diyAudio (
-   Tubes / Valves (
-   -   938 power triode (

abj1 16th July 2006 05:08 AM

938 power triode
Hi All,
Anyone out there with more experience than I with these big power triodes got a pointer to some information about the 938 power triode? It definitely has 938 screened on the glass (not 838), only 1 grid and a big graphite anode. Made by United.


jlsem 16th July 2006 02:06 PM

It is the same as an 838. United Electronics used a "9" instead of an "8" for 800 series transmitting tubes. Same goes for the 966. If you have any 945's, I am interested.


Tubelab_com 16th July 2006 04:08 PM

I once built an amp using 838's. They are like 811A's on steroids in an 845 bottle. A2 mode is required, but they sounded good.

I used the same circuit as my 845 amp except that the 22K resistor that is connected from the bias pot to ground is now connected to the +150 supply, and the 1K resistor that connects from the bias pot to the 150 volt supply was changed to 22K 5 watt.

This circuit operates at high voltages (over 1000 volts) don't try this unless you are an experienced builder.

abj1 18th July 2006 04:20 AM

Woohoo! I had suspected that they may be 838's (and the 966A's might be 866A's). Fantastic. Thanks to all who replied.

I'm right in the middle of learning (by doing) how to handle A2 valves. Have a test mule set up with PP 46's running with positive grid bias (around 28mA each valve with a little under 350V filament to anode which needs about +10V bias (the 4 bottles are quite different ranging from 8 to 12V to give the same anode current) and the sound is intriguing to say the least. Love to be played loud! 20K anode to anode load so it stays in A2 all the time. Major challenge is to drive them so I've been putzing around with a cathode follower design (all I know about CF design can be scratched on an aspirin with a crowbar but I'm learning fast). One thing I have found is that pentode connected 12SG7 with a bias of -1.7 V and around 140V on the screen sounds the best I've found so far (slaughters all the triode arrangements I'd tried to now). I realise that this is not really a pentode connection as the plate voltage is referenced to anode rather than cathode which is swinging around a tad (perhaps +/-10V) but is close enough for an initial test. I've read in several places over the years that theoretically pentodes should make excellent CF's. Anybody interested in elaborating? Also I'd truly like to know what are the do's and don'ts of CF design - I'm not afraid of pentodes and appalling power efficiency but I'm not quite ready to embrace FET's yet (sorry TubeLab) but it may come to this. Perhaps this should be a new thread now that I've rattled on a bit.


Tubelab_com 19th July 2006 12:58 AM

I read some theory about how pentodes make the best CF's. I don't remember it all, but I do remember the fact they can get closer to a gain of 1.

I experimented with CF's before I went over to the dark side (mosfets). I found that in order to drive the grid out of a big transmitting tube you need a CF that can pump some current. To me that sounds like a sweep tube. Look for a tube that has a high peak current capability. I don't what is available in your part of the world, but I would look for a small horizontal output tube, or a vertical output pentode.

abj1 19th July 2006 03:22 AM

Thanks Tubelab - that's pretty much how it seems to me. Alas I don't have enough power supply herbs on the current chassis to run 4 big CF's (its push-pull) but what I have running at present is a good pointer to what should work and I suspect that the ability to handly man-sized peak currents without the grid (of the CF) going positive is the key.

I like full-time class A2 sound enough to push on with a single ended design (on a new chassis) with lots of power supply mojo (efficiency is not emportant...) using either 809's or 811's as outputs (saving the 838's er 938's till I have things ironed out a tad more) as I have a brace of both. With a bigger supply I can work with things like 6L6 or 6W6 or other highish gm pentodes/beamies for CF service with a lot more ability to handle peak current than the 12SG7's I'm using at present . To drive A2 triodes the limiting factor isn't voltage swing its the ability to handle current swings that I reckon is critical. Thoroughly enjoying the challenge of A2 - turns a lot of standard design decisions right on their heads! Right off the beaten track....

You have one of the more impressive websites around - thanks for sharing your work with the great unwashed.


jlsem 19th July 2006 04:08 AM


Look for a tube that has a high peak current capability. I don't what is available in your part of the world, but I would look for a small horizontal output tube, or a vertical output pentode.
For a high current pentode actually designed for audio, try this inexpensive number:

All times are GMT. The time now is 11:32 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2