Jawari!!

[Mark Hammer]

Put one together last night.  Can't put it down.  Just a clever, clever circuit, Tim.  You have my undying appreciation.

Takes a bit of getting used to to get it to behave properly but that's mostly picking technique.  Sound best with a Tele bridge pickup, but then if you look at pics of the Coral Electric Sitar, you'll see a lipstick pickup by the bridge so that makes sense.

I'm curious about how it does what it does, and whether there is anything that could be added to make it behave in a more predictable and perhaps recordable fashion.  I'm all for simplicity, but if a little more complexity turns this into the kind of thing you'd be proud to turn on during a gig, why the heck not.

BTW, Constantin, Aharon and all the other DIY-ers in the Toronto area, Supremetronic has the Mouser 42TM018 10k:10k transformers in the back with all the other transformers for a buck each.  They're red.

[Jason Stout]

I STILL haven't built the jawari! It has been on my list from the beginning.

[Marcos - Munky]

The Jawari is very cool. With a distortion between guitar and Jawari, you get a crazy octave sound.

[Tim Escobedo]

Though it really is just a weird distortion device, it behaves like no other. I basically describe it as a underpowered full wave rectifier. It seems the diodes are not completely "turned on" when the signal level is low. Which seems to translate into a kinda exaggerated dynamics. This is tamed a bit by the diode to ground, something I'm always fiddling with, to get just the right amount dynamics while keeping the output level high enough.

The unique harmonic spectrum at the output is the result of a reasonably harmonically complex input, along with the unusual resulting harmonics from a FWR. This is one reason I recommend the treble pickup over the neck pickup. The neck pickup, being generally more sensitive to center node vibrations, gets more fundamental tone than the treble pickup. The rule of thumb for this is opposite of what one generally wants from a octave up device, where more fundamental translates into a octave without too much hash. The Jawari, without much gain to start with, generates clearer (and unusual) sidebands with odd dynamics, resulting in it's buzzy tone. And less sustain.

Add gain stage on the input, and play with the neck pickup, it morphs into a regular octave up.

Be warned. A few have not been able to get the circuit to work.

[Mark Hammer]

Hmmm, so would some sort of insertable treble boost improve its performance across variations in pickup and/or guitar?  Perhaps a dual-input-cap-plus-panner arrangement like Joe Gagan has cleverly implemented on his various FuzzFace extrapolations?

Is there any sort of relationship between the Ge diodes and silicon diode that results in optimal sound?  Alternatively, since the centre-tap is used on the diode side of the transformer, and since diodes vary in their forward voltage, maybe some sort of quasi-nulling arrangement like JC Maillet used for the Green Ringer is called for here.

[Tim Escobedo]

A nulling trimmer might be useful. I've never tried it, however.

Some of the original prototypes did some severe (and adjustable) bass cutting at the input. Eventually, the circuit was honed down to the final version, though. I'm not sure such a passive dual cap input would be as effective on the current design since the single FET doesn't have as much gain to compensate for the losses. The older designs used several different gain devices, op amps, BJTs, etc. Perhaps a more effective way of implementing such a control would be a smaller cap bypassing the source resistance.

The Si diode is only there to limit the output peaks. I've fiddled a bit with series resistances or subbing the diode/resistor with a LED.

The Ge diodes are pretty critical. They seem to "turn on" at just the right level. Regular Si diodes don't really work in there in this particular circuit. You might be able to sub Schottky diodes, as they seem to behave similarly to the Ge. But there is a noticable difference when A/B compared. Not necessarily better or worse. Just different. May be worth a try.

Finally, it's nice to use Ge diodes from the same "batch" if possible. I've found Ge diodes from different sources to be different enough that a nulling control might be necessary.

[Mark Hammer]

Thanks for the info.  I'm almost out of 1N34A's, but have a half dozen or so 1N60's.  I'll try and pull the best matched pair from what I have left out of that and put together another one, perhaps with a simple nulling trimpot just to see what difference it makes.

The suggestion about input cap panning was predicated on a more complicated pre-transformer input stage.  I recognize your preference for least-technology-needed.  This thing sounds so good when it works flawlessly, though, that I feel compelled to insist on anything that will MAKE it work at its best all the time with any guitar.

Ironically, when I was working on something else yesterday, I had the radio on, and they started playing "Cry Like a Baby" by the Box Tops.  Just had to plug in and join along.