Stuart Smith’s bias comparison tester…

[TomCarlos]

MEF Members….

UPDATE - Just updated the EL34 mention below.

I came across a YouTube video where Stuart Smith of Berkshire Guitar Amplifier Repairs (in the UK) built a bias matching device. (Perhaps Mick knows him!!).

Youtube Video link. The video portion to watch starts at the 24:25 point.

Actually, there are two devices used together. As Stuart mentions in the video, the device sets a predetermined current flow for one tube, and then you simply drop in a second tube to see how closely it matches in terms of current flow.

The first piece of this setup is a high voltage power supply. I have a Heathkit SP-2717a for this. The second device is a jig where he drops in tubes into sockets. As you can see, there are three tube sockets. On the device, Stuart has notes for four different tubes – (1) 6L6, (2) 6V6, (3) EL34, and (4) EL84. In addition, he also shows target voltages.

In the video comments, you will see some exchange I have with Stuart, wondering how he is testing the tubes (circuit wise).

For the 6L6, the plate is going directly to the HT B+ supply. The screen goes to HT B+ using a 470 ohm 2 watt resistor. The grid goes directly to the Bias supply (negative voltage). I put a 1 ohm, ½ watt resistor on the cathode to measure voltage (which in effect is current).

For the 6L6 and 6V6, the first socket and components work great. I have replicated this on my board.

As you can see from his sticky notes, his settings are as follows:

6L6: HT = 360, Bias = -38

6V6: HT = 330, Bias = -25

EL84: HT = 350, Bias = -16

EL34: HT = 330, Bias = -28

And what about the EL34? That is an 8 pin tube, a slightly different pinout, Pins 1 and 8 are connected, so it needs a different socket. But is there a need for different components?

My other question is, how does this change when testing the EL84? I see a separate socket for the EL84 but I am wondering if it uses the same circuit as described above? Perhaps it does? I know that most EL84 single ended applications use Cathode biasing. And in that configuration, I am believing the Plate voltage is higher than the screen? But I am sure you can connect it in a fixed bias mode for this test. So I am wondering about the plate and screen connections and components.

Any ideas on this would be appreciated !!

[Helmholtz]

I don't see a reason for additional components (for EL34s or EL84s).
You should always start with large negative bias and then turn down to desired current.
The 470R screen resistor makes sure the screen voltage is always a bit lower than the plate voltage.
I don't see a special reason for choosing these exact currents.
Always best to stay close to the designated operating conditions in the amp. I would opt for same cathode current rather than same dissipation. Cathode current is a real operating condition while dissipation is just a thermal thing.

This is a very basic power tube tester. Actually you could just use an amp as well.

[TomCarlos]

Thank you Helmholtz....

Circuit board is done and working. I finished wiring up the EL 84 socket 3. Socket 2 for the EL34 is similar to number 1 but pins 1 and 8 are shorted.

You are correct, best way to test is in the amp. But when you have a bunch of old tubes and want to see which ones match up, this is a handy test board.

Using new JJ matched tubes, I was able to establish the following bench mark readings. so I will now have something to work from when testing other tubes.

Socket 1, 6L6: HT = 350v, Bias = -38v, Current = 23 ma

Socket 1, 6V6: HT = 330v, Bias = -25v, Current = 35 ma

Socket 2, EL34: No yet tested

Socket 3, EL84: HT = 350, Bias = -16v, Current = 20 ma

[Helmholtz]

Is there a reason why the current with 6V6 is largest?

[TomCarlos]

Good catch...

I adjusted the -Bias on the 6V6 (-30v) to attain 23 ma.


Socket 1, 6L6: HT = 350v, Bias = -38v, Current = 23 ma

Socket 1, 6V6: HT = 330v, Bias = -30v, Current = 23 ma

Socket 2, EL34: No yet tested

Socket 3, EL84: HT = 350, Bias = -16v, Current = 20 ma

[Helmholtz]

I don't understand the different HT voltage either.

Limit for 6V6 is 350V, while only 300V for EL84. Often these limits are exceeded in amps.

[TomCarlos]

I was working from the arbitrary voltages that Stuart used in his video. But I see your point.

Thank you.

Revised targets -

Socket 1, 6L6: HT = 350v, Bias = -38v, Current = 23 ma

Socket 1, 6V6: HT = 350v, Bias = -31v, Current = 23 ma

Socket 2, EL34: No yet tested

Socket 3, EL84: HT = 300, Bias = -12v, Current = 23 ma