Electronic compressor limiter with valves

HWV · April 16, 2006, 02:12:00 PM

I've been working with vacuum tubes (again) after a few year hiatus, and was trolling through all of my links on tubes. I came upon this schematic: http://www.triodeel.com/comp1.jpg
from this page of interesting stuff: http://www.triodeel.com/studio.htm
I figured it looked easy enough & have it breadboarded. It runs very quiet; so quiet that no signal is passing through!

I have enough knowledge with tubes to scratch-build a simple class-a amp, but the compressor schemo is a bit different I am used to. Does anyone have any insight they would like to share?
Does the circuit need more imput than a guitar (or bass) can give?

(BTW, I'm running the B+ at 250v, and it looks like the 1st (bypassed) pot is 1k - might be 1meg?)

Thanks for any info!

-Hans

zachary vex · April 16, 2006, 02:53:21 PM

the audio doesn't pass through the tubes on the way to the output jack. it passes through the 470k resistor between the input and output jack, and is divided (and thereby "compressed") by the last tube stage's action as a variable resistance.

basically, the first tube stage acts as a cathode follower, the second acts as a rectifier(note that plate and grid are shorted), the third acts as a DC cathode follower, and the last stage is essentially a voltage-controlled resistance that is greater when the signal is quiet and smaller when the signal is loud.

there are factors to consider in this circuit: how loud does the signal have to be before compression action begins to take place? what is the best setting range for those two pots? (that one will have to be determined empirically... just adjust them a lot, but watch for the low resistance settings... magic smoke may escape!) what is the basic function of the two pots? it looks like the first pot might adjust the release time somewhat. the attack time seems fixed. the second pot seems to adjust the threshold, but i'm guessing at this point... there may be details i've ignored.

HWV · April 16, 2006, 03:07:40 PM

Cool - thanks for the info! I will keep working on the circuit & update when I have a bit of progress. So far, the magic smoke has stayed in the bottles.

I have redrawn the schematic to make it a little easier to read - here is the link:

http://furtive.us/images/compressor_limiter.pdf

It's quick & dirty, but readable.

Thanks again,

-Hans

puretube · April 16, 2006, 03:22:29 PM

I never trust a schematic on the web...

HWV · April 16, 2006, 03:55:47 PM

Quote from: puretube on April 16, 2006, 03:22:29 PM
I never trust a schematic on the web...

I hear ya...

I'm sure the text that references 'circuit 14' would help a lot.

But after some more messin' about, it does seem to work, after a fashion. It wants more than a straight guitar signal to overcome the 470k resistor (or some sort of a recovery gain stage) if you want a ~avg. amount of output. Larger input & output caps probably won't hurt, either. The pots, as Mr. Vex surmised, do control the threshold and the release(kinda).
I doubt I'll take the circuit any further than the breadboard, but I am glad that as a 'proof of concept' it works (and didn't let any magic smoke out)!

Thanks again for the help & info. Maybe someone else will take this further than the breadboard!

-Hans

Incubus · April 16, 2006, 04:42:16 PM

That pot value in the jpg looks to me like 1Meg, not 1K.

zachary vex · April 16, 2006, 05:31:20 PM

i agree... first pot is probably 1M.

puretube · April 16, 2006, 05:52:40 PM

Quote from: Incubus on April 16, 2006, 04:42:16 PM
That pot value in the jpg looks to me like 1Meg, not 1K.

Incubus wins, being first answering the

...

1 Meg.

no offense, HWV, but tube-diodes don`t get loaded with 1k/smoothed by 10nF;
(even when fed from a cathode follower)

hope you can correct that in time,

before your schem is going to enter the web-history like

the "MXR Hot Tubes"...

HWV · April 16, 2006, 07:01:44 PM

Quote from: puretube on April 16, 2006, 05:52:40 PM
Quote from: Incubus on April 16, 2006, 04:42:16 PM
That pot value in the jpg looks to me like 1Meg, not 1K.

1 Meg.
no offense, HWV, but tube-diodes don`t get loaded with 1k/smoothed by 10nF;
(even when fed from a cathode follower)

hope you can correct that in time,

before your schem is going to enter the web-history like

the "MXR Hot Tubes"...

Fixed!
Umm, the whole circuit works much better when you use the correct values, too. (DOH!)

-Hans

zachary vex · April 17, 2006, 04:34:51 AM

i used this very circuit diagram to build a compressor into a microphone preamp i built years ago when i had a recording studio... evidence: vocal track (sorry, me singing) on this song. never mind the rest of the processing... the vocal track has this particular vacuum tube compressor on it during the verses. band: fauna. label: twin\tone. year: 1993. lineup: zachary vex guitars/vocals, peter von grossmann bass, todd lipelt drums/vocals.

http://www.twintonedigital.com/mp3/89266-04.mp3

Paul Perry (Frostwave) · April 17, 2006, 05:05:58 AM

Anyone hazard a guess as to how this works? Is the second half of the first tube acting as a diode? Because, there is no + on the anode!

puretube · April 17, 2006, 05:09:38 AM

yes... see reply#7

puretube · April 17, 2006, 05:23:14 AM

2 chains: "envelope follower" to the left, comprised of cathode-follower (1st tube, 1st half) passing signal through 10µ cap to rectifier (1st tube, 2nd half).
This CV is buffered by a 2nd cathode-follower (2nd tube, 1st half),
whose output varies the bias of "variable resistor"-wired 2nd tube, 2nd half at that one`s grid.
The (varying) R_i of this triode is part of signalvoltage divider 470k/plate-cathode.
So there is signal-loss when compressing.

(that`s just my view).

AARgh, sometimes I hate sitting @ this non-sound PC...

George Giblet · April 17, 2006, 06:36:09 AM

Does anyone else think:
- the 10uF electro cap is around the wrong way?
- the compression control is a little dodgy?
- the forward biased grid for the voltage controlled resistor has a poor range of control?

puretube · April 17, 2006, 06:57:18 AM

Quote from: George Giblet on April 17, 2006, 06:36:09 AM
Does anyone else think:
- the 10uF electro cap is around the wrong way?

maybe it`s an *sterix...

puretube · April 17, 2006, 10:47:40 AM

ooops - due to strange forum behaviour/unavailability(easteregging hide&seek),
I couldn`t see that some stuff I replied had already been said before...

fikri · April 17, 2006, 01:10:55 PM

I'm keeping the schematic for years but havent try it yet. I'm glad you've done it succesfully

Any experience with tube mic preamp ? That is what i need now ! maybe a combination between altec mic preamp and this compressor would be charming !

zachary vex · April 17, 2006, 01:28:57 PM

Quote from: puretube on April 17, 2006, 10:47:40 AM
ooops - due to strange forum behaviour/unavailability(easteregging hide&seek),
I couldn`t see that some stuff I replied had already been said before...

i'm glad to see that the tube-meister is in basic agreement with my guesses. 8^) i neglected to mention that the audio circuit is basically passive and lossy. you can hear from the song i posted that the vocal during the verses is suffering from what sounds like intermodulation distortion, which i believe this circuit generates in abundance, which is a very good tool for a recording engineer to have around.

puretube · April 17, 2006, 01:44:33 PM

sure, Z.
(PT lights up cigar, before going italian easter dining...)

to me, that 0.01µF smoothing the diode`s "CV", is a bit small.

(haven`t found the text to that circuit, but I guess that`s why it carries
an asterisk, too...)

zachary vex · April 17, 2006, 09:08:33 PM

hence the intermodulation, eh? 8^)

News:

Electronic compressor limiter with valves

George Giblet