My first home-made tube project. Uses all "AA5" (All-American Five, refers to the dime-a-dozen line operated table radio tube lineup) tubes, in this case a type 35W4 rectifier, 50C5 output, and a 14GT8 or 12AX7 for the driver.
We now return you to your regularly scheduled tired, confused droning... ;)
Original schematic diagram, as you can see it's pretty old. The style I used (font, symbols to a certain extent, etc.) also dates it a little. ;) Anyway. The first thing that might hit you looking at it is - WTF?! No seriously, if you can read it, then you'll notice it's line operated. At the time I didn't have any suitable power transformers, and figured that if line is good enough for these tubes in their intended use, I might as well use that here. It's not very safe, and as noted can easily cause some problems. In fact, one time trying to connect it to the computer, I actually burned up a patch cable trying this!! Scared the crap out of me, thought I might've fried the sound card or something, but all that happened was a burn mark on the cable (a 1/8" headphones > stereo RCA homebrew (read: held together with masking tape) splitter cable), rendering it useless of course. Seems the ground line was so thin it made the perfect fuse, if not for being 3' long... Anywho, yes this incident is what prompted the "if plug is reversed" notice on the schematic. ;)
I will say it for the last time: ANYWAY! Circuit details:
Is line operated so is dangerous; either all inputs and outputs, or the power input, should be isolated with a suitable transformer.
Yes, 100uF is waaayyy too much for a 35W4. In fact, another scary moment was testing it with a 50DC4 (my only one :'( ). It worked on powerup, but testing it for ruggedness - turning it off then on again - popped the rectifier! The blue-white flash sure gave me a start. So after that I put in some extra resistance with the rectifier. And yes I now know that I would get better results with even two 22uF capacitors and the same resistor (well, probably) as a pi filter arrangement. (As is, it has a little hum.)
That covers the power supply; on to the signal-handling section. Come to think of it, C1 is a bit small, for a 220k grid leak (on V1) it should be at least .047uF (for 20Hz response). Oh well, not like the OPT can handle ten times that anyway... but I'll get to that in a moment.
V1 is shown as 1/3 - this is because the 14GT8 has two seperate signal diodes in it, something like a 6AL5 I guess. If you use a 12AX7 it would be 1/2 since there are two triodes in an 'AX.
V2 - 50C5 - output stage. Biased for 30mA (at least as I recall; the schematic shows 21mA, that seems low), wired for pentode mode in the schematic. For triode mode (making it SET, or Single Ended Triode, as advertised ;) you connect the screen to the plate through a small resistor, say 100 ohms (to prevent oscillations). The important part here, IMHO, is power: pentode mode gives significantly more power than triode, about 12dB worth (12dB = 16x; measured triode-mode power, assuming I didn't goof on something, was somewhere around 1/8W!) in this case.
One last thing, the OPT (Output Transformer) is probably 1 or 2W, 250 to 10,000Hz response (voice basically), and 2.5kohm primary to 8 ohm secondary.
But enough about the schematic, I'm sure you want to see the cursed thing, don't ya? ;) (Sorry about the crappy photos.)