Lots of advice. But pentodes in general are very much dependent on the plate & screen voltages to arrive at the optimum load impedance.
For maximum output at any plate voltage the output load impedance is dependent on the screen voltage used.
That becomes obvious when one looks at the plate voltage curves for Zero grid one voltage while the screen voltage is varied.
The ends of the loadline need to go up into the knee of the plate current curve. Otherwise max power is compromised
& screen dissipation increases when large signal occurs.
Curves for the 6L6 are here. The 807 is a 6L6 with a plate cap & a 5-pin base. That's All. 🙂
For maximum output at any plate voltage the output load impedance is dependent on the screen voltage used.
That becomes obvious when one looks at the plate voltage curves for Zero grid one voltage while the screen voltage is varied.
The ends of the loadline need to go up into the knee of the plate current curve. Otherwise max power is compromised
& screen dissipation increases when large signal occurs.
Curves for the 6L6 are here. The 807 is a 6L6 with a plate cap & a 5-pin base. That's All. 🙂
Attachments
Indeed they are over specified. The previous owner had them set up with EL84's. I had heard them perform. Beautiful sound but they ran out of puff with his large vintage Pioneers with 12" drivers.
- Technically not a 100% gift. More of a barter. But I am happy as the new cost is prohibitive. These are atleast 15 years old.
- Honestly I would like to future proof an aim for at least 20 watts. A large Tannoy Devon could be making its way to me in due order. Will be a bummer to have only 10 watts on tap when that heppens.
- I think UL would be the way to go. I will measure the exact % at which the UL's are tapped for you to make a recommendation.
- 807 look beautiful and more vitally, I have a about 6 pairs in sealed NOS condition. Another power tube NOS or otherwise with involve shelling out crazy money.
- What would be the preferred method to keep the screen voltage down if going full tetrode mode? I was thinking of using Zeners.
Philips Philips AD-12100 12" in a 50 lt sealed box. My EL34 SE in triode drives these to insane levels. DCR is 8.8 & 9.3 ohms respectively as measured with my Fluke multimeter.
Last edited:
What a superb set of reading recommendation ! That Dalmura resource is a goldmine.
Will be doing some detailed reading as days go by but at the moment I strongly feel that the the Sowter opt could stand upto the Williamson task.
I shall redraw the Musician's amp and the subsequent mods into a easier to read schematic. It looks messy in a cute vintage way right now.
And thanks also for the gracious offer of the DM facility for any help. I am planning to go easy and hope to start ordering parts in next 2 months.
Thanks again for a fantastic set of pointers.
The early post-War amplifiers, like the DTN Williamson and the one in Radio and Electronics, are (and were in their day!) very challenging for home constructors. Very high open loop gain and fairly high feedback around too many stages plus the output transformer, make them poor choices for folks without some measuring equipment and a burning desire to learn about Bode plots and stability.
Modern DIY has been trending towards fewer stages and feedback, if any, taken before the output transformer. Type 807 has fallen out of fashion, but is an excellent choice with your output transformers (also excellent by reputation). A Williamson topo with the middle stage and the loop feedback removed might be a good start. Maybe leave room on the chassis for an extra valve socket per channel, to give yourself something to worry about in five or ten years.
All good fortune,
Chris
I am not exactly compelled to follow the Williamson. But most searches about 807 PP bring up a Williamson.
I am sort of newbie despite a couple of successful tube projects. It would help me immensely if you can suggest a non Williamson type schematic that can use the 807 in UL or triode. As I understand the 807 probably needs a better drive than a pure audio tube such as EL34.
There were literally thousands of DIY 807 Williamsons built in the early 50's. I've personally restored early 807 Williamsons with the original 807s still in good condition. There's ample documentation that with 400 volts across the tube, the screens hold up fine in either triode or UL mode. I wouldn't exceed that, but there should be no problem.
Perhaps the many schematics of the Williamson amplifiers that have output tubes such as the 807 and 6L6, are because of the vintage of Williamson design and the vintage of the 807 and 6L6 family design,
The KT66 is also used, because it has very similar characteristics as the 807 and 6L6 family.
Generalizations:
The 6CA7 and EL34 characteristics are very close to each other.
And, they both have about 1.5 x the transconductance (Gm); and about 1.5 x the g1 to g2 u (mu) . . .
When compared to the 807, 6L6 family, and KT66's Gm and u.
The KT66 is also used, because it has very similar characteristics as the 807 and 6L6 family.
Generalizations:
The 6CA7 and EL34 characteristics are very close to each other.
And, they both have about 1.5 x the transconductance (Gm); and about 1.5 x the g1 to g2 u (mu) . . .
When compared to the 807, 6L6 family, and KT66's Gm and u.
Last edited:
Type 807 was manufactured in the gazillions to win the War with, and was a drug on the market post-War. Higher perveance post-War valves like the EL34 were marketed into a horsepower race in the 1950s. They were perfectly fine, but hardly of the construction quality of the made-to-win-the-war 807. The very similar type 1625 is still affordable and available today, and a sleeper for DIYers.
