Last edited:
There is a very long thread that dealt with the development of a driver circuit for use with 6L6GC's. During the course of that development I tested the circuit with several different tube types. The 3D21WB was included. I could get 75 watts from a pair in triode mode. This is discussed starting around post #331. The last driver schematic is in post #277. It works good and was not changed since then. The power supply was not completed for this amp, and I have not had time to work on it in a while. I will return to it, but I can't say when.
I have a totally new driver circuit (similar design but uses pentodes) to try that isn't done yet. I want to test both with the 3D21WB's and several other tubes when I get time.
http://www.diyaudio.com/forums/tubes-valves/133034-6l6gc-ab2-amp.html?highlight=7403
I have a totally new driver circuit (similar design but uses pentodes) to try that isn't done yet. I want to test both with the 3D21WB's and several other tubes when I get time.
http://www.diyaudio.com/forums/tubes-valves/133034-6l6gc-ab2-amp.html?highlight=7403
Check out what Tubelab did with his Octal Driver Board. Scroll down near the bottom. Looks like they can perform similar to KT-88s and a pair may be capible of near 100W in PP.
Octal Driver
Dang, George beat me to It. I spent too much time googling I guess....
Octal Driver
Dang, George beat me to It. I spent too much time googling I guess....
Last edited:
And I am usually the slow one.
I complained that I went through all of 2008 without building a single amplifier. I vowed that I wouldn't let 2009 go by without building at least one amplifier. Well as of Dec. 31 I didn't have any amplifiers that I had built in 2008 or 2009. I realized that I had built a Simple P-P in Nov. but I took it all apart so I could rebuild it while taking pictures for the assembly manual, so I guess that I did build one amplifier in 2009. After I put it back together, and build at least 2 more Simple P-P's, I plan to go back to the higher powered P-P stuff.
The Octal Driver board that is mentioned here is a good start for building several high powered P-P amps using conventional audio tubes (6L6GC, EL34, KT88). It also works well with sweep tubes, and other unusual tubes like the 3D21WB and 7403. I also plan to test it on some bigger tubes like the 813, and do some more screen drive experiments.
About two years ago I breadboarded a little 10 watt P-P amp using all pentodes. It used 6AU6's and 6AQ5's because I have hundreds of them. Local (Schade) feedback and direct coupling was used. It worked very well, but wound up on the shelf forgotten.
Recently Pete Millett designed the "Engineers Amp" using a similar all pentode with local feedback (but not direct coupled) circuit. I bought a board (I refer to it as the "red board") and built it. It works very well and I have already hot rodded it to 125 WPC. It also reminded me about the 6AQ5 amp. A search of my hard drive revealed a nearly complete PC board layout for that amp. I am thinking that it would make a good driver board too.
I still have the Universal Driver board too, so it will be time to see what they will all do when connected up to all sorts of different tubes. I bought 5 more of the 7403's (like a 3D21WB) so I expect that they will be part of the test.
tks tubelab.com,
this schematic resemble your idea,
and driver Schematic here:http://www.pmillett.com/images/UNIAMP.PDF,
is it ok, with KT88 opt : 6,6K->8Ohm, with 3D21WB, OPT=?
Ps: sorry, my English is not good.
this schematic resemble your idea,
and driver Schematic here:http://www.pmillett.com/images/UNIAMP.PDF,
is it ok, with KT88 opt : 6,6K->8Ohm, with 3D21WB, OPT=?
Ps: sorry, my English is not good.
Last edited:
That question was asked back in the 6L6GC in AB2 thread. Along with this one:
My answers in posts # 339 and #341. Were pretty much focused on making a big amp using a triode wired pentode tube, without any feedback. The discovery of the 3D21 / 7403 with its 850 volt screen grid rating seems to be the enabler for this idea and the initial experiments proved that it would work.
Since then I have spent some time cranking some stupid amounts of clean good sounding power out of the Red Board, I may abandon the "triode only" requirement. I set up a breadboard for testing all sorts of different octal tubes with the Octal Driver board last year. I have not powered it up yet. Once the Simple P-P stuff is out of the way, I hope to explore the upper limits of a bunch of different tubes.
The 6146 got a bad name in the audio world (post #339 again) back in the late 60's and has been pretty much shunned ever since. I gave all of my 6146's and 6159's to a friend who wanted to build a broadband distributed RF amplifier several years ago. I know that there is at least a dozen sitting in a box somewhere from a military surplus pulse generator that I parted out a year or so ago. If I can find them, I can crank them up. They should do OK in pentode mode.
lieukiem, I think that the driver board shown will work. It uses a transformer for the phase splitter, which I prefer to avoid. Transformers that work good are not cheap, and Sowters are not common in the US. If you chose to use this board, I would look for a quality transformer.
