Orange Squeezin's

[willienillie]

I've never played an Orange Squeezer, or clone thereof, but I started looking into it after a buddy's recommendation.  Much googling ensued.  The first schematic that came up was this one from Beavis Audio:



My first thought was "Gee, that looks easy!"  My second thought was "Why didn't they use a 741 instead of half of a 4558?"

Well, turns out they did on some of the early ones:



I think that's a 1973 date code on the chip.  The 4558 models I've looked at have later codes.  If Jeff Baxter did use one of these on "My Old School" (released '73), seems it was probably a 741 model.

I bet it sounds the same, maybe lower noise with the 4558, and anyway that's NOT the point of this thread.

What sent me searching further was this line from Analogman's site:

The factory schematic and other schematics on the internet are NOT the same as our Juicer and the original Dan Armstrong builds, not sure why the factory messed up the schematic. But the pedal seems to work OK with the error though we still build them the same way the originals were built.

I couldn't find a factory schematic, but the GGG, Geofex, and Madbean schematics all seemed to agree with the Beavis.  Were they all right, or all wrong?  Was Analogman pulling our leg?  Most of the schematics omitted the filter cap across the power rails (usually 100µF), but that's not such a big deal on a battery-powered effect.  I doubted that was the error AM was referring to.

Then at some point I stumbled onto this shematic, from a now-gone Japanese website, but still showing up in Google images:



Hey look it's got the filter cap!  And also, check it out, one of the JFETs (Q2 in the Beavis schem) is reversed compared to every other schematic I've seen.  Is that right?  We know JFETs usually work either way, so that would jive with what Analogman said.  Well, I found component side and trace side pictures of several original units, figured out the JFET pinouts, and sure enough that last schematic is correct.  Even for the early 741 models.

It probably makes little or no difference.  But I thought I'd share my findings anyway.  I think I'm gonna build one soon.

Rock on.

[highwater]

It probably makes no difference. The drain and source are effectively interchangeable in most of the JFETs we use in pedals (and most JFETs period they way we use them in pedals).

Where it *can* matter is at high frequencies -- MUCH higher than audio. The reason is above my pay-grade, but my understanding is that they can be designed to have less capacitance between the gate and drain than between the gate and source. Since G-D capacitance (unlike G-S capacitance) is effectively multiplied by voltage gain (the Miller effect), a little bit less goes a long way if you're dealing with radio-frequency signals.

On another semi-related note, most (at 9v, probably all) BJTs will still work, but with reduced gain, when the collector and emitter are swapped. That actually *will* make a difference, and is worth trying the next time you have a fuzz on a breadboard.

[willienillie]

Not really out to bump my own thread here, but I've done a little more reading up on JFETs.  Apparently some types are totally symmetrical, and 2N5457 is one of those, even says D and S are interchangeable in the datasheet.  My impression before was that they "usually work" when reversed, but I didn't know why.  So anyway, it seems the orange squeezer would work exactly the same either way, and there's really no point to this thread. 

But since I'm back in here, I'll point out the other "error" on the Beavis schematic, suggesting that a matched pair of JFETs is required.  Obviously not.

I have this on the breadboard now, except with 2N5485s, because I don't have any 2N5457s.  It sounds nice, the effect is much more subtle than a Dynacomp.  Maybe it's a little stronger with the proper JFETs.  I'm not 100% sure I'm going to build one, but I probably will, and I'll get some '5457s from Steve if I do.  And install them the "correct" way, just because.  Maybe I'll even go with a 741.

[Ben N]

Only slightly OT: The interviews I have seen don't really nail this down, but my sense is that Baxter used the OS more for his pedal steel. Guitar solos like My Old School were, AFAIK, direct to the board--he preferred (then, at least) an overloaded mixer input to fuzzboxes and overdriven amps. Maybe an OS was part of that, too, I wasn't at the session . But I think his GP interview may have unintentionally mislead people on that score by mentioning both in the same sentence.

[willienillie]

Yeah, I have no idea.  I have seen old live vids of him with the OS plugged into his guitar jack.

Killer guitar work on that one either way.  I'm not a huge Steely Dan fan, but I'll always stop what I'm doing and listen if "My Old School" comes on the radio somewhere.

[PRR]

The basic JFET is totally symmetrical.

For HIGH frequency (radio) work, they may offset the Gate a wee bit toward the Source for a part-pF less D-G capacitance. This never matters in audio.