The 12GN7 / 6B4G PP Thread

[jaymanaa]

It seems like Sylvania had the market cornered on these small pentodes. Every one I've found so far (7) are Sylvys.:thmbsp:

 

[bricktop]

I'm glad only a few audio amps used video pentodes, otherwise the stashes of them might not be around today. A couple of stu hegeman's amps did, but not a whole lot else....

 

[Tinkerbelle]

Believe that GE and others made 12GN7s.

 

[www.records]

I have some Tungsols and Raytheon 6B4Gs in factory boxes. Don't know much about tubes yet, but thought I would chime in hoping to learn something. Do you think they are rebadged?

 

[jaymanaa]

Believe that GE and others made 12GN7s.

Yep, I just found some GEs and RCAs.

Micah, I tested those Fishy outputs and they are 6.4K pri. Z (just so ya know).

I'm assuming around 320 or so for the main B+. I should be able to get a power supply cobled together this weekend.:thmbsp:

 

[jaymanaa]

I have some Tungsols and Raytheon 6B4Gs in factory boxes. Don't know much about tubes yet, but thought I would chime in hoping to learn something. Do you think they are rebadged?

I'm not sure on this one. I have RCAs, CEIs, and older Svetlana Winged Cs.

Tunsol and Raytheon had some of the finest vacuum tubes around IMHO.

 

[jaymanaa]

I think I have the iron gathered up; the 6.4k Pri. Z Fisher OPTs, a nice potted 10hy choke (500wv, 150DCR, 150mA), and this power TX. It has one very beefy (large wire size) 6.3 tap for the 6B4Gs, a regular looking 6.3 tap for the 12HL7s, a 5v tap for a rectifier, and 730vct secondary. I'll start working on the power supply tomorrow, what do you think about choke input for this?

 

[wajobu]

Jay's busy again!

 

[bricktop]

I think I have the iron gathered up; the 6.4k Pri. Z Fisher OPTs, a nice potted 10hy choke (500wv, 150DCR, 150mA), and this power TX. It has one very beefy (large wire size) 6.3 tap for the 6B4Gs, a regular looking 6.3 tap for the 12HL7s, a 5v tap for a rectifier, and 730vct secondary. I'll start working on the power supply tomorrow, what do you think about choke input for this?

Choke input will do just fine. It will be a pretty much constant current load with class A push-pull, so it's a relatively easy job for a power supply. The 6.4k OPts will be great. Not huge power like with a 5k or so, but it should be pretty low distortion since the load-lines will be flatter.

With a 6.4k load and class A PP, it's a 3.2k load to each tube. Bias it at 270V and 50mA with a 320V supply and 50V bias via a cathode resistor of 50/0.1 = 500 ohms (bypassed with 100uF, or can use two separate 1Kohm cathode resistors for better balance and 50uF bypass caps) -> which should make you about 10 watts output power with 13.5 watts idle power per tube. Can use that 50V for the screen supply of the 12HL7s with ~12k loads (higher since we dropped the screen V a bit). I initially picked 62V because that's what the 2A3 datasheet said for PP, but we are running all class A, they assumed class AB (one tube cuts off, thinner bias for more power - you could run class AB if you wanted to for more power, but class A is always sweet and simplifies design [no kinked load lines]). This should be a pretty sweet rig :yes:

 

[jaymanaa]

Choke input will do just fine. It will be a pretty much constant current load with class A push-pull, so it's a relatively easy job for a power supply. The 6.4k OPts will be great. Not huge power like with a 5k or so, but it should be pretty low distortion since the load-lines will be flatter.

With a 6.4k load and class A PP, it's a 3.2k load to each tube. Bias it at 270V and 50mA with a 320V supply and 50V bias via a cathode resistor of 50/0.1 = 500 ohms (bypassed with 100uF, or can use two separate 1Kohm cathode resistors for better balance and 50uF bypass caps) -> which should make you about 10 watts output power with 13.5 watts idle power per tube. Can use that 50V for the screen supply of the 12HL7s with ~12k loads (higher since we dropped the screen V a bit). I initially picked 62V because that's what the 2A3 datasheet said for PP, but we are running all class A, they assumed class AB (one tube cuts off, thinner bias for more power - you could run class AB if you wanted to for more power, but class A is always sweet and simplifies design [no kinked load lines]). This should be a pretty sweet rig :yes:

Holy cow! I understood every word of that.:yes: Ten watts will be a "monster" amp around here, so that's plenty. Thanks for that post, it really helped a few things come together in my "P" brain. Another ten or twenty years, and I'll have this stuff figured out.:thmbsp:

 

[bricktop]

Here's a scribbled schematic, missing a few things. At the very bottom needs to be minus (-) 20 V. If you have a spare 6.3VAC filament winding, you can get it with a voltage doubler with some 1N4007s. :thmbsp:

 

[Kegger]

Very nice, this should be fun to see!

Only issue I see is if Jay's using "that" choke for input duty I don't think you'll be
anywhere near 320v B+, the resistance is to high and you'll drop to much V I bet.

Do you have 2 of those power tranny's Jay and 2 of the chokes to, Mono blocks?

 

[jaymanaa]

Thanks Micah :thmbsp:

Keg, I could snag a couple control TXs easy enough, but no center taps. Of course we know how to work around that. I can't wait to get started.

 

[bricktop]

I can build the driver stage later on this week and test it out for distortion and if there are any quirks. :yes:

 

[jaymanaa]

You da man Micah! :thmbsp: Thanks, Jay

 

[bricktop]

Depending on when I get done babysitting, may have a 12HL7 diffy amp by tonite :thmbsp:

 

[bricktop]

I gave the circuit a test. With a 12K load, 280V B+, ~22 mA per pair with the CCS, 50V screen voltage, there is about 2% THD at the 37Vrms needed to drive the 2A3s to full power. That's total THD, so the portion of it that is 2nd order will cancel in the output stage. The circuit give a single-ended output of 57Vrms at 5% THD (160Vp-p), we'll only need 100Vp-p max prolly. No oscillation problems, but I did have a 47ohm screen resistor on each screen plus 510 grid-stoppers on grid 1s. I also had a 2.5K pot with the wiper to the 50V supply each end to a screen for balancing the tubes at bias. I can scribble up another schematic if you want.

 

[kvflyer]

I have some Tungsols and Raytheon 6B4Gs in factory boxes. ....

Most tube manufacturers bought tubes from each other to fill their needs. Not unusual to find an RCA tube that is obviously a GE tube. GE used etched numbers with date code dots below the numbers. RCA usually used the "stop sign" around their numbers. Westinghouse numbers on the top of the envelope and Sylvania numbers on the side of the tube near the top. All of which could be labeled "RCA" Just adds to the confusion!

 

[Kegger]

Way coool Brick, I'd like to see the schem.

 

[bricktop]

Schematic, and it's about time. :yes: