Solution for scarce 838 tubes

[Opcom]

The 572B can be used for ham audio service if the 838's are too hard to find for a reasonable price. The 838 is an older graphite plate variable mu Triode (mu of 50, varies with drive)  and the 572B is a high mu Triode (mu=160). The plate current curves at 1000V, where 838's usually are operated with zero bias, is very close to the 572B. Driving should be compatible, and the grid current when the grid is driven positive is similar.

Attached is a chart comparing the two according to published curves.
The 838 is rated only 1250V (2KV peak when plate modulated) so the 572B would have no trouble with the voltage.
The 838 has very low distortion in class "B" push pull operation. Less than 572B's (see the curves)
There is an SV572 for audio service, if it is still made, for those who are more demanding.

This experiment came about because of an audio amp here that takes the rare and costly 838's, and having only three good ones, I tried a pair of 572's wired up with jumper leads on the work bench to see if it was a good long term solution. They sounded fine to me with some music on a dummy load with a tap for a hi-fi speaker, and the power was as expected so I think it is just fine for voice. There was a slight amount of crossover distortion but I thought it was not worth worrying about. It might be because of the low voltage. The regular 572B is usually to be run at 2000V. The idling current was lower on the 572Bs at this low-ish 1KV level, only 60mA for the pair instead of the 838's 110mA for the pair. The plate voltage run was 1000VDC. The driver was a pair of 6B4 triodes. I suppose 811A's, which are an older type that works like the 572B and has straighter curves, could also also be used if less power is wanted. Or use four in pp-par, because it is the current rating that is the cause of the dissipation requirement for that swap.

[WBear2GCR]

The Chinese are making decent 838s, fwiw.

What the heck amp do you have that takes 838s??

OUght to look down in this section a bit more often...

[Opcom]

It's the Norton Chimes amplifier. I thought it was posted but don't see it.

[WD5JKO]

   The 838 with a plate cap on it becomes an 805. Same tube with a loftier plate voltage ratings (1250 versus 1500).

I wonder how the 808 compares to those two sets of curves you plotted? The 808 is a medium mu triode as well. Some of the Single Ended audio guys run them in class A where they need positive grid bias to set the DC operating point. This is at a plate voltage of around 800 volts. All this for 10-15 watts of audio from 30-40 watts static plate dissipation.

The guy at tubelab designed "Power Drive" for driving big triodes. Here he uses a special triode as a driver, such as a 417A or even a 45, and uses a solid state CCS on the driver plate, and then couple the driver through a power mosfet source follower to the next stage big bottle triode. Look around that website...he uses it to drive an 833 with positive grid bias to set the operating point. 

http://tubelab.com/articles/circuits/power-drive/

Yes the 838 is a special tube.

Jim
Wd5JKO

[Opcom]

And the 3-500Z as well, up to the limit of grid dissipation, which is the limiting factor at reasonable plate voltages.

The Norton Chimes have some links here. I guess I have to fix the site due to my insouciance.

http://www.bunkerofdoom.com/00_QRZ/Norton-Chimes_4x838/nChimes.pdf
http://forums.qrz.com/showthread.php?421566-Norton-Speech-Amplifier-Driver

[W2PFY]

The Chinese are making decent 838s, fwiw.

Bear, any hint as to the manufacture or source for the 838 tubes. Found nothing on Giggle........

[WBear2GCR]

Terry, not sure, I have seen them in casual browsing from time to time. Maybe a check for 805s will yield the elusive 838?

Nothing on epay??

What do you use them for? One of your Out-of-round-dyke rigs?

                      _-_-