Edit: Direct-coupled valve compressor

[Winston OBoogie]

"I have built negative 300V supplies just so I could use a dumb resistor."[/color]

Me too.  ALA Philbrick K2's.  A few -V and a piece of sand is easier though.

Well, I just fired up a dumb resistor from a -tve 200 (all I have on hand) and, at the same current as my active CCS, 3rd is down by about 4dB on a 7N7 diff. amp.  No more time today to fiddle with it...

 

[Winston OBoogie]

No further tests on the "PRR current mirror to C.M. voltage" idea but the latest fiddling has been: 

I shifted the transformer I was using purely as a balanced anode choke for the VCA to being a balanced grid choke for the output valve
The VCA is back to resistor loads and I still have a current sink on the output 7N7's.

Since DCR of the choke is much lower than my usual grid R, the DC time constant is much faster but AC impedance for signal is high. 
I do still see a shift when I probe the test jig (Hi voltage DC burst to C's to chokes) but it recovers so much faster than the C to grid R that I don't hear it when I have it in a unit.  I'm not in Fairchild territory as far as attact time anyway so ?
I suspect I'll possibly need to damp the chokes but it looks pretty good for a quick lash-up

F.W.I.W. I'm using a pretty inexpensive centre-tapped output transformer primary as the chokes.  DCR is about 187R per leg and I can't easily measure the inductance with the stuff I have on hand but would guess it as being about 350H per leg.

A next test for me will be using a cheap and cheerful hi-impedance 1:1 with both pri and sec windings wired together in series as the centre-tapped grid chokes.  No doubt I will see some imbalances since pri and sec are not exactly the same but we'll see what happens...

Ciao

 

[edanderson]

you might try the XSM series from edcor; i believe they make a 15k to 15k model where the primary and secondary are wound bifilar, which should result in good balance between the two "halves" if you put them in series.  if you could find a mains iso transformer wound on a split bobbin, that might also serve in a pinch, something like the VPS230-110, though i doubt the inductance (and thus impedance) is very high.

if you end up needing more than a few, let me know... it shouldn't be too hard to knock together some CT chokes with a split bobbin for good balance and less capacitive interaction between the two halves.  i'd like to get some of those 610 outputs from you as well, but your PM has been all jammed up (with good wishes, no doubt).

ed

 

[Winston OBoogie]

Hi Ed,
Well, upon pondering this a bit more, I went back to a scheme that was direct coupled but kept everything nice and simple:

VCA is again choke loaded and fed to the diff. amp.  I have two discrete 200V H.T. supplies that are stacked so that the reference for the supply that feeds the diff. amp sits on the +V of the supply for the VCA. 

I'm now onto playing with more ancient valves as alternatives to the 6ES8 VCA. 

I may take you up on finding a good balanced anode choke though when it gets time to cost reduce this thing.
I'm currently using the same transformer 3 times -  It's flipped for the input, it's used as a load for the VCA (secondary currently either resistor loaded, or, used to feed the side amp) and it's used conventionally as the output. 

610 outputs:  I ordered 12 about a week ago.  Not sure when I'll get them back but, if they turn out well, I'll order more for sure.

Cheers 

 

[Winston OBoogie]

Here's the basic idea for the Audio Path I have going right now.
There are details missing from the schem. and no side chain, time constants, meter etc shown
but I thought I'd shove something up for now.
Side chain will, probably, be quite simple (Hi-V MOSFET) but have some grunt to drive
the diodes and caps.

Sorry for the big size, couldn't shrink it!

Edit: Wasn't latest rev. Fixed.

 

[emrr]

no coffee yet, but I like the army of chokes.  Good use of vertical signal flow techniques.  ;D  I've not played with this sort of thing before, so more thought required here. 

 

[Winston OBoogie]

no coffee yet, but I like the army of chokes.  Good use of vertical signal flow techniques.   ;D   I've not played with this sort of thing before, so more thought required here.  

Hi Doug,
Yeh, I'm using an old Tekronix or H.P. (?) transformer that has two identical H.V. windings and two 6.3 C.T. windings.  I needed to drop less voltage on the bottom so I had enough V for the C.C.S. to work from so I threw in some Hammond chokes to get the ripple down with less drop.  

Hmm? I haven't had any coffee yet either (and not much sleep) so I probably should wait before I open my mouth, but, now that I look at this thing... I'm wondering if I can't just get rid of the choke on the 6ES8 anodes, put a stiff current source under the 7N7 and just let the 6ES8's work off the 7N7 anode to grid feedback R's.  Make the supply a single 300V or thereabouts.  I'll wait until I'm a bit more chipper and look at it again.

Cheers.

 

[emrr]

I was looking at those feedback resistors and considering their wee bit of voltage stabilization, but hadn't gone that far yet.    What if the feedback R were chokes? 

 

[Winston OBoogie]

I was looking at those feedback resistors and considering their wee bit of voltage stabilization, but hadn't gone that far yet.    What if the feedback R were chokes?  

Actually, I was using the feedback resistors to cheat a bit!  They were giving a bit of 'fake' compression from the output stage.  I had a bit of a nap and woke up thinking this was a crap way to do things.

Latest iteration has a real tapped plate choke (was using an output transformer before although it wasn't shown as such) and has some R's to shunt the anode r and define the source a bit.  I removed the F.B.  Mo' better, not fake.

Also simplified the power for the lower stage by using in a couple of Triad 68X isolation transformers in reverse.  They're nice because they have static shields and are only about $11 apiece.   It's a bit more of a brute force supply now but I can't see or hear any difference comparing it to the supply with the chokes.  No doubt because I'm isolated from it with the C.C.S. and the shunt reg anyway.
Mo' better, now cheaper.  

FYI, the side amp is currently an old Langevin 138M Bridging amp which feeds a 6AL5.  It's overkill but I'll replace it with some sand when I have time.  

Last thing:  I thought I saw a post from Rowan earlier and was going to respond with thanks etc. when I got back online but it's now gone...  
Hey Rowan, was I hallucinating?  If not, thanks for posting.  Come on back man  

Gotta go and catch up on some other stuff...

 

[emrr]

I saw it, it went away. 

 

[Winston OBoogie]

Current iteration of this compressor has gone through many tweaks and changes since my last post although the basic principles are still there.

It seems to be a winner.

Enjoy,

 

[mitsos]

610 outputs:  I ordered 12 about a week ago.  Not sure when I'll get them back but, if they turn out well, I'll order more for sure.

Hey John, how'd these trafos turn out?

Is the comp in this thread going to be released as a new product or was it a custom job for someone?