Dan Armsrong - Orange Squeezer Mods

A forum devoted to mod, tips and suggestions for upgrading and rehousing your VERY CHEAP commercial stompbox to near boutique excellence.
Post Reply
User avatar
markm
Diode Debunker
Information
Posts: 787
Joined: 26 Jun 2007, 14:41
Location: U.S.A.
Been thanked: 10 times

Post by markm »

EdG asked a couple of questions about the O/S and I thought this would be a better place to put this info rather than start a new thread.

From Mark Hammer;
--------------------------------------------------------------------------------
Changes to the response of the envelope follower inthe OS *can* be effective in getting what you want. Just be realistic about what you expect, since simple half-wave rectifier circuits are limited in their flexibility.

If you follow the signal from the "branch point" after the first op-amp, you will see that it goes through a 1k5 resistor and a diode, with a cap and resistor to ground after the diode.

The diode assures that only one half of the signal passes on to the FET gates. The cap is charged up by signal, and continues passing on a constantly diminishing version of that voltage for a brief instant until it is fully discharged. The rate at which the cap is charged up with each burst of signal is determined by what limits the current flow to it, as well as the actual value of the cap. A much bigger cap would become fully charged to capacity at a much slower rate, and a much smaller one would be full up much quicker. Since the 100k resistor in parallel with the cap provides a path for the cap to discharge (remember, the cap is acting like a battery here, and the two ends of this "battery" need to have a path to connect to each other), the rate at which the charged cap *dis*charges after a signal impulse is set by that resistor (and the leakage properties of the cap).

So.....if you up the value of that resistor, the decay/discharge time will be longer. If you hold the resistance constant but increase the cap value, the decay will be longer. If you insert a small resistance between diode and cap to limit how quickly the cap can become full of current, you slow down the attack/charge time.

The problem one acqures in inserting any variable-attack mod is that anything which restricts current flow will decrease apparent sensitivity - i.e., make the unit behave like you aren't picking as hard, hence producing less compression. The simple solution is to compensate for that by having variable gain ahead of the rectifier. In this case, the envelope gain section is the audio path as well. Smart guy, that Dan Armstrong.

Personally, I try to have as much bang and variation for the fewest knobs that I can. I love to be able to change things around, but I also want to avoid having pedal controls feel like some sort of combination lock. So, it is exceedingly rare that I would stick both an attack AND variable decay control on such a thing, though that may be just me. Given that the decay AND attack time are affected by the value of the cap, I find it provides a reasonable amount of variety to simply have a toggle that can pick one of 2 or 3 cap values. In this case, being able to tack a 10uf and 22uf cap in parallel with the existing 4.7uf would provide a wide range of squish feels. certainly the parallel value of 22+4.7=26.7uf would extend decay time/lag out enough to provide extreme squish (might be nice for cymbals, too!). A small value pot of 500R to 1k would be reasonable to stick in between the diode and cap.

Note that the amount of squish or apparent compression a unit can provide is not only a function of how much it can reduce gain or level, but also how UNresponsive it is to any new transients. This can not e better illustrated than by the infamous "attack" mod to the Dynacomp. Although that mod actually alters the decay time of the critical cap in the rectifier circuit, anything you do to make it more likely that the cap will not have finished responding to the last note before the next one comes along, will create the impression of impaired attack.

I suspect one of the reasons why the OS has a reputation for "transparency" is the 100k resistor for discharging the cap. The result of this, in theory, ought to be that gain/level is reduced for the initial transient of the pick attack (i.e., the big peaks) but the device instantly becomes ready-for-duty for anything else immediately afterwards, rather than still caught up in discharging. In general longer decay times will increase the obviousness of the effect.

Between a toggle for decay time, a small pot for finer attack-time adjustment, and a variable gain pot to compensate for the attack control, you should be able to get a lot of different personalities out of the unit.

Hope this is clear enough.



Hope this helps.
Join together with the "Banned"!

User avatar
markm
Diode Debunker
Information
Posts: 787
Joined: 26 Jun 2007, 14:41
Location: U.S.A.
Been thanked: 10 times

Post by markm »

More info here;



Mark Hammer
Posts: 9923


Re: Orange Squeezer...very subtle compression?
« Reply #8 on: Today at 09:16:14 AM »

--------------------------------------------------------------------------------
1) The OS gets in and out of the way of the signal fast, so the squish is less noticeable than with many other compressors.

2) The op-amp gain that sets the amount of compression achieved is fixed in the stock design. Does one size fit all? I'll let your ears decide. I made the gain variable on mine and it makes a difference.

3) The 82k resistor and the FET+trimpot behave like a voltage divider where one leg (the 82k) is fixed and the other leg (the FET) variable. How much impact the variable leg makes on attenuation depends on the trimpot but on the FET and 82k leg as well.

4) The 2N5457 appears to have the appropriate properties, and like the other Mark I have had only positive experience with it. It is also used as the variable half of a voltage divider in the Jon Gaines Noise Gate project I have posted on my site.


Join together with the "Banned"!

Post Reply