...a hardware question: I'm in the second half of building a simple 12AU7 line stage designed for SS amps, a variation of Matt Renaud's well-known "Tube Color" arrangement.
My 12AU7 collection is extremely random. Not a matched pair in the bunch. Being OCD-afflicted about symmetry, I'm thinking of rewiring (not redesigning) so that instead of stringing the plate follower in the same tube with the cathode follower, the first gain stage would occupy one tube; the second in the other. IOW, V1's plate follower would drive V2's cathode follower. Ditto for both channels, bearing in mind that a fair amount of crosstalk would be somewhat desirable for my hearing-impaired purposes. Apart from arranging the four triodes side-by-side instead of in tandem, it should be exactly the same circuit as printed.
Might there any hideous consequences to this scheme?
My 12AU7 collection is extremely random. Not a matched pair in the bunch. Being OCD-afflicted about symmetry, I'm thinking of rewiring (not redesigning) so that instead of stringing the plate follower in the same tube with the cathode follower, the first gain stage would occupy one tube; the second in the other. IOW, V1's plate follower would drive V2's cathode follower. Ditto for both channels, bearing in mind that a fair amount of crosstalk would be somewhat desirable for my hearing-impaired purposes. Apart from arranging the four triodes side-by-side instead of in tandem, it should be exactly the same circuit as printed.
Might there any hideous consequences to this scheme?
Or, bite the bullet, and go to Eurotubes and purchase a couple of JJ 12AU7s that are both tested then matched and balanced?
Yes, it's better to use a single dual triode for both followers so you can elevate its filament properly.
Well, I'll give you some feedback based on experience having done some testing and amplifier design with the 12AU7, 5963, 5814a, etc. and at least a quantity of 50 or more tubes.
First, there's no need for the 1K grid resistors. The 12AU7 and variants don't have enough gain to require them. I don't use them ever with these tubes.
Second, I think you're running each section quite high on current... there's no need, depending on what you're attempting to drive at the other end. Using your voltages the cathode follower, you're running around 1.7 watts of dissipation on the plate. Likewise you're running the voltage gain stage close to 3 ma.
I would suggest trying something a bit simpler... direct-couple the cathode follower to the first stage... you'll eliminate 3 resistors and a coupling capacitor. Try running the first stage around 1ma... 150K plate resistor and 3.48K cathode resistor bypassed with a 6.8uF film capacitor. Use a 240 volt supply to drive it. This will give you around 93 volts on the cathode of the second section. Using a 33.2K resistor will result in about 2.8ma of current.
A 10K volume control is pretty low... unless you need that doe driving some amplifiers with lower input impedances. But only you can decide on that. In any case, the circuit changes above can easily drive a fairly high peak-to-peak output voltage and distortion should be quite low.
If you decide to try and find other tubes, suggest you look for some NOS JAN spec 5814a... much better tubes in general.
Have fun.
First, there's no need for the 1K grid resistors. The 12AU7 and variants don't have enough gain to require them. I don't use them ever with these tubes.
Second, I think you're running each section quite high on current... there's no need, depending on what you're attempting to drive at the other end. Using your voltages the cathode follower, you're running around 1.7 watts of dissipation on the plate. Likewise you're running the voltage gain stage close to 3 ma.
I would suggest trying something a bit simpler... direct-couple the cathode follower to the first stage... you'll eliminate 3 resistors and a coupling capacitor. Try running the first stage around 1ma... 150K plate resistor and 3.48K cathode resistor bypassed with a 6.8uF film capacitor. Use a 240 volt supply to drive it. This will give you around 93 volts on the cathode of the second section. Using a 33.2K resistor will result in about 2.8ma of current.
A 10K volume control is pretty low... unless you need that doe driving some amplifiers with lower input impedances. But only you can decide on that. In any case, the circuit changes above can easily drive a fairly high peak-to-peak output voltage and distortion should be quite low.
If you decide to try and find other tubes, suggest you look for some NOS JAN spec 5814a... much better tubes in general.
Have fun.
Lose the 10k ouptput pot - you’ve already got a volume control up front. Lose the bypass cap on the first tube’s cathode you don’t need so much gain with today’s sources. Keep the grid stopper on the cathode follower they have a tendency to oscillate in fact I would increase it to art least 4k7, bandwidth is not an issue since followers have no Miller capacitance. There’s nothing wrong with the given operating points, higher current gets you into the more linear part of the transfer characteristic’ 3mA is pretty low for a 12AU7 as a gain stage and, 8mA is good for a follower in order to drive the cable capacitance (long cables) and input capacitance of whatever power amp you’re driving. Nixie62 is correct - bias the heaters up by 45-50V and you’re good to go!