Pentode pre/compressor (warning nube attempt at circuit drafting) --suppressor grid

[Blue Jinn]

You might want a small cap from screen grid to cathode on the 6GY6 or whatever to hold the Vg2 constant. The high value screen grid resistor is commonly used to drop the voltage from B+, but it doesn't hold Vg2 constant against current change unless the signal is really small.

If there is a resistor from B+ to screen would that capacitor bypass the resistor and go to cathode or to ground? Also, from what I've read so far, it seems a regulated screen voltage from the PSU is preferred to a dropping resistor off the B+. The latter is easier though, and I'm not seeing the difference in something like this.

 

[smoking-amp]

A cap from screen to cathode would maintain the most gain. Going to ground would provide a little neg. feedback ( cathode degeneration from resistance there). A regulated screen supply is good for output stages due to the larger current draw and for idle current stability with fixed (regulated) bias. With cathode resistor bias on this small signal stage, the resistor from B+ should work fine. You want the voltages to all track to keep idle current stable.

 

[Blue Jinn]

A cap from screen to cathode would maintain the most gain. Going to ground would provide a little neg. feedback ( cathode degeneration from resistance there). A regulated screen supply is good for output stages due to the larger current draw and for idle current stability with fixed (regulated) bias. With cathode resistor bias on this small signal stage, the resistor from B+ should work fine. You want the voltages to all track to keep idle current stable.

Do you think this is worth breadboarding or is it not worth the effort, as too narrow a compression ratio to be useful? This really is my first foray into any kind of design, and is new territory for me. One of PRR's criticism was how much gain reduction do you want? And to quote

IMHO the 6BA6 is not the ideal audio squasher; that has not been designed yet (and never will be).

However, IIRC the Dolby A is only a 2:1 ratio and there is a mod to the CAT-22 card to use it to good effect. (But eliminates some lower bands to get the desired effect) Although then you have the noise/distortion problem with this one.

Don't see where I need a whole lot of gain for a compressor in this instance, as it would likely be used from a line source or maybe from a guitar into either an amp or a line in on a board. Also, wouldn't lower gain translate into lower noise/distortion? I also need to be able to make up the gain reduction at that 12AU7. Although changing the position of the cap would not be a lot of effort in a prototype.

There were some other thoughts on a couple of other forums about better loading the 12AU7 and in the process increasing the gain in the side chain, with a higher turns transformer. A 1:4 there I could in theory recycle backwards as an output transformer in another McTube or NYDave type preamp.

I plan to get some tube prototype boards off of Ebay and changing out the configuration shouldn't be too big a hassle. Using fixed resistors to start out in the attack/decay network on the side chain, there doesn't seem to be a whole lot of parts to this. Edcor xfrms are only about $4 more a piece than the the car isolations used in the PRR, and they would come with a center tap. The Edcors are only $10 or so. If I use 1:1's at the input and the makeup gain interstage, I can save them for the D-AOC (which is a much more ambitious project) if this doesn't work out. The tubes are pretty cheap. PSU isn't a problem if I use AC on the heaters to start with.

 

[vinyvamos]

Very interesting thread Blue Jinn! Have you made any progress/discoveries since your last post?

I too have been experimenting recently with a pentode and trying to control the gain using the Suppressor grid. I started off building a tube pre-amp within an existing chassis which has EF50 pentodes and a pair of EA50 diode tubes (Circa 1940s!). It started out as an overdrive/clipper unit but am now trying to control the gain of the EF50 using negative DC from a charge-pump circuit fed from a 12AT7. All rough experimentation really, and good fun . I am however quite sceptical of the linearity of gain control via g3, and am sceptical of distortion figures, although if my unit is still an overdrive with the bonus of compression a little distortion aint gonna hurt ...

