line with the two bias points. Very close to your measurements, and I just used the layout's 270V. Not a huge difference, and consistent with your testing. Channeling Leo, I think using a single tube type across the whole preamp is a plus.

[rp]

Martin, right now I'd happily trade a plate of pasta, even my 95y/o aunt's fettuccine, for an enchilada w/ mole. But don't tell my aunt.

I'm sort of understanding how tube curves are drawn on but I still have no idea what makes one operating point "better" than an other. The only thing that seems bad is something that plots out at the extremes. I don't understand how a tube that should have 150-300V on it can function just fine at 65V! I guess that's why tubes work so well for so many sophomore builders.

Looking at your plots it looks like 1/2 of 1/2 half of the other?

Channeling Leo, I think using a single tube type across the whole preamp is a plus.

Never occurred to me that was intentional, but in MI amps you very much see, for example, 12AX7s straight across. But it did occur to me that in hifi preamps and amps you see a more intersting mix up of tubes, especially in the old days when there was infinite availability. Hifi amps full of 12AX7 are a post '80s phenomena. Or maybe MI amp builders are/were just much less iconoclastic than the hifi guys.

[rp]

I went with my ears and did a goofy A/B switched set-up. You can hear the difference, 1.5K adds a touch more gain and hence sounds a bit meatier, that's always good so I went with that. Looks like it wasn't a 62y/o mistake, just an option. So for future 5C8 reference, V3 6SC7 paralleled 100K shared/1.5K seems optimal.

[martin manning]

For a small difference ears are the best test gear. The load lines give you an idea of how the stage is going to clip, but the proof is in the listening. I would say that the 1k5 Rk point is better centered considering the AC load line will be steeper. There are more grid curves above 0V, and dynamic shifting of the bias point too. Gain is not the right word really, as the Rk has little to do with that. It's about more or less of different kinds of harmonic distortion.

[Firestorm]

I had a 5D8 built into a 5C8 chassis. Made right before Summer NAMM when Fender was transitioning to the new circuit but still using old chassis. With no hole for the presence it wasn't quite 5D8, but it used the 5D8 circuit card. Also 9-pin tubes on adapters.

But they did something out of the ordinary (to compensate for no presence, maybe) and it made the amp ungodly loud and harmonically rich.

Now I just have remember what it was...

[martin manning]

I had a 5D8 built into a 5C8 chassis. Made right before Summer NAMM when Fender was transitioning to the new circuit but still using old chassis. With no hole for the presence it wasn't quite 5D8, but it used the 5D8 circuit card. Also 9-pin tubes on adapters.

But they did something out of the ordinary (to compensate for no presence, maybe) and it made the amp ungodly loud and harmonically rich.

Now I just have remember what it was...

5C8 has NFB routed to one power tube grid directly, and in a roundabout way to the other. 5D8's presence runs NFB to the PI's input cathode, a step closer to modern technique. That could certainly have been connected without the presence control. Perhaps that's where the magic is?

[Firestorm]

I had a 5D8 built into a 5C8 chassis. Made right before Summer NAMM when Fender was transitioning to the new circuit but still using old chassis. With no hole for the presence it wasn't quite 5D8, but it used the 5D8 circuit card. Also 9-pin tubes on adapters.

But they did something out of the ordinary (to compensate for no presence, maybe) and it made the amp ungodly loud and harmonically rich.

Now I just have remember what it was...

5C8 has NFB routed to one power tube grid directly, and in a roundabout way to the other. 5D8's presence runs NFB to the PI's input cathode, a step closer to modern technique. That could certainly have been connected without the presence control. Perhaps that's where the magic is?

That's what I would have thought, but this one had a crazy value resistor somewhere. And it really made it crazy loud. This was a strange era for Fender: the were behind with the Strat introduction and still screwing with the Bassman, which was the favorite child. And then there was Leo.

I might have notes.