Well, then here's your chance!
After collecting RCA 808 triodes for over fifteen years I finally have three pairs, enough to build a SET with them.
808 is a close relative to the much more common 811A and must be pushed far into class A2 in order to produce any useful output power.
The easiest way to do this is by using a decent-sized power tube as a cathode follower, direct-coupled to the grid of the power tube.
Here I have picked a selection of tubes that should be up to the task (807, 6L6GB, 6AV5GA and PL36) and now I would like your opinion about which one that would be the best visual match to the 808 (the one to the left of the 807).
After collecting RCA 808 triodes for over fifteen years I finally have three pairs, enough to build a SET with them.
808 is a close relative to the much more common 811A and must be pushed far into class A2 in order to produce any useful output power.
The easiest way to do this is by using a decent-sized power tube as a cathode follower, direct-coupled to the grid of the power tube.
Here I have picked a selection of tubes that should be up to the task (807, 6L6GB, 6AV5GA and PL36) and now I would like your opinion about which one that would be the best visual match to the 808 (the one to the left of the 807).
Thanks for your input!
PL36 is a strong candidate, possibly the best choice performance-wise and I have a good stash of them from old quality brands. The top cap is a plus, but a tube with a 6,3V heater would be more convenient in that aspect.
I asked a friend to have a look and she suggested the 6AV5GA which I also think would be a good match, both electrically and aesthetically.
I had already pretty much decided to use the 807 but it would probably look better witha driver tube that isn't almost as big as the output tube.
Technically, a mosfet or at least a more modern, high gm tube would be the best choice but I would love to have the slow startup from an IDH tube and in this case I'm actually prepared to compromise a bit with the performance to get the looks just right.
PL36 is a strong candidate, possibly the best choice performance-wise and I have a good stash of them from old quality brands. The top cap is a plus, but a tube with a 6,3V heater would be more convenient in that aspect.
I asked a friend to have a look and she suggested the 6AV5GA which I also think would be a good match, both electrically and aesthetically.
I had already pretty much decided to use the 807 but it would probably look better witha driver tube that isn't almost as big as the output tube.
Technically, a mosfet or at least a more modern, high gm tube would be the best choice but I would love to have the slow startup from an IDH tube and in this case I'm actually prepared to compromise a bit with the performance to get the looks just right.
Good idea, or maybe even do it the other way around and put the 808s in an elevated position? Long ago I saw a picture of a big transmitter tube amp with the power tubes on top of the output transformers.
I've also been looking at 2C34 / RK34 / VT61, nice looking tubes with the same bulb shape as 807 but smaller and with dual top caps. A bit shy on plate current though, the 808 seems to need 70mA or more into the grid for full drive, and they are a bit expensive and hard to find here in Europe.
Yes, I see what you mean. Perhaps the PL36 is the best visual match to the 808, even though I think 807 and 6AV5GA are better looking tubes on their own but the 807 is a bit too big and the 6AV5GA looks a bit too modern.
This question has been on my mind for over fifteen years, spawning a few rather wild ideas over time including the fugly RCA 815 "R2D2" tube, several miniature 7-pin tubes in parallel and even various solid state solutions. An LM3886 with asymmetrical rails could get the job done, but they dont look quite as good as tubes with top caps.
I don't have a full schematic yet but the output stage should look something like this:
With a little luck I can get away with just the 4-5dBs of negative feedback from the 16R tap straight into the cathode. If so, the input stage can be any generic triode gain stage that can swing +-70V or so.
With a little luck I can get away with just the 4-5dBs of negative feedback from the 16R tap straight into the cathode. If so, the input stage can be any generic triode gain stage that can swing +-70V or so.
Thanks for your vote!
Mosfets are not out of the question, in fact that's what I've been using to drive my 808s in various test setups through the years. They make excellent followers but I would very much prefer to have the slow startup from an IDH tube. If the grid voltage goes positive before the B+ has risen, the grid will draw excessive currents.
Fets also have a nasty habit of failing as shorts, which definitely would melt the grids of my rare and fairly expensive triodes.
Both these problems can of course be avoided with some clever eningeering but an reasonably dimensioned IDH cathode follower would provide both startup delay and some current limiting withouth further efforts.
Mr Nobukazu Shishido used inverting interstage transformers to drive positive-biased triodes in a very clever manner where the grid current partially cancelled the plate current of the driver tube. AFAIK this required highly specialized (and expensive) transformers but on the shelf behind me I have a pair of LL1671 / 50mA ITs. Perhaps it would be worth trying them out connected as 2:1 and non-inverting. With a 2:1 stepdown, 20mA standing grid current through the secondary should leave room for 40mA through the primary which should be enough. Worth a try once I have a working prototype. The required input signal at the 808 grid should be around 25Vrms or 35Vrms if there is CFB from the 16R tap.
Hard to tell. Assuming that all of them can handle enough plate dissipation, voltage and peak current, then the best choice is the one with the highest gm. 6L6/807 is 1930's technology while 6AV5GA and PL36 are sweep tubes, presumably from the early fifties. Any of the tubes mentioned here should work but high gm would be very benificial for a CF working into a low Z, non-linear load.
That's one book I would like to own!!
And probably the highest gm too, if I may take a wild guess without looking at any curves.