3 legged dog build report :D

[vdm]

hey all!
i finally completed my first effect - after a failed attempt at a tubescreamer, a mxr dist+ and an abandoned perf attempt at a blue box.
so i have my new 3 legged dog and although it seems to be a bit of a one trick pony it does that trick extremely well!

the few mods and changes i made were:
2n5457 instead of the j201
6.8k bias resistor (not 8.2k)
560k resistor between 2nd and 3rd stages (not 470k)
10uF cap after 3rd stage (not 4.7uF)
and i accidently wired the pots backwards.

i havent done a whole lot of tinkering with it, as i've been too obsessed playing the hell out of it, but i've socketed most of the coupling caps, and the bias resistor on the back. at the present moment i don't have an MM so i can't test the voltage on the jfet but it sounds very nice at the moment.

the sound is quite harsh with a strat but has a nice smooth bite on the neck pup of a les paul - moving to a more defined attack with the bridge pup. through my marshall avt100 it sounds nice on the clean channel and really opens up the pick-attack with the overdrive channels (though it gets fairly noisey).

my drive pot seems to be best about half way - it just goes to mush and horrible gluggy distortion at max volume. i get the feeling this is because the jfet isn't biased properly, but it works very nicely as is.

the increased output cap was done simply because i didn't have a 4.7 on hand and wanted to get the thing working quickly - yet the box still is not overly bassy and im tempted to increase the other ones as well to give it a little more bass - though it is nice and defined at the moment.

the distortion itself is very tight, and on a clean channel seems fairly quiet. the intermodulation distortion is kept very low on this pedal and is much better than the distortion on the amp.

i built it all on one of those ic socket boards like the picture on run off groove, and it works quite well - my best soldering job as yet:)

i put the effect in this cute little carousel box, of which i have 2 others each one successively larger - so they'll be for my next effect - my family of effects.

well any ideas on the mushy gain control would be good help, but other than that, adieu.. and i'm off to go plan my next pedal suggestions are welcome.

seeya guys

trent

[gez]

With the Dog, input impedance of the first stage increases as gain increases, which is usually the opposite to how most distortions work.  I've used a similar approach in the past to get rid of the fat soupy sound you get at low gain in CMOS circuits, but it's a fine balancing act.

One thing you can do is wire up the first inverter stage similar to the Anderton circuit over at AMZ (Guitar Player one).  Stick a pot and stop resistor in series at the input and use a fixed value resistor in the feedback loop.  If you set the pot up to decrease resistance as it's turned then you can cut the low end as gain increases by choosing a suitable input cap.  

Sticking small value caps in parallel with the feedback resistors (this mod courtesy of Mark Hammer) will get rid of some of the high-end harshness.

This is entirely subjective, but you may find that removing the 1M & 2M2 DC offset resitors from inputs to ground smooths things out too.  I used to do the offset thing using 4007 chips and it made intermodulation worse.  I should think that the 4049 is more forgiving though.

[gez]

Here's the link:

http://www.muzique.com/schem/tubesnd.gif

I found that this circuit cut too much bass, but you can tinker with the values of everything (it's not rocket science!).  I'm not suggesting that you build this one, just steal from it!    

P.S. As shown, gain decreases as the pot is turned.