Anytime you chose an uncommon tube you must be prepared to experiment since there aren't many examples available to copy. I still haven't had the time to fully investigate the 3D21WB tube, and it may be a while before I have the time to do so. Most of my experiments were done with a 3300 ohm load. This was obtained by using a 6600 ohm transformer with the 8 ohm load on the 16 ohm tap. That worked good in triode mode. My output transformers don't have a UL tap, so that was not tested. you will need to adjust the values of the cathode bias resistors for proper current depending on the choice of tubes and power transformer.
From what I've read, the RCA 6146B introduced in 1964 was a problem for transmitters as well. It was supposed to be a drop in replacement for the 6146/6146A with higher plate dissipation rating. However, it had a nasty habit of parasitic oscillations in transmitters and often resulted in burnt circuits. I think there were some Collins Motorola and Heath Kit transmitters. I got my tubes from my dad who used to have a Collins (what model I don't know).
It seems the biasing requirement for the 6146B is different than the bias for the 6146/6146A.
All my tubes are 6146Bs except one 6146.
It seems the biasing requirement for the 6146B is different than the bias for the 6146/6146A.
All my tubes are 6146Bs except one 6146.
I have Some 3D21WA, I have searched the web for Data on this tube; I can find the 3D21WB but not the WA. Just looking at the WA the plates look identical to the WB{ I have 1 3D21WB} exactly the same. Can the WA stand 40 watts or only 15 like the 3D21A.
thanks
phil
thanks
phil
puremilk.
What impedance do you use with the 3D21?
I hope to use the 3D21WA/WB class Ab2 Push pull triode. I have 4500 P-P out put transformers.
Phil
What impedance do you use with the 3D21?
I hope to use the 3D21WA/WB class Ab2 Push pull triode. I have 4500 P-P out put transformers.
Phil
Multi,
I have been in the process of moving all my stuff and Tubelab about 1200 miles, so I haven't been able to answer too many questions.
I have several 3D21 type tubes.
The 3D21A's that I have are all in coke bottle "ST" shapes and have tiny rounded off plates. They are rated at 15 watts and I would tend to believe this number, but I didn't try them.
The 3D21WB's have straight glass and a brown phenolic base. These are rated for 40 watts and have large rectangular plates. I had about 5 of these, all used, but only one was good. The others had low emission or were gassy. I would test all of these tubes at the voltage and current you intend to use, or at least try your circuit with junk OPT's and a bench (or fused) power supply. The gassy tubes tended to run away without warning and these tubes can pass some serious current.
I had one 7403 which have straight glass and a metal base like the KT88. It was good so I paired it with the good 3D21WB and did my testing. The results were good so I ordered some more 7403's.
My experiments and results were buried somewhere in the "6L6GC in AB2" thread, but I used a 6600 ohm OPT because I have a bunch. I used an 8 ohm load on the 16 ohm tap for a 3300 ohm load to the tube and explored operation up to about 500 volts, yielding nearly 100 watts in triode.
A 4500 ohm OPT is probably a good choice and would make a bit less power than I was getting with better distortion and damping factor.
As for the 3D21WA's, I don't have any, but if they have large rectangular plates they are probably good for 40 watts. It was common back then to use the latest guts to stuff the glass of earlier tubes since it makes for easier manufacturing, but a military contract probably called for a specific type number. Your WA's may actually be WB's.
The most common example is the 807. These should all have 6L6GA plates, but the 807 lived on long after 6L6GA production ended. I have seen "807's" with 6L6GA, 6L6GB, 6L6GC, and 7027A guts.
I have been in the process of moving all my stuff and Tubelab about 1200 miles, so I haven't been able to answer too many questions.
I have several 3D21 type tubes.
The 3D21A's that I have are all in coke bottle "ST" shapes and have tiny rounded off plates. They are rated at 15 watts and I would tend to believe this number, but I didn't try them.
The 3D21WB's have straight glass and a brown phenolic base. These are rated for 40 watts and have large rectangular plates. I had about 5 of these, all used, but only one was good. The others had low emission or were gassy. I would test all of these tubes at the voltage and current you intend to use, or at least try your circuit with junk OPT's and a bench (or fused) power supply. The gassy tubes tended to run away without warning and these tubes can pass some serious current.
I had one 7403 which have straight glass and a metal base like the KT88. It was good so I paired it with the good 3D21WB and did my testing. The results were good so I ordered some more 7403's.
