OT upgrade recommendation for SF Deluxe Reverb

[Tone Meister]

I don't have any further plans for the amp past being able to run 6L6GC and 6V6 power tubes. I went completely through the circuit a couple of years ago, so it's being all it can be as a stock Deluxe Reverb.

[pdf64]

This is a situation where it may be better to set the static conditions to be within (or at least not too far over) what the PT can accommodate.
So I suggest seeing if a current draw more in line with what would be normal for 6V6 rather than 6L6 can provide a reasonable tone / sufficient freedom from crossover, eg 20-25mA.
The 6k6 load may facilitate colder operation whilst keeping the transfer function linearish.

[Tone Meister]

I used the scope to adjust bias and the notch disappeared right around 30 mA. The amp puts out right at 24 watts with the 6L6GC and GZ34 with the bias set that way. (13.91V into 8-ohm load) Line voltage = 120VAC

Filament voltage 6.3V

[Tone Meister]

10 gallons in a 5-gallon bucket

OK, here's what I ended up doing after studying the circuit and my options and coming up with some realistic expectations. Ended up installing the following components and here are the main details:

Allen TP40D power transformer (320-0-320, 5V @ 3A, 6.3V @ 4.5A, center tapped heater, internal hum shield brought out for grounding)
ClassicTone 40-18006 output transformer (40-watts, for Vibrolux, 4K primary to 4-8-16 ohm taps)
Ruby 6L6GCMSTR power tubes (matched for high headroom)
Yellow Jackets YJR solid state rectifier

Installed open J11 jacks
Installed 8-ohm ElectroVoice EVM12L speaker
Wired 8-ohm tap to main speaker jack (6L6 sees 4K)
NFB to 8-ohm jack
Wired 4-ohm tap to extension speaker jack (6V6 sees 8K)
Biased at 442V @ 37mA = 16.5 W = 55% dissipation

Painstakingly went through the amp to make sure all voltages were consistent with AB763 DR schematic. All voltages verified to within 5 percent, most are closer than that. Dropping resistors in dog house ended up at 10K/10K. (I had run the silverface values of 10K/2.2K previously)

The amp was previously producing 20 watts into an 8-ohm load with 6V6 and a 5U4 rectifier and 24 watts with 6L6GC and a GZ34.
The current setup produces 42 watts with a perfect sine wave on the scope. All power tests were done with exactly 120.0V AC line voltage from the variac.

I replaced a (very well loved) Tone Tubby 40/40 with the EVM speaker to ensure maximum headroom. Monster tone, exceptionally quiet amp, full and crisp with gobs of headroom. Ended up with TREBLE on 6.5 and the BASS on 4.5 and the tone is just what I was looking for. More headroom than a Pro or a Vibrolux, more reminiscent of a very manageable blackface Twin Reverb. Plenty loud if need be, but it sounds full, tight, and clear at "conversation volume" also. Looks like I was fortunate enough to hit a bullseye on the very first shot. I've "shut the hood" on it and have no plans to work on it again until it needs service or has a problem.

And there you have it: Ten gallons in a five-gallon bucket

[Tone Meister]

UPDATE: April 28, 2015

1) Changed the phase inverter plate load resistors from 100K/82K to 43K/43K, which removed all the low end flatulence. Silverface values for these resistors are often 47K/47K in some amps. I used 43K because that is what I had on hand. The combination of the BF phase inverter values (1megs, 470, .001uF cap) and the 43K plate loads really tightened up the bottom end.
2) Changed the 3.3M resistor in the reverb mix circuit to 1M, which lets in a bit more of the dry signal. Noticeable difference that I like.
3) Changed the speaker from the (very heavy) EVM12L to an Eminence Swamp Thang
4) Except for the changes above, the circuit is stock AB763 with upgraded transformers
5) Heated up power tube bias slightly to 17.5W which is about 58% dissipation

Played three shows after making the above changes, one with a 5-piece horn section, and the amp was all but perfect. Very full, thick and warm, with just the right amount of high end and ample headroom.