Here is an interesting investigation/write-up on g3 gain control of the EF50 pentode. It looks quite linear based on the 2V steps in Joe's scope captures:

http://www.radiomuseum.org/forum/ef50_suppressor_action.html

 

[smoking-amp]

Varying g3 voltage mainly collapses/expands the plate curves on the knee end or low plate voltage side. So for a signal introduced on g1, it will see compression/expansion control toward the knee end, but little effect toward the high plate V end. Causing considerable 2nd harmonic (and higher) distortion in SE mode.

The EF50 curves shown on Joe's site with g3 stepped, look evenly spaced but they are showing the effect of g3 signal for g1 = 0V only, not the linearity of g1 signal gain. What is needed is a set of the usual plate characteristics, each with stepped g1, and a new graph for each g3 voltage. Then the non-linearity would be apparent as just the left side collapses versus g3.

But you could put your signal on g3, and then use g1 as the gain control, giving a non-linear gain control function. ( like a log pot say)

There is a set of graphs shown there for g2 voltage changes between graphs (g1 stepped on each) of the EF50, and that set does show nice linearity (well, actually 3/2 power). So g2 control would make for a reasonable gain control of g1 signal.

The 6LE8 tube was designed specifically for differential g3 signal insertion, and has some datasheet graphs quite similar to the EF50 ( still not linear gain control by g3, but g1 can be used as the semi-linear gain control of g3 signal):

http://frank.pocnet.net/sheets/135/6/6LE8.pdf

For real linear gain control, the Beam Deflection Tubes (BDT) (or sheet-beam tubes) are the go to device. 6JH8, 6ME8

http://frank.pocnet.net/sheets/093/6/6JH8.pdf

http://frank.pocnet.net/sheets/049/6/6ME8.pdf

 

[Blue Jinn]

Very interesting thread Blue Jinn! Have you made any progress/discoveries since your last post?

I too have been experimenting recently with a pentode and trying to control the gain using the Suppressor grid. I started off building a tube pre-amp within an existing chassis which has EF50 pentodes and a pair of EA50 diode tubes (Circa 1940s!). It started out as an overdrive/clipper unit but am now trying to control the gain of the EF50 using negative DC from a charge-pump circuit fed from a 12AT7. All rough experimentation really, and good fun . I am however quite sceptical of the linearity of gain control via g3, and am sceptical of distortion figures, although if my unit is still an overdrive with the bonus of compression a little distortion aint gonna hurt ...

Here is an interesting investigation/write-up on g3 gain control of the EF50 pentode. It looks quite linear based on the 2V steps in Joe's scope captures:

EF50 suppressor action

Right now it's still on the drawing board. I also need to learn a bit more about pentodes. I'm not sure, but to my limited knowlede don't see why not, you can tie the screen to plate, but keep the suppressor separate. This sorta does that, but the GDIY folks thought the 6BA6 arrangement was odd:

An externally hosted image should be here but it was not working when we last tested it.

The folks at GroupDIY gave it a real quick pass, didn't see any problem using the suppressor; one poster at HomerRecording.com thought the side chain was a little weak.

It was designed to work with ham radio, which perhaps wasn't intended to have a lot of gain reduction. Some additional thoughts:

Use a 12DW7 instead of a 12AU7;

Use a step down after the pentode to the "12AU7" section to keep the overall gain lower;

Use the "12AX7" and a step up to drive the 6AL5 harder

 

[Blue Jinn]

@smokin-amp

Our posts crossed, thank you as well. More to think about. I've read other threads about using g3 as in input rather than g1.

using a pentode either way seems (and correct me if I'm wrong) to have the advantage of not introducing dc into the signal path, as you would putting the cv on the grid of a triode?

 

[Blue Jinn]

Another quick thoughts: using a fixed screen supply?

Wouldn't that help somewhat with distortion?

 

[Blue Jinn]

Took a quick look at the EF50 pages. It looks like an EF80/6BX6 would be equiv, with a more standard base. Or a 6EJ7 close enough. Both are US$5.00 at tubedepot.

 

[vinyvamos]

smoking-amp: Thanks for your detailed reply . I'm learning quick with all this experimenting. I might try signal on G3 then and alter G1 voltage for gain control. Will see then how non-linear it is. If I don't get usable results I'll just leave out the comp/limiting option and just leave it as a pre-amp/overdrive unit .