My experiments and results were buried somewhere in the "6L6GC in AB2" thread, but I used a 6600 ohm OPT because I have a bunch. I used an 8 ohm load on the 16 ohm tap for a 3300 ohm load to the tube and explored operation up to about 500 volts, yielding nearly 100 watts in triode.
A 4500 ohm OPT is probably a good choice and would make a bit less power than I was getting with better distortion and damping factor.
As for the 3D21WA's, I don't have any, but if they have large rectangular plates they are probably good for 40 watts. It was common back then to use the latest guts to stuff the glass of earlier tubes since it makes for easier manufacturing, but a military contract probably called for a specific type number. Your WA's may actually be WB's.
The most common example is the 807. These should all have 6L6GA plates, but the 807 lived on long after 6L6GA production ended. I have seen "807's" with 6L6GA, 6L6GB, 6L6GC, and 7027A guts.
I used them w/ a 4.2k output tx. somewhere between 450-500v on the plates. Triode.
It sounded very good, but I didn't do much testing.
It sounded very good, but I didn't do much testing.
Tubelab
Thanks for your reply; while you are under the horror of Moving; can’t even imagine Moving with all my junk.
I compared the 3D21WA, to the WB and they have the same plates.
I have 3D21A & B totally different; small plates to the WA& WB.
I purchased some of the transformers recommended by Peter Millet{Dynaclone A431S}
for his 807 & 815 amps, 4500P-P. Intending to use them with 6550’s, but the 3D21 sounds more interesting. Even the 3D21A &B would be ok at 300 volts.
The A431’s have ultra linear taps, so I can try both Triode and Ultralinear.
I find testing the 3D21’s the 3D21A &B test very well, but the WA & WB read very low on first test, running the heaters Alone they creep up.
Phil
Thanks for your reply; while you are under the horror of Moving; can’t even imagine Moving with all my junk.
I compared the 3D21WA, to the WB and they have the same plates.
I have 3D21A & B totally different; small plates to the WA& WB.
I purchased some of the transformers recommended by Peter Millet{Dynaclone A431S}
for his 807 & 815 amps, 4500P-P. Intending to use them with 6550’s, but the 3D21 sounds more interesting. Even the 3D21A &B would be ok at 300 volts.
The A431’s have ultra linear taps, so I can try both Triode and Ultralinear.
I find testing the 3D21’s the 3D21A &B test very well, but the WA & WB read very low on first test, running the heaters Alone they creep up.
Phil
That's the output tx I used. (sorry, meant 4.3k, above.. not 4.2)
I also used the power tx from Mr Millett's 807 project. I was building something else though, so just played w/ the 3d21wb's while it was on the workbench for a couple days.
I only tried triode, which worked very well. The tube has some ugly pentode curves.. not sure if that means anything for UL operation.
I also used the power tx from Mr Millett's 807 project. I was building something else though, so just played w/ the 3d21wb's while it was on the workbench for a couple days.
I only tried triode, which worked very well. The tube has some ugly pentode curves.. not sure if that means anything for UL operation.
Testing the 3d21's i have; all appear to be NOS. 3D21A, B test as new. All the 3D21WA & WB have problems. I have a CATHODE bias amp; I can earth the grids via a 1k resistor so the tubes can't run away. So far the WB's &WA's Tungsol all show grid leakage when the 1k resistor is removed, leaving the 470K resistor. Maybe they need a lower value? I get reading of grid leakage up to 3 volts.
Phil
Phil
I had runaway issues in some of my tubes. I was using a mosfet follower with a 10 K resistor to -100 volts on its source tied directly to the grid. Even with this setup a couple of tubes tended toward runaway. One of them decided to glow blue as I cranked up the plate voltage, so they were gassy.
Wicked 1
I used them w/ a 4.2k output tx. somewhere between 450-500v on the plates. Triode.
It sounded very good, but I didn't do much testing.
Would you have the circuit you used?
Thanks
Phil
I used them w/ a 4.2k output tx. somewhere between 450-500v on the plates. Triode.
It sounded very good, but I didn't do much testing.
Would you have the circuit you used?
Thanks
Phil
I was just playing w/ Pete's PP driver PCB
Push-pull driver board
At the time, triode wired 12gn7 drivers, no feedback. I might have only been using two stages, but don't remember.. (I'm sure I had it set up that way at some point).
Push-pull driver board
At the time, triode wired 12gn7 drivers, no feedback. I might have only been using two stages, but don't remember.. (I'm sure I had it set up that way at some point).
- Status
- Not open for further replies.