807 and EL34 both have triode (G1-G2) mu of about 8 to 10, so need similar drive voltages. Any driver circuit you like that specs type EL34 could be used with your OPT (output transformer) and type 807. I tend to like something like the first two direct-coupled stages of the Williamson, but with a 12AX7 or 6SL7 type first stage and a 6CG7/6FQ7 or 6SN7 second stage. Leave out the Williamson's middle push-pull stage. Keep its bias and balance adjustable cathode bias, without capacitor bypass. With triode connected outputs you won't need loop feedback; with partial-triode ("Ultra-Linear") you may want some, may not.
All good fortune,
Chris
807 and EL34 both have triode (G1-G2) mu of about 8 to 10, so need similar drive voltages. Any driver circuit you like that specs type EL34 could be used with your OPT (output transformer) and type 807. I tend to like something like the first two direct-coupled stages of the Williamson, but with a 12AX7 or 6SL7 type first stage and a 6CG7/6FQ7 or 6SN7 second stage. Leave out the Williamson's middle push-pull stage. Keep its bias and balance adjustable cathode bias, without capacitor bypass. With triode connected outputs you won't need loop feedback; with partial-triode ("Ultra-Linear") you may want some, may not.
All good fortune,
Chris
I think the EL34 and 6CA7 g1 to g2 u is close to 11; and Gm is about 11,000 micro Mhos.
807, 6L6, and KT66 g1 to g2 u is close to 8; and Gm is about 6,000 micro Mhos.
Push Pull:
At one time the Triode wired 807 and KT66 Williamson were very popular.
I am not sure if the Triode wired EL34 / 6CA7 was ever 'popular' in commercial designs.
"Ultra Linear" has been used over a range of from about 20% to 80%.
Triode Wired "is" 100% "Ultra Linear"
Beam Power / Pentode wired "is" 0% "Ultra Linear"
807, 6L6, and KT66 g1 to g2 u is close to 8; and Gm is about 6,000 micro Mhos.
Push Pull:
At one time the Triode wired 807 and KT66 Williamson were very popular.
I am not sure if the Triode wired EL34 / 6CA7 was ever 'popular' in commercial designs.
"Ultra Linear" has been used over a range of from about 20% to 80%.
Triode Wired "is" 100% "Ultra Linear"
Beam Power / Pentode wired "is" 0% "Ultra Linear"
Last edited:
An 807 is a "pure audio tube" and needs no special drive requirements when used in this way. Typical Williamson amps bias at 34-38 volts, be they 807s, 6L6GCs, KT66s or EL34s.
The article you included in post #10 is, in fact, very close to a Mullard-type design, with a pentode voltage amplifier and a long-tailed phase splitter. It's a bit simpler than a Williamson, but since we don't know the ouput transformer used, it's still anybody's guess as to how well the Sowter will work. When those articles were written, it was a given that a competent DIYer had a 'scope and other analytical tools at hand, and could order a Triad or a Stancor or a Chicago or what have you, and tune the circuit accordingly.
Just about any PP amp using a beam power tube or pentode is going to need feedback to some degree. I've seen no-feedback 6550 designs, but I assume that's possible because the plate resistance is so low. The 807 in triode operation will have a plate resistance of about 1700 ohms. That's pretty high for a no-feedback amp. A quick search for PP 807 amps yields nothing that doesn't call for feedback, but there might be something out there.
I'm not saying you can't experiment, of course. But you have some very nice OPTs there, it would be worth learning how to build a quality amp with them.
Last edited:
By the time the EL34 came on the market, no one was interested in a 10 watt PP amplifier. ;-) To the best of my knowledge, the Mullard 20 watt amplifier was the first to use the EL34. In the American market, Dynaco and Eico followed with 30-40 watt designs.
I thing it will be easier to draft build with with the standard designs you have suggested and see where it leads to. This would be my first line of attack before approaching something complicated.
Thanks for wishing me success.
Can you suggest a PP schematic with minimum stages to drive the 807 in PP.
Am keeping the pure Williamson topology as a 2nd option.
The Heathkit W4AM is basically a slightly modified Musician's Amplifier with added stability measures and an updated power supply, as were all the Heathkits up to the W5. The schematic is easier to read. This is actually what I pretty much use now. The only thing I add is 15 ohm test resistors between the output tube cathodes and the main cathode resistor, to make it easy to balance the output tubes with an ohmmeter. The feedback values won't be the same, and the input cap isn't necessary with modern preamps, but the rest of the circuit is pretty much a standard Williamson.
http://www.tubebooks.org/file_downloads/W4-AM_Manual.pdf
Last edited:
This is a key learning. And just for clarity which exact area in that graph is the "Knee"? Pardon my ignorance.
I also remembered having a pair of identical power transformers with HT winding of 320-0-320 at 200mA each. I recon these will be good for about 400 ~ 420 vdc as final B+ with tube rectifiers and CLC. I just need to order a couple of 5H chokes and bases for the 807's.