Bluejinn: The whole reason I'm using the EF50's with their strange bases is cos it's all in the chassis already . Otherwise I would go with something a lot more common and less archaic...

Seeing as your circuit was only designed for HAM radio use I really don't think it will be "hifi" and without distortion (same with mine probably!), but there's no harm in trying it!

 

[Blue Jinn]

The guys at GroupDIY, didn't see (or didn't mention) distortion as a factor but thump in the lower frequencies, and suggested push-pull to compensate for that. (The ham radio would be limited to voice freq, bass "thump" wouldn't be a problem.)

I have been taking a look at the links at radiomuseum.org. I admit, I have just figured out triodes, so the curves for pentodes I'm still learning, but I *think* I get the non-linearity, between the valve-wizard site and some other documents I have on pentodes.Also, as for distortion, if the gain reduction were kept moderate (like the 12AU7 PRR "vari-mu") could it still be kept in the linear range of the curves?

I am going to spend some time looking at that other info.

 

[Blue Jinn]

@vinyvamos: understand with the EF50; good find to have something like that to experiment. A guitar pre/overdrive with a bit of compression may not be such a bad thing.

The original ARRL side chain has always intrigued me. I've another rough sketch, with a 12AU7 for the make up gain and a 12AX7 driving the side chain. (12DW7)

There is also this schematic, which seems a similar idea, using the suppressor for gain control in a single ended configuation.

 

[Merlinb]

Suppressor-grid compression by Amos and Johnsone:
https://dl.dropboxusercontent.com/u/57831278/1951_Suppressor_Compression_AGC_Amos.pdf

 

[Blue Jinn]

The discussion about 3/4 the way on page 415 seems to validate the approach. A poster at homerecording.com made the point about the amplification before the 6AL5.

 

[asfi]

necroposting, sorry.....

Unsurprisingly, by now the link has gone stale. Is it possible to re-post the article?

 

[jfetter]

Yes, please put up the pdf here so is no longer lost.
i too would like to see it.

thanks

 

[flysig]

I, too, am interested if the article is still available somewhere. A search has not found it for me.

Fred Nachbaur has an interesting circuit on his page which is worth looking at. He taps off of the speaker output to charge up a passive circuit which is then connected to the control grid of the first preamp valve.

The preamp valve is a remote cutoff pentode 6BJ6. Using diodes and capacitors his compressor circuit captures peak voltages which then feed back to the grid circuit to adjust the gain of the pentode.

Here's the compressor circuit. "Li'l 4x4" Compressor Option Since it taps off of his speaker output, it would need to be modified to be suitable for whatever amp/preamp it is being used in. I am modifying it to feed back from the first stage of the preamp itself, to provide compression prior to a second stage which can be driven to distortion. But the theory should still apply.

Here is his preamp section. Note that it is the older version with a different valve. He describes in the compressor section what changes to make to component values. "Li'l 4x4" Tube Amplifier - How It Works

Here's a link to his full schematic except for the compressor. Amp, preamp, and power supply. http://www.dogstar.dantimax.dk/tubestuf/graphics/4x4large.gif

Note that the power amp is unconventional! +- 140 volt power supplies, with push-pull throughout 3 stages! The preamp is run off of just +140V and ground, without a negative supply.

 

[jazbo8]

The article is available at American Radio History: Documents from the History of Radio TV and FM
broadcasting and Electronics
Look for Wireless World 1951, Oct. issue.

 

[flysig]

Thanks jasbo.

Here's the direct link for those interested: http://www.americanradiohistory.com/Archive-Wireless-World/50s/Wireless-World-1951-10.pdf

The article begins on page 33 of the pdf as it displays on my browser. The magazine shows it as page 415.

 

[jfetter]

Here are selected pages. Thanks for the link.

Interesting in that implies sharp cut-off pentodes may be better for compressors.

-