[pdf64]

Thanks for the update!
Alternative or additional to lowering the plate resistor value, consider reducing the power tube grid leaks, eg down to 100k.
They're effectively in parallel to the plate resistors, so have the same (blocking distortion mitigation) benefit of slightly reducing gain and max voltage swing, but also they reduce the charge / discharge time constant with the 0.1uF coupling caps, so reducing bias shift at overdrive.

[Tone Meister]

Are those the 220Ks after the PI coupling cap? Can you post a schematic with the correct resistors marked? That is my goal, to reduce/eliminate blocking distortion. The amp is really, really good now but I would like to try adjusting those grid leaks to see what effect it has. The amp is very touch sensitive now, much more so than the stock amp.

[g1]

Yes, those 220K's after the PI coupling caps.

[pdf64]

Yes, see Lots Of Mods To A BF Fender Photo by pdf64 | Photobucket I tweaked an old schematic to include my preferred mods.

[Tone Meister]

Thanks G1 and Pete. Really cool how you labeled the schematic to your own AA1113 circuit revision. I have to admit it threw me for a loop initially.

Q: Given the way my amp is currently configured, what should I expect from changing those grid leaks to 100K?
Q: Is it possible to lower the value of these resistors too far and what would be the negative effects, if any, of doing so?

[pdf64]

Pleased you appreciated it, I thought that the schematic designation may be a nice finishing touch!
Those old Fender schematics are so well crafted that other styles pale somewhat in comparison.

Given that you are happy with how your amp responds, you may not perceive much (if any) benefit from lowering the grid leaks. But if you notice a bit of blocking distortion creeping in at certain settings / scenarios, eg when it gets pushed into overdrive, then it may be worth experimenting.
Some SF models had 47k LTP plate resistors and 68k grid leaks on the power tubes. I suspect that if the grid leak value went too low, eg <47k, then the voltage swing at the LTP outputs would be constricted, and the power tubes wouldn't be able to be fully driven, so limiting the amp's power output (similar to a crossline master volume turned down a little).

[Tone Meister]

I have shows again all the way through the weekend and into next week, so I'll swap'em out and see what difference I can detect. It'll get driven pretty hard, so I'll certainly get a good sense of the amp's response. Thanks again for the info.

[Tone Meister]

Swapped the 220K grid leaks for a pair of 100Ks. My initial impression is that it may have neutered the bass response a little too much when combined with those 43K plate resistors. The bass control seems to have lost a great deal of effectiveness past about "3" on the dial; however, the only way I can truly assess overall performance will be at show volumes. The limited bass response may turn out to be a positive, depending on how the amp overdrives in action.

I didn't use the amp the past two nights, but I will use it for the next several shows, starting tonight, and will see how it performs in action. That's the only place it counts anyway. Look for another update by Tuesday or Wednesday.

[Tone Meister]

... Given that you are happy with how your amp responds, you may not perceive much (if any) benefit from lowering the grid leaks. But if you notice a bit of blocking distortion creeping in at certain settings / scenarios, eg when it gets pushed into overdrive, then it may be worth experimenting ...

As predicted, I didn't detect any audible difference with the lowered grid leak resistors when combined with the 43K plate loads. At show volume, the bass response was ample although the BASS control is all but ineffective past "3" on the dial. Overall feel (touch response) is the same and the amp overdrives nicely without the flatulence associated with many stock Fender amps. I may increase the grid leaks to 150K to ensure the power tubes are driven fully and call it a day.

To summarize, the circuit is basically stock AB763 except for the 100K grid leaks, 43K plate loads, and the 1 meg resistor in the reverb mix circuit - plus upgraded PT and OT - which produces a 42-watt output into 8 ohms. The stock AB763 phase inverter arrangement combined with the 43K plate loads -OR- 100K grid leak resistors seems a great way to reduce or eliminate low end flatulence caused by blocking distortion.

[pdf64]

'the BASS control is all but ineffective past "3" on the dial'

Many SFs had a bass control with a ~30% taper, same as the treble and volume.
The old BFs generally had a 10% taper; that provides much better resolution / useful control range over the dial.