Altec comp attack/rel modifications

[mitsos]

Hi, I've got an Altec 1591 mic/line amp,mixer,compressor thingy that needs some love.  In the middle of recapping it and thought I'd expand the attack and release times.  The release is switchable between 0.5 or 1.5 seconds, controlled by a front panel switch (switches between C15 and C16 in the schematic).  I am going to remove those and put 6 caps on a rotary switch to get a few more times.  Seems easy enough?

But I'm not sure what controls the attack time.  Can someone shed some light on the workings of the sidechain? 

thanks in advance!

 

[iampoor1]

Very interested in learning as well! WOndering if it has to do with the photocell and impedence?

 

[JohnRoberts]

attack looks fast all the time, with hold/release based on R36 and C15.C16.

R3 and C15,C16 slows down how fast the hold voltage gets charged up, so a brief transient will limit the transient without causing the gain to clamp down and over react. Longer loud sounds will charge up the hold cap(s) through R3

You could just drop in a spread of capacitor values bracketing those. Does the switch select C15, C16, and then both for third position?

=====

A subtle consideration, electrolytic caps with dielectric absorption in side chains  can affect the CV envelope since they are charges with low impedance and discharged with different higher impedance. I seriously doubt that is worth worrying about for that design, or even that it is bad, it just is. 

JR

 

[mitsos]

Hi John,

thanks for that.  R3, do you mean R37 (kinda hard to read the 7, but it is a 4K7 resistor in series with the caps). I imagine that's the one.

Attack is 30msecs according to the manual, wouldn't it change with the cap value?  The switch chooses one or the other, no "both" option as far as I can tell.

So is my plan of having a series of caps on a switch the best way to do this?  Or will varying R36 affect attack?  It would seem to me that making it smaller would allow more current through it and less to the caps, so longer attack time?  If that's so, maybe make it a pot with resistors on either end to limit travel?

Thanks again!

 

[JohnRoberts]

Hi John,

thanks for that.  R3, do you mean R37 (kinda hard to read the 7, but it is a 4K7 resistor in series with the caps). I imagine that's the one.

yup R37

Attack is 30msecs according to the manual, wouldn't it change with the cap value?  The switch chooses one or the other, no "both" option as far as I can tell.

It looks to me like no lag in the side chain attack for transients.. only lag to the hold up circuit. I suspect there is audio path attack time lag due to lamp/gain cell time constants. 

So is my plan of having a series of caps on a switch the best way to do this?  Or will varying R36 affect attack?  It would seem to me that making it smaller would allow more current through it and less to the caps, so longer attack time?  If that's so, maybe make it a pot with resistors on either end to limit travel?

Thanks again!

Simplest is to change the caps..There are multiple moving parts. I suspect you will only affect the release.

FWIW  I designed a cheapo compressor back in the '80s with caps and 3 position switch for fast/medium/slow att/rel.

JR

 

[mitsos]

Yes, I'm going to stick a few caps on a switch and probably leave it for now.  For edumacation purposes though, to reiterate a bit. The caps (c15, C16) charge through R37.  They then discharge through R37 +R36 + part of P4 ( the zener and cap throw me off).  The release times don't quite work out to what's on the panel using just the resistors, do the zener/cap play a role as well?

What about simply varying R37?  smaller should be faster charge time, larger should be slower, right?  I am having a hard time seeing this thing, like you say, many moving parts.

thanks again

 

[PRR]

It's 1967. The opto lamp is incandescent. That's all your attack time. SLOW.

R37 also affects charge (attack), and *should* have been your first suspect (R37 C15/17 R36 is a classic attach/release network). However R37 C15/16 happen *after* the lamp gets fed, and won't slow its attack, and for rational uses we don't want to slow an incandescent.

If you must bodge-up: the inky lamp could be replaced with an instant-on LED. Q15 can certainly drive it, maybe to blow-up. Now we have to move the A/R network to before Q15. R13 may be a likely place. But considering that excellent VCA limiters are $100-$300, and passable LDR limiters not a WHOLE lot more and sometimes less, I'm opposed to chopping-up even less-loved Classics.

 

[mitsos]

Thanks, that kind of clears things up. When I looked at R37 the first time it seemed too simple, and when I saw the single attack time in the manual, I thought it couldn't be R37 because attack would change with the release cap (and drastically at that, since it's more the 3x change in capacitance). But then, if the light bulb is the limiting factor in the attack time, then that's the main reason it can't be made much faster. 

I'm not really going to delve into changing the circuit drastically, but I was hoping there was an easy way of changing the behavior like there is with release time.  I like adding more options, but I think it might be better to look into the ratio/compression control.  Instead of having two 10:1 positions (one at 0dBm and one at +8) and one 5:1,  I thought of eliminating the 0dBm option and having 3 ratios, 10:1, 5:1 and 2.5:1, all at +8dBm.

On the switch, the middle section changes the threshold from 0dBm to +8dBm.  The other two sections effectively only change in the last position, which is the 5:1 ratio (the other two are 10:1). So if we wanted another ratio (say 2:1 or 2.5:1 for starters), could it be done by changing the 180K and 68K resistors on the other parts of the switch? 

thanks again!

 

[mitsos]

all done.. Nice and quiet.  Ended up with a 2x6 rotary with 6u8, 3u3, 2u2, 1u5, 1u, and 0.47u.  First two caps are tants, the rest are films.  Will do some testing to see if these stay or if they get swapped for other values.  Thanks everyone, I'll post back if I make any changes.