To those OS Comp builders

[petemoore]

To those whov'e built the OS comp I have a question.
 Having built it, I'm close to having It 'working', but it barks.
 Question: What are You're Jfet drains connected to?
 Or put another way: What are you're Jfet sources connected to?
 Or which pin on the Jfet goes to ground?
 Because the Fet's Ppl use may be different, by determining the circuit connections for me Of Drain or Source avoids the Different Jfet Mfr. crosspin confusion. I know where the Gates are in the middle of the Fet figure on the schem. they look perfectly symmetrical [nothing I found that indicates drain from source]...
 Could someone please help me clear this up?
 I could explain why I'm confused better, once I get unconfused.

[lightningfingers]

i got the same confusion you are currently suffering from, heres what i did:

i replaced the bit with the jfets with a LED/LDR
mark hammer wrote a great post on it
ill find it if your interested

[petemoore]

Any theories or results on how that's supposed to work?
 What the heck, I might just re-OS, so I can A/B.
 Of course to do that with any conviction, it will still require my feeling that the Jfet's Are properly oriented.

[petemoore]

I forgot to type this.
 Does anyone know what some acceptable pin voltages would be for the OS Jfets?
 The OS build I have so far shows very good pin voltage readings on the OA, passes clean signal with good boost, however upon hard attack [last night] throught a MkII, I was getting undusable barking. I had to back the OS volume below unity to get around it.

[petemoore]

I knew it was there...
 The little G's on the layout...trace those out and write them on the schemtic...N/P.
 I should be able to get it now...

[petemoore]

That arrow pointing up...Tha's +V, in my case ~9v+...should work slightly different than that being connected to ground !!!
 I just noticed [after decoding using the layout, schematic, and Jfet pinout] ... [And still hoping that the layout works with J201's with straight legs >GSD< laying on the table, with J201 facing you], that the little arrow is same shape/size as something around the 9v+ area [DC Jack?].
 I don't use DC Jacks and didn't really look closely and compare that area to the little arrow over the top of the left Jfet.
 I go look around for schematic decodeing texts and see if I can find an arrow like that. I just don't remember seeing one like that before.

[strungout]

Check out GGG for voltages on those jfets, the have them listed in the OS build instructions(along with those of the IC).

I was wondering about those little arrows too at first, until it hit me that an arrow pointing up was the opposite of an arrow pointing down (ground symbol) . It's always a good idea to cross-check schematics. Keeps you from going "is it ok to use this instead of that, cause I don't have this?" sometimes too.

And aren't drain and source the two pins that are interchangeable?

Ciao.

[petemoore]

The OS is a happy circuit. Nice tones. Good character.
 Does a nice job driving an amp, seems to even the individual string volumes when going up and down on a string or three.
 The only noise problems I had was the barking with the Drain to Ground, other than that it seems to function perfectly, and on this one, it seems there's a very slight, but just discernable effect with turning the 'other' pot. :wink:
 Now back to wiring the 4th pot on the Overdrive Pro.

[lightningfingers]

Consider using an LED/LDR combo instead of FETs, if possible. You may have to up the gain of the op-amp a bit, but it should work.

The entire subcircuit involving the FETs simply creates an attenuator on the input of the opamp. The 82k resistor forms one leg of this imaginary input-volume pot, and the pair of fets and associated components forms the other leg.

Run an LDR from the junction of the 82k resitor and .047uf cap to ground (replacing the 2 FETs, 2 caps, 2 resistors and trimpot that normally go there). Replace the 470k resistor from diode/cap/resistor junction in the envelope follower with a 1k resistor and an LED, and that should be enough to do it.

The existing circuit uses component values that will produce a certain amount of input attenuation for a certain input level. As such, depending on the LDR you end up using, you may have to a) reduce the 82k resistor a bit, or b) run a 500k-1meg resistor in parallel with the LDR, or c) stick a variable resistance in series with the LDR.

What you want to end up with is enough of a change in resistance to ground that there is audible gain reduction (produced by attenuating the input signal) without the attenuation effectively turning the input signal off. So, for example, if your LED and LDR choice illuminated the LDR enough to drop the resistance to ground down to, say 5k or less, that would likely produce too much attenuation, so you'd want to stick some additional resistance in series. If the amount of resistance change was great, but still not enough to produce usable attenuation (e.g., an LDR that went from 10meg to 200k), you'd want to stick some resistance in parallel to produce a more effective range. With a 680k resistor in parallel with our hypothetical LDR, the resistance change would now go from about 637k to about 155k. In conjunction with the 82k resistor, this would produce modest attenuation possibilities.

[/quote]

[Mark Hammer]

The current newstand issue of Vintage Guitar has a big two-page ad for new products from E-H, and one of those listed for April release is a "White Finger" pedal, described as a "multi-FET compressor that yields transparent long sustain".

Hmmmm, do I smell the scent of oranges in there somewhere?