Space-charge tetrodes in Inverted Mode

[zesak]

Is it possible to connect space-charge tetrode in inverted mode ?

This would be for possible SE parallel OTL mini amp (inductive or resistive load, and, say 16 or 40 ohm speaker).

Any existing and/or working designs like that out there ?
(I am aware of several space-charge mini power amp circuits but have not seen any space-charge tetrode inverted mode designs)

 

[wa2ise]

If you were going to use tubes like 12K5, there's a grid 1 that is connected to 12VDC and it draws 85ma. Which what gets enough electrons loose off the cathode to do something with. And the plate current is about 8ma. And the data sheets say the tube can make 35mW of audio into 800 ohms on the plate. Might work for a headphone amp? Doing an OTL circuit via the cathode may be hard, give that there's the grid 1 current to deal with.

 

[Tubelab_com]

Inverted mode is an unusual configuration pioneered by Steve Bench. Cathode is grounded. Grid is the output, and is tied to B+ through the
OPT. The plate becomes the control "grid". Mu is the recriprocal of the tube normal Mu. So a 6SN7 with a Mu of 20 would have an inverted Mu of 0.05. So obviously you want to use a low mu tube. I never tried space charge tubes. My best success came with 6AS7's. About 1 or 2 watts per tube is possible. So is melting the grid.

 

[Paul Scearce]

Now you have my curiosity. Perhaps a tetrode could be used with g1 tied to the cathode, g2 connected as the anode, and the plate as the control element?

 

[TheGimp]

Sounds like a job for a tube with a hefty grid that can handle some drive.

 

[zesak]

Actually, the pioneer of inverted mode would be FREDERICK EMMONS TERMAN, allthough S.Bench has done a lot of great follow up work in this and other areas as well.

I just downloaded his article on Inverted Tube Operation from IEEExplore :

Proccedings of IRE, 1928, "THE INVERTED VACUUM TUBE, A VOLTAGE-REDUCING POWER AMPLIFIER"

I could not find anywhere articles referenced by the "Bible" (Langford-smith) :
Brian, W.S. "Experimental Output Tetrode", Electronics 20.8, August 1947 121.
and
Pickering, N. "Space-charge tetrode Amplifiers" Electronics 21.3, March 1948 96.

Any idea where to find these references ?

Thanks.

 

[zesak]

I was thinking g1 positive with respect to cathode, otherwise connected .as Paul described

 

[zesak]

wa2ise,
not true for inverted mode:

Since output is taken from the grid, then per Terman:

"The dynamic grid resistance of the inverted vacuum tube is
much lower than the dynamic plate resistance of the same tube
used in the usual manner."

 

[zesak]

I was also thinking about using this beefed up space-charge grid (g1), instead of g2, as output, giving Zout around 150 Ohm.

Would this arrangement have a chance to work ? ( with g2 tied to plate as control grid )

 

[Tubelab_com]

I vaguely remember Steve Bench crediting someone else for discovering inverted mode. Unfortunately Bench's web site is gone now, and I don't remember all the details.

I experimented with inverted mode when I read about it about 5 or 6 years ago. Most of my experiments used triodes back then. I have hundreds of used 6AS7's and some 600 ohm SE OPT's so this is what I played with the most. I tried some triode wired sweep tubes(G2 and G3 connected to the plate), but the 6AS7 worked the best. The sweep tubes have too much Mu and required more drive than I had.

If I were to try this today I might get different results. I now know how to get hundreds of volts of clean drive, and I can work with lower loads than 600 ohms.

Any time you venture off the beaten path, you must be prepared to experiment, and in this case, maybe melt a few tubes. I don't see why it wouldn't work, but the only way to find out how well it works is......

 

[zesak]

That is the thing. I do not feel like melting anything but I do not have the max grid power allowed info to limit the current if needed.

Is it inherently safe from melting when all tube terminals operate with maximum of 24V on them, or do I still need to insert current limiting resistor ?

 

[zesak]

Before I start melting grids I decided to try to simulate the circuit with LTspice using 12K5. I found a model that might work here on diyaudio but I do not yet know how to associate drawing of a tetrode with its correct model so that I do not get a message "model not found" or "library not found" when I try to run the simulation. Does anyone know an easiest step-by-step method to do that, or can you provide a link that shows the steps required in LTspice ? Thank you so much !