• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Simple follower with 12AX7

I am working on adding a tube stage to a solid state preamp I have already built. I am not looking for any gain, since it's already designed into the pre-amp. I only want to add a unity gain (or close to unity) follower circuit using a single 12AX7 for both channels. I am using this tube because I have one laying around and I'm a cheap skate =).

I would like to avoid negative feedback ("degeneration" is ok of course), and I will probably be using 300-350 volts for the plate supply - which is high, but available from the transformer I am going to use. It will need to swing +- about 30 volts to maintain headroom in the preamp. The tube will drive the high non-inverting input impedance of an op-amp.

Are there any links to simple followers using this tube without negative feedback, with calculated resistor values -- or at least something to start with?

i am a complete newbie to the whole tube scene so I might not even know what I'm talking about. Please correct me


a cathode follower has the highest local negative feedback possible.

If you insist on using a ECC83 / 12AX7 for a cathode follwer, then please do it. It will work; it drives a very easy load. But this tube is the last i would choose for that purpose.
OTOH, the 12AX7 is one of the best sounding tubes for gain stages, provided the input capacity doesn't hurt. (Hint: Spare it for that purpose)

The 12AX7 is meant to run at 1.2 mA with a plate supply of 250V=. As the tube has high amplification (µ=100), its bias at this current is only -2V. Of course it is a follower and will follow the input voltage, but for a very short period it will draw grid current if the signal swings too fast and too high.

You may want to have more headroom at the input. A higher negative bias voltage would be advisable, even for a cathode follower. Better go with a ECC82 / 12AU7 / 5814A. It runs at 10.5 mA with 250V= supply and -8.5V bias at the grid.
Or use a 6SN7 if you have one.

If you run this stage as an insulated additional linestage, you will need an input coupling cap and also an output coupling cap.

Have a constant current source as "load resistor" for your cathode follower and adjust it to exactly the current the tube is meant to run with. Have the grid biased at a voltage that ensures that the tube has atleast, say, 150V, better 175 or 200 across it when no input signal is there.
(For that you need the input coupling cap)
Then the tube and the CCS forms a voltage divider, each having about the same quiescent voltage across it and the whole assembly can swing easily to both sides with lots of headrooms. If there is too much voltage across the CCS, put it in series with a resistor.

Avoid resistor-loaded cathode followers.
Unlinear and sounds like crap.

Hoping that helps.
dice45 said:

The 12AX7 is meant to run at 1.2 mA with a plate supply of 250V=. As the tube has high amplification (µ=100), its bias at this current is only -2V. Of course it is a follower and will follow the input voltage, but for a very short period it will draw grid current if the signal swings too fast and too high.

Thanks for the help. I have decided to use the tube right on the input stage (preceded by an op-amp buffer I dont know yet) since it will not likely see more than 2 volts. I guess I could design some sort of regulator for the plate supply maybe even using tubes for that too. Now I'm probably getting way over my head! :rolleyes:

I definately want the sweet tube sound so I don't want to run the plate too hot or too cold, or use the wrong bias current. Will the recommended (250v, 1.2mA) bias give the best sound?

Here's a circuit I found (to use one tube for both channels since it is a dual-triode). I want a fixed gain anywhere from 0-10db, so what should I use for Rk? I assume an input coupling cap would be nice.


  • img38.gif
    2.4 KB · Views: 1,307

in that circuit the tube has unity gain.
The cathode practically follows the grid, it is always 2V above the grid (the bias!). And it is sitting in the corner; nearly no voltage across Rk, whole supply voltage across the tube. Its death is near with 350V= supply.

If you introduce a voltage divider of 1 MOhm at the tubes grid, divided into 2 halves, tube and Rk each have half the voltage across it. Then you can adjust Rk to 1.2mA, this would be 146kOhm.
It may be better to decouple the volatge divider with a cap from the PS.

But you need an input coupling capacitor.

Please re-read my last post. And you can replace Rk with a constant current source adjusted to 1.2mA.
OK I built a test circuit today and tried various B+ voltages. The 12AX7 only worked well with voltages >100V to give me the swing I wanted. I tried a 12AU7 and it worked fine down to about 20V, which is near the regulated rail voltage of my pre-amp. I am going to try using the rail voltage as well as the 5V digital supply already present in my pre-amp to run the heater and see what happens. This would save me from adding another transformer and having dangerously high voltages present.

Would using only 20V for the B+ ruin the "tube" sound? It looked great on a scope but I can't compare it to anything. I measured the bias current somewhere around 0.1mA with that voltage with a 22k Rk.


me being a tube person striving for freedom of coloration :) , i
slowlyslowly am realizing what you want to achieve (correct me if i am wrong).

As i understood it, you want to feed a cathode follower by opamp or SS devices and this cathode follower feeds another opamp.
So you want this cathode follower produce tube sound, tube coloration for you, right?

Assuming so, in contrary to my former recommendations i recommend to use low operating voltages, you can use your SS preamp's +/ supply of say +/15V (the higher the better, but 2x15 will do.)
You already have tried out that the 12AU7 works fine at 20V supply, so 30V will do better. Hook the anode of the tube to the + rail and the lower end of Rk to the - rail. Then the tube's grid is centered voltagewise between + and - and you need no input coupling cap. But the output coupling cap is necessary.
You want coloration, at this supply voltage you will get it. 20V or 30V will not ruin the tube sound, it will produce it :)

You can try out whether you like the sonics better with resistor or with current source in the cathode. I would bet on the CCS.

Underheating: at those supply voltages it is no problem to use 5V= heater supply instead of 6.3V=. It only will enhance your tube's life to almost eternity :).
And it will help to smoothen your digital power supply as a "bleeder resistor" as it adds a totally constant current load.
You've got my plan right on the money. Right now I am only using the +18v supply of the + rail and both coupling caps, and 12V for the heater from a bench supply at this point, realizing that the 5V supply doesn't have enough current to handle the heaters in parallel at this point (well, it does but the regulator nearly shuts down).

It looks as if using the entire 36V and no input capacitor will "force" the tube to bias at 0V and allow a much greater swing than I have currently. Since I am using it before any op-amps at this point (as the input stage), it looks as if I need to add another op-amp before it to hold the bias steady. No prob.

I have a few 12AU7's to choose from here (they are all used I think). What brands do you think are best? And since I am holding the bias at 0V now, would a 12AX7 or AT7 be better suited now? I ditched the original 12AX7 because it would "self-bias" (is this the correct term?) too close to 0V using coupling capacitors and only 18V.

I have learned quite a bit thanks to you, so thank you very much for being patient with a tube virgin such as myself.

if u use the cathode follower at the input without a coupling cap, but with, say, 50 - 100 kOhm tied to signal ground, then this resitor acts as input impedance and centers the tube when idle. As soon as a signal is there, the cathode will follow that signal.

Please do NOT introduce another opamp after you worked so hard to get rid of that §$%&in' coupling cap :)

IMO, opamps hurt more sonically than coupling caps.
John Camiile

I met John a couple of times at Dennis Boyles. He is a local legend with area audio nerds. Dennis is way cool also. The last time I saw Dennis ( it has been a few years) he was living and working in a converted movie theater and had motorcycles parked inside and the largest collection of vintage tube equipment you have ever seen. It was the Disneyland of vacuum tube audio.

Thanks, I was doing some tests and the 12AU7 seems to perform best in this application so thats what I used.. BTW i ditched the 350 volts and simply used the combined rail voltage of of my preamp (36 volts). It really sounds quite transparent with even this setup. I honestly expected more "color" but I guess since there's no gain or clipping there won't be much.