New Build - SilverClone

[Hazcat]

I have started on a build that is using parts from a 1937 Silvertone console radio that was used occasionally and had good volume & tone when it was working but the cabinet was ruined during storage. It has a 6V6 output into an electrodynamic speaker, and it looks like a fun challenge. I'm building it into a cabinet & chassis from a noisy Fender Sidekick 35 (approx Princeton-sized).
FYI re my skill level, I was a E-tech (aka TV Repairman ) back in the days when electronics weren't disposable ('72-'82), but I really didn't get into the design part of things, just troubleshooting and repair of radio, TV and my and other folks' guitar amps. But in my retirement, I'm having some fun taking up amp building, and learning a lot from this forum.
That said, For this project, I'm using the radio's detector/AF amp 6SQ7 for the first preamp (similar to 12AX7) and the oscillator 6J5 as the driver (like 12AU7?). For the output, I'm planning on using the common Champ/GA-5, etc -type output - the Silvertone originally had a 250 cathode resistor and no bypass cap.

I had it on my to-do list to build a GA-5 amp as well, and I'm wondering whether it would be reasonably easy to build this into a two-channel amp, one channel GA-5 and the other Octal. I'm attaching my first cut at a schematic - the component numbering is from the GA5 schematic with "A" on the matching octal components. Does it look feasible? Here are a few questions that have occurred to me:

1. Output tube grid matching/mixing(?) resistors for the two channels - I was going to start with 220K like the Fender AB763 channels, but I'm wondering whether the driver plate resistance needs to be taken in consideration for showing the 6V6 a balanced input. The coupling capacitors are both .022.

2. RE B+ filtering, I've never worked with field coils, but I'm treating it like a choke, with a filter cap on each side. Any old timers know whether the field coil needs that much smoothing? The Silvertone used a 25uf off the rectifier (before the field coil) and a 16uf after on the screen -but with the 5Y3, I thought a 10uf first filter would be easier on the tube. I wouldn't think the flux was responsive to 60Hz, but I've learned things about assuming . I'm planning to use a 10uf there, which is closer to the 6V6 spec. The GA-5 has a filter on the back of the 10K screen resistor, too, then another after the 22K preamp supply resistor. If I include those, am I going into overkill on the filtering? I'm not sure how the other capacitance on the B+ would load the supply. BTW, the Silvertone's power supply ran 7 tubes, so I think my filament and primary current should be ok. On the other hand, signal tubes don't draw a lot...

3. One last general question - Is there a rule of thumb for choosing the initial values for plate resistors when designing an amp? If I knew the plate current I could use ohm's law, just used the published value? In looking at tube characteristics and knowing how voltages changes with load, it looks like a moving target to me. A lot of older amps look like the manufacturers bought a boatload of the same value resistor, and used the same wherever they could, and tweaked cathode bypass values to get the B+ in range. I have a Silvertone 1333 that seems to use 390K for everything, Gibson seems to like 220K, and Fender 100K. I did a search on this forum, but came up empty. I see Uncle Doug has a series on Resistors in guitar amps, so I'll be checking those out as well.
Thanks!

[PRR]

> ...1937 Silvertone console radio .....I've never worked with field coils, but I'm treating it like a choke, with a filter cap on each side. Any old timers know whether the field coil needs that much smoothing?
> ....rule of thumb for choosing the initial values for plate resistors when designing an amp?

I would not assume a speaker almost as old as my dad works today. Have you tested it?

I would not assume a speaker suitable for one 6V6 in the parlor playing 1930s music (without compression; that came later) will stand-up to hard-clipped rock. Or even heavy blues, K-Pop, Peruvian Polka, whatever. I'm saying I would not assume the field coil will always be there for you. (OK, you can bust it off and nail it inside the case, put a JBL E-110 or whatever the killer kone is today in the grille.... but that's weight.)

But assuming you go that way, I would not assume you know better than Silvertone's designers.

I don't know what overkill is on filtering, especially single-ended.

The plate resistor can be 2X (to 5X) the load it must drive. Notes below. The rest of the values can be taken from the Resistance-Coupled Amplifier Tables--- it was good enough for Leo.

I suspect the NFB will fight (try to eliminate) the 6J5 channel. But it won't do nothing unless you ground the far side of the winding.

Can that image be any bigger?

[shooter]

Is there a rule of thumb for choosing the initial values for plate resistors when designing an amp?

I start with 40yrs of designer's values, take the average of the difference's then adjust as needed

[Hazcat]

Sorry about the image size, this is my first time posting and it came up as a usable size on my laptop. I'll resize it.  The speaker was reconed about 35 years ago, when the radio was serviced (the customer never picked it up, the tech was closing his shop and gave it to me).  I realize that it's not optimized for guitar music, but I've got it, and I'm going to try it for fun and experience. I've got other amps, too.  The PT is good, and if the speaker craps, I've got a SE output xformer and a 10" guitar speaker I can use. This is strictly a "Can I turn an old radio into a guitar amp?" project. My main concern about the filtering is whether it would stress the PT - I plead ignorance, but glad it's not an issue.  Not using the NFB on the 6J5 side, was going to leave it off altogether, but just copying the GA-5 on the first 12AX7 channel as a starting point.  Thanks for your input, I'll get to work on that schematic resize.
re your suggestion next to the 6J5 volume control, I'm assuming you mean a 1 meg pot and 320K grid stopper; correct?

[ac427v]

Hmmmm, I don't see a filter capacitor on node C of the power supply. A 20uf or 22uf at 450 volts would work nicely there. I like that node A has two capacitors separated by a coil. Good design in my experience!

[Hazcat]

Thanks, I'll put one there. I've been brushing up on tube amp design from Uncle Doug's channel, his series on resistors is perfect for my level, and I already answered some of my own questions!  I'm going to lay out my terminal board tonight and get started! Thanks for the encouragement!

[Stuff4bikes]

I am fairly new to this amp building hobby also but I have learned a few things....

If you are just going to leave the radio exactly as is and just plug a guitar into it
Then the field coil speaker will probably work for you. I have had several field coil
Speakers that sounded nice with the stock configuration but when I put the tone after
The first stage, grid leaked the input and by-passed the output tube(s) the speaker(s) just
Didn't handle the extra umph very well....

[Hazcat]

Hmmm, I'll have to look at the schematic - having a hard time imagining what you described!  I have the radio schematic, so I'll try the original output circuit if the modern 6V6 circuit is mush.  Just tinkering. It's likely I'll end up with a new OT & PM speaker when all is said and done, but I'm curious about the old gear. The radio didn't have a lot of high end response, but that wasn't desirable in the noisy AM environment. Don't know if the electrodynamic dynamic contributed to that, but I'll see!  The coil flux gets stronger when the power supply is drawing more current, so that might dampen things as well.
Thanks!