diyAudio (
-   Tubes / Valves (
-   -   Mono push-pull driver PCB (

pmillett 24th December 2011 08:02 PM

Mono push-pull driver PCB
1 Attachment(s)
Merry Christmas all!

I've been working on a new push-pull driver PCB. I mentioned it in the "Engineer's Amplifier" thread, but figured I should start a new thread for discussion, since it is a different project.

I wanted to come up with a driver that has some of the same characteristics as the Engineer's amp, but was a bit more flexible. Lots of requests for a similar amp that can use normal output tubes, for example. I also wanted something that has an extra gain stage, so NFB could be applied... partly for my own education, since I have not worked much with global NFB.

Like the engineer's amp, what I've come up with uses pentode drivers (either E80L or 6EJ7 or tubes with similar bases, or 12BY7 / 12HG7 / 12GN7) arranged as a diff amp with a CCS in the tail. A single triode (6J4/8532) gain stage sits in front, so you can apply gNFB to the cathode.

There are many more details on my web page: Push-pull driver board

Right now, I have a proto board built, running a pair of triode-connected KT88's. So far it looks pretty good.

At this point, before I build any more boards, I'd love to get some feedback. Anything I should change or add? Within reason, of course :)

One thing I already thought of was to add some jumpers so you could triode-connect the driver tubes.

If you have suggestions, please post them here, or shoot me an email.


GloBug 24th December 2011 08:12 PM

A triode switch would be nice. Lot's of people seem to have interest in being able to switch between the two.

Tubelab_com 25th December 2011 03:38 AM

I have a few different driver boards in various stages of completion. One that is currently breadboarded uses a pair of 9 pin pentodes in LTP for the driver stage and a dual triode in LTP for the input stage. I have stuffed several different tubes into it and also tried a few things that I'm sure you haven't...yet.

My observations:

Some 6EJ7's can be good TV jammers (VHF oscillations). Grounding pin 6 helps calm them down. Pin 6 can't be grounded on all tubes, so a short jumper is good.

6EH7's are considered useless for audio but work OK for LTP if you don't need a lot of drive voltage.

I have experimented with the screen bypass cap (and zener where used), and found that for most tubes connecting it to the cathodes generally works better than ground. If you think about it connecting it to ground applies some negative feedback to the screen in a manner similar to UL operation. The VR tube acts like a bypass cap, but you might want to add a .01 or so across it for noise control. Too big a cap may make them unstable.

If you really want to try something crazy, take the local feedback from the output tube plates to the driver screens. Since the screens invert the phase, you need to "crossover" the feedback resistors (take the feedback from the other output tube's plate). I first learned this on my red board and it can compensate for output tubes with mismatched Gm. It doesn't work with all tubes, and you need a resistor to ground to keep the screen voltage down with big plate voltage on the output tubes.

I tend to run things like mosfets and CCS's off of the negative voltage bias supply. A simple half wave rectifier doesn't work very well if you need 50 ma or more. Some builders will be in 50 Hz countries making matters worse. In the case of the typical Hammond transformer with the single tap, it's all that's available, so it's your call. If you have room on the next itteration 4 diodes and a jumper will allow both.

Mismatched output tubes are more common than matched ones. A seperate bias adjustment for each tube, or an offset adjustment like the red board is a worthwhile addition.

dgta 25th December 2011 04:14 AM

Suggestions: I would want

1. provision for CCS in the first tube's plate circuit (or not)
2. provision for 2 series components in the first tube's cathode so that one can use a LED and a resistor for NFB (or not)
3. If there is enough room, I would have dual patterns for the tubes so one can use noval or octal tubes (my preference).

pso 25th December 2011 06:01 PM

Plans to offer a PCB fot the power supply ?

pmillett 25th December 2011 09:16 PM

Some good ideas. Here is what I've implemented:

1. Jumpers to triode-connect (or pentode-connect) the driver tubes
2. (Optional) cap across the VR tube (0.01uF)
3. Changed bias rectifier to FW bridge (you can always jumper a diode for half-wave)
4. Added an output balance pot
5. Added an (optional) CCS in the plate of the input tube (replaces resistor)
6. Added an (optional) LED in series with the cathode bias resistor of the input tube
7. Added an (optional) RC from input tube plate to GND (additional FB compensation)
8. Added jumpers to connect pin 6 of the driver tubes directly to the GND plane

Sorry, an octal alternate would be very difficult - it would make the board bigger (I want to keep it compact), and there are not a lot of good octal pentode drivers (IMHO).

An updated schematic is here:

Keep the ideas coming!


pmillett 25th December 2011 09:18 PM


Originally Posted by pso (
Plans to offer a PCB fot the power supply ?

There really isn't anything to the power supply, so no PCB planned. I'll use a tube rectifier and CLC filter. I did design a "cap board", just a little PCB with two snapmount electrolytics, just because I'm tired of trying to find a good way to chassis mount them...


pso 26th December 2011 12:10 AM

This little PCB are already for sale ?

Tubelab_com 26th December 2011 03:13 AM


and there are not a lot of good octal pentode drivers (IMHO).
The drivers job is to take a small signal and amplify it to a level capable of cranking big power out of some insensitive tubes. You may need 300 volts of drive to run 845's in class A2 or AB2. To do this with a pentode requires a good bit of Gm. The usual 6SJ7 comes in around 1600 uMHOS. I had the best luck with a 6K6 (2300)or a 6V6 (4100).

There are several good 9 pin driver candidates with Gm as high as 25000.

bayermar 26th December 2011 05:27 PM

Hi Pete,

do you think it would be possible to apply plate-to-plate (aka Schade) feedback with this driver board?
It might be interesting to use this in a Baby Huey-sytle setting i.e. with KT88 output tubes. I am not sure if your driver could be used instead of the diff amp in the Baby Huey.
(link to Gingertube's circuit here )
This "Huey" might not be much of a "Baby" any more... ;)

Or is this beyond the scope/idea of the circuit (I don't want to go too far off-topic)?


All times are GMT. The time now is 02:55 AM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio