Re: Modifying W Plugin to use 6DJ8 Tubes

Dave Wise


Production below S/N 7000 uses a pair of 8056 Nuvistor at the input.

At 7000 they are replaced by a dual JFET.

V114 is on the A (plus) side, V214 is on the B (minus) side.??

One section is a cathode follower that bootstraps the first-stage amp anode or drain,

the other section is the top half of the second-stage hybrid transistor-tube cascode amp.

V124 is current sinks on the first-stage cathodes or sources.


I'm still investigating powering 6DJ8's, but I also experimented with a couple of common 12V tubes.

First I tried 12AT7. It's a no-go. The cascode top's grid bias is Q154's Vce. 12AT7 is only 0.5V, so Q154 doesn't work.

Then I tried 12AU7.

Like 8416, it's about 2V and Q154 is happy.

12AU7 works! Albeit with reduced 3dB bandwidth.

I see 22MHz with 12AU7, and 20ns rise time. Compare with the spec: 26MHz and 14ns.

I see a flat top on the square wave. I didn't bother adjusting amplifier HF comp. Maybe I can speed it up.

I see 33MHz with 8416.

Common-mode rejection spec is 500:1 at 500kHz; both tubes make 1000:1.

(Just for fun I tried 20MHz. Atrocious with either tube. Don't throw out your Type G.)

It's a simple mod. Unless you're a purist or a stickler for specs, I say go for it.

To modify your Type W to use 12AU7:

1) Cut off XV214 pin 9 solder tail. For assembly convenience Tek ran a component lead through it.

This 8416 pin is internal shield but for 12AU7 it's the heater center tap and must be disconnected.

To reach the solder tail to cut it, I had to lift the grounded terminal next to the socket, with a skinny right-angle #1 Phillips.

This step is backward compatible.

2) Add 1200 ohms 2W across the V114-V214-V124 heater string. I hooked one end to the chassis end of R290 and, with a wire, extended the other to XV124 pin 4 . To pass shake test, add a tie point at the flying end of the resistor.

3) Adjust DC Level. I hit spec with R280 at minimum R, i.e. full CW. (Same as 8416.)

4) Adjust amplifier HF comp. I have not done this.

Dave Wise

My signal sources were HP 8640B and HP 8601A. Using an HP 436A/8482A power meter I confirmed less than 0.5dB variation from 1MHz to 35MHz.

From: <> on behalf of Dave Wise via <>
Sent: Tuesday, February 23, 2021 9:28 AM
Subject: Re: [TekScopes] Modifying W Plugin to use 6DJ8 Tubes

In post #108873 (<>), David Hess suggested wiring 6DJ8's in parallel and running them off 6.3VAC from pin 13-14.

Unfortunately, that's out. Pin 13-14 is elevated to +100V and that will exceed the heater to cathode voltage rating on V124.

To recap, today I see exactly two approaches.

One, reverse a transformer. Unless you get lucky and find one with 6V and 18V windings like the 1A4's 120-0481-00, you have to wind your own, or put up with wire sized for 110V, i.e., at least 20VA even though the load is less than 4W.

Two, run a switching power supply off +225V. I will investigate this if the 12AT7 sub fails.

Join to automatically receive all group messages.