Using alternative grid for VariMu control (G3, G4, G5?)

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

rafafredd

Well-known member
Joined
Jun 3, 2004
Messages
2,409
Location
Rio, Brazil
There should be some advantages to use an alternative GRID for ControlVoltage in a VariMu compressor, like not needing a center tap on the input transformer secondary, sidechain not interacting with the transformer winding capacitances and resistances, and maybe even trying a transformerless VariMu.

Has anyone ever tried to control gain of a Pentode (or maybe Hexode / Heptode) by running the CV to G3 (or G4, G5???). Maybe even G2 with a buffer, I don't know.

If you did this, what were your findings?
Did it work well?
How does it impact tube performance?
Did you set G3 to 0v with signal grid (G1) and aplied negative voltage to gain reduct or did you go from a positive voltage like 18v and applied control voltage down to zero?

I don't really understand how the tube will work with G3 (or other grids, like in a pentagrid) in positive and negative potential in relation to G1, so I'm really afraid to just try it and burn some tubes, or worst, blow some big caps or starting a fire.

So... before I start experimenting with high voltage setups, any hints about how the effects of CV on these non traditional grids will bend the tube curves?

As always, thanks on advance for any inputs and discussion.
 
Using the pentode suppressor grid for control voltage works but has some problems. For regular pentodes you need a lot of control voltage for g3, much more than you would need for g1. Large audio signals at control grid will clip easily when control voltage is only at g3. The cathode current will stay fairly constant with g3 control, as the plate current decreases the screen grid current rises. Constant cathode current can be an advantage. Check out " Single-ended surgeless volume expander" from RDH4. Same idea can be used for compressor. I have tried it. It thumps unless you find perfect match of tubes and has high 2nd harmonic distortion, which can be good for sound effect.

There are also dual control pentodes where the suppressor grid is meant to be used as control element. With dual control pentodes you will need much less control voltage for g3 compared to regular pentodes.
https://bms.isjtr.ro/sheets/093/6/6AS6.pdf

In a tube compressor you are reducing the plate current to reduce transconductance. If you are not changing the voltage between control grid and cathode, it's hard to keep distortion at acceptable levels unless you know your input signal is fairly small and wont have large peaks. For best results you will need to have the control voltage on g1.

The easiest way to get rid of input transformer is to use solid state sidechain and take the control voltage to cathode. If solid state is evil maybe experiment with circuit I have attached.

 

Attachments

  • comp2.jpg
    comp2.jpg
    1.1 MB · Views: 41
The tubes used in vari-mu compressors have a spiral control grid the pitch of which varies. This is what makes the gain of the tube variable over such a wide range. AFAIK, the other grids in those tubes do not have the variable pitch so they would be less effective for controlling gain.

Cheers

Ian
 
Gates SA-39 using 1612/6L7’s feeds the same control to 2 grids. Probably true in the WE and Altec, maybe a Presto also.  I believe RDH4 calls it the lowest distortion vari-mu tube. Not many used it. 
 
Thanks EmRR. Gotta check those circuits...

Well, seems like driving the cathode with more CV current is a good solution.
 
rafafredd said:
Thanks EmRR. Gotta check those circuits...

Well, seems like driving the cathode with more CV current is a good solution.

In what way is that better than driving the grids?

Cheers

Ian
 
EmRR said:
Gates SA-39 using 1612/6L7’s feeds the same control to 2 grids. Probably true in the WE and Altec, maybe a Presto also.  I believe RDH4 calls it the lowest distortion vari-mu tube. Not many used it.

That seems to be the way 6L7 was used in limiters. Audio signal to g1 and control voltage to g1 (remote cut-off grid) and g3 (sharp cut-off grid). Advantage there looks to be that you need much less control voltage compared to remote cut-off pentodes.
 
ruffrecords said:
In what way is that better than driving the grids?

Cheers

Ian

Like I said in the original post:

"like not needing a center tap on the input transformer secondary, sidechain not interacting with the transformer winding capacitances and resistances, and maybe even trying a transformerless VariMu."

Those are some of the goals.
 
rafafredd said:
Like I said in the original post:

"like not needing a center tap on the input transformer secondary, sidechain not interacting with the transformer winding capacitances and resistances, and maybe even trying a transformerless VariMu."

Those are some of the goals.
OK, understood. It will be interesting to see what you come up with.

Cheers

Ian
 
Here's one to experiment with
https://patents.google.com/patent/US2766331?oq=patent:2766331

The patent says:
Maximum compression ratio-------- Greater than 10:1.
Maximum input signal (no compression) 1 volt.
Maximum input signal (10:1 compression 3 volts.
Gain at Zero compression---------- 0.9.
D. C. level shift at plate (maximum). 2 volts.
D.C. bias change for 10:1 ratio------ 0 to -6.5 volts.

Input is screen grid and control voltage is brought to control grid and suppressor grid. DC voltage shift at plate is low, but it's large at the input (screen grid).

Somebody tried it couple years ago and it seems to work, but it seems like he left out resistor R3 from the patent drawing.
https://www.diyaudio.com/forums/instruments-and-amps/310677-birkemeier-audio-compressor.html
 
From that thread: No free lunch - it seems that circuit gives you

Problem is can't get distortion below 2%. Interesting it remains fairly constant across the compression range.

and the solution:
..unless you drive the control grids at low impedance

Then you're back at where you could as well just drive G1 as we always did..

/Jakob E.
 
I don't know Jakob, I still think there's advantages on having CV and audio input on separate points.

Now I had another idea.

Feeding audio to g3 and CV to G1. That sounds like a rather good one to try out...

Still, driving CV to the cathodes or g2 with a lower impedance wouldn't cost more than a single transistor and a resistor.

I'll start experimenting.
 
Problem with feeding audio to pentode g3 is that the suppressor grid has very little effect on plate or screen currents. You would need huge input signals to get any output from plate. Unless you use dual control pentodes, but then you might as well feed audio to g1 and CV to g3. If you do enough experimenting you'll probably come to the conclusion that for best results you need to take audio and control voltage to control grid or positive control voltage to cathode.

Some years ago when I was playing around with g3 voltages I did notice that it could be used for some nice effects. I don't remember the exact circuit but I was able to make a circuit that produced only 2nd harmonic distortion up to 5% before any other harmonics showed up. When music was played through the circuit it made everything sound better.
 
Thanks Heikki.

So audio on g3 sounds unfeasible unless I use another tube stage in front of this, which is kinda pain.

I don't know if g2 would be the same, but I might also just try it. I could also try a really steep input transformer like 600:60-80k or something, but I also don't know if I would need more current drive to push g2 and g3, which would make this idea impossible. Seems to me like they would be much lower impedance then g1, right? Yet, something like 60k seems alright...

Well, in the end I might just stick to cathode drive, or use the beam deflection tube trick as discussed here some time ago.
 
Screen grid is low impedance, often around 20k depending on operating conditions. Suppressor grid is high impedance and probably draws no current until the voltage approaches screen grid voltage.

With 6SK7 pentodes I have tried takin the control voltage to both g1 and g3 and audio to g1. It does have a small effect on distortion. At some signal and compression levels the distortion is lower compared to control voltage only at g1, but at some levels it's higher. I didn't come to any conclusion if there was any advantage  taking the control voltage to both grids.
 
Maybe using the screen G2 as audio input as it's adjacent to G1. Who knows? Lots of things to try...

But then it wouldn't be really working as a pentode without G2 grounded or at cathode potential, would it???
 

Latest posts

Back
Top