A *useable* external bias control for the Orange Squeezer solution :)

[midwayfair]

I decided to work on some squeezer mods the last couple days. I shared this over on Madbean, too, since much of the modding was done using his Cupcake PCB.

A usable external bias control
The weird thing is, the solution ended up being quite simple (and perhaps obvious), but I couldn't find it anywhere online.

The goal was to make an external sustain/ratio pot with the following sweep:

1) Unity volume with both pots fully (or nearly fully) CW.
2) Fully CCW on the sustain pot is a boost (i.e., minimal, if any, compression).

Testing shows that the resistance range between "compressed to the point of getting almost no sound" and "no compression at all" is very, very small. It's less than 1K in the middle of a pot's travel. I don't have a spectroscope, but my ears and my multimeter and trimpot put the actual range at only about 600-800 Ohms. Any further to one extreme or the other and you might as well be using a jumper or 10K resistor.

A couple ways to go about it:
The simplest way is to put the trim pot in series with a B1K. This is not ideal, but it will get you very close with minimal parts. Pretty much the first half of the pot will work as a booster, and then you'll get some nice compression, and the very last mm of the pot will sink into inaudible land.

Even better, you can use a reverse C1K pot instead. Fortunately, most of us have these around for fuzz faces, and checking the schematics reveals that the OS's bias is wired up pretty much just like a fuzz control. This will give you a much bigger range of compressed sounds and a much smaller range as a booster, about 1/4 of the pot. The last turn of the pot might still be below unity even with the volume maxed. You can also use an audio pot instead and it'll just work backwards from the way you'd expect.

What if you want an "ideal" sized pot? Obviously the thing to do, if we want to reduce the "boost" range to almost nothing, is to reduce the size of the pot. I opened up my Hartman Compressor, which is of course the inspiration for doing this mod, and measured the Ratio pot. It was ~750Ohms in circuit, pretty close to the value I measured above. My Hartman also has about 1/4 of the pot working as a boost with no audible compression, so this makes sense. (For the record, I don't know if what follows is how the Hartman accomplishes this, but the end result is pretty much identical.) Obviously no one's getting a 750Ohm pot ... so we have to stick a parallel resistor on there. A 3.3K is a good value to use if you want a little more fudge room at both extremes, but you can use something as low as 1K if you really want limit it to the "good" biasing range (honestly, though, I think if you don't have a booster setting, there's not much point in putting the bias control outside). I used 2.7K. It goes from boost at one extreme to just below unity volume at the very last turn. Pretty nice!

As an added bonus, strapping a parallel resistor to the pot creates a concave taper, which I found gave a lot of good control in the middle of the pot and kept unity volume when both controls were at 9:00 (very subtle compression, lots of boost), 12:00 ("normal" sounding), and 3:00 (squished). This made it VERY easy to find balanced settings.



Yes, this will still fit easily in a 1590A.

---------------

Bonus mods
I don't think I've ever built a "stock" Orange Squeezer. Here are some things worth doing:

1) Decay control: Replace R11 with a 500K pot. This seems fairly extreme when the stock value is 100K, but I've found there's not a big difference until you get close to 330K. After about 3:00 on the knob, you'll start to get an enhanced subtle bloom effect because the capacitor won't quite drain fast enough. At the CCW end of the pot, you'll get a very natural/correct sounding decay. Turn the pot down further or off and you'll actually end up with very little compression, because very little of your sound is getting "caught" in C7. I like controls that do more than one thing at once, and this is a good one: it affects the decay of the compression, but also the attack (just because of the way the OS creates its compression), the tone (darker as it's turned up), and the overall compression level. I'm surprised it's not done more often, actually.  I ended up replacing an external bias pot (which was useless without the above mod) on one of my Squeezers with this control and it might end up back on my board -- I like it that much.

2) Diodes -- the diodes don't affect the tone, but they CAN affect the compression level. 1N100s really do compress a LOT, but they're hard and expensive to source, so it's definitely worth thinking of alternatives. My testing shows that if your diode's Fv is below .35v, you might as well not be using it. Under .3v is much better. I found that 1N60Ps, which have a lower Fv than typical 1N34As (Fv = ~.25v), are very good and also very consistent. The nice thing is, you can pick them up for pennies. And lo and behold, the Hartman compressor uses them.

3) Reducing C7 (to 2.2uF or even 1uF) will give a slightly more natural attack and decay (less duck and bloom) without changing any other values. It'll sound less squished when you're not playing fast but still have lots of squeezer character. I usually adjust R11 (the decay resistor) a little (150K) when doing this -- it keeps the limiting effect on fast runs, but you still get the benefit of a fast attack recovery. I built my Clementine like this and it's pretty much perfect as a booster with a bonus.

There are other good mods out there (RG Keen and Mark Hammer have both written good posts about them). This circuit's a ton of fun.

Edit: SPAZ sent me his favorite mod, from one of Mark Hammer's posts, which is a really good one for people that use guitars with different pickup strengths:

"You can extract other sorts of performance characteristics or feel from the unit by varying the gain of the op-amp, such that you "drive" the FET less hard (or more) for the same picking strength. Replace R9 (10k) (on the Madbean schematic -- this will be R10 on GGG*), with a 6k8 fixed resistor in series with a 25k pot to achieve gains ranging from 7.9 to 33.5x, as opposed to the stock gain of 23x.  Note that gain changes will also affect output levels such that lighter compression (less gain) will provide less output as well."

*There's only one 10K in any given layout, so it should be pretty easy to find.

[Keppy]

Cool! I'll have to replace the crappy near-unusable external bias pot on mine one of these days.

[Kesh]

I'm currently tuning an OS, my first build.

I find the bias pot does an interesting thing near the ideal setting. At about 1.72V or 4.1 ohms, across the pot, a very low noise appears, that increases in pitch up to white noise type whistling that disappears above audible frequency at about 2V/4.8 ohms. This noise can be heard cutting out when the compression kicks in when I apply input.

I have no idea if this sound is a sign that this is the useful range, or whether I should tune it below this to avoid any noise.