Pushme Pullyou Octave effect build questions

[Top Top]

I've been experimenting with a build of the Tim Escobedo's pushme pullyou octave up effect.



In a post by RLBJR65s that I found while searching, I followed his reported success using Q1 2N5088, Q2 2N2222, and Q3 2N5087, and 470K resistors rather than 500 though I didn't have a 2N5087 on hand, so I used a 2N5086 (2N5086 has 150hfe vs 2N5087 which has 250hfe).

Well, I don't get an octave effect at all. All I get is a sort of semi-phase canceling anemic fuzz. If I don't set the gain very high, then I get very little signal and just sort of a low sustain, clicking almost fuzz sound.

I checked my solder joints and double checked the layout against the schematic.

When I probe around, I get an clean, amplified signal after Q1, and can trace it all the way to the base pin of both Q2 and Q3, so my gain stage is working fine.

Then, after it has gone through Q2 & Q3, probing at both the collector of Q2/Emitter of Q3, and collector of Q3/Emitter of Q2, it comes out with this fuzz... no octave effect...

I have tried switching the transistors on 1&2 and also a bunch of different transistors that I have laying around, most of which are "switching" transistors (I have a bag of old transistors I got from a TV repair shop rummage sale), but I get basically the same result no matter what I have tried so far.

Any idea what I should try? I have heard two sound samples of the pushme pullyou, both sounded different from eachother, and mine sounds nothing like either of theirs - no ringmod like sound when playing two notes, no apparent octave effect...

I used Munky's layout of this project as a basis for my layout, BTW:

[Top Top]

One thing I forgot to  mention was that I experimented with lower resistance values in place of the 500K ones.

With lower resistance values, it cleaned up the sound somewhat and made it louder at the output (similar to what the signal is like straight after the gain stage), but still no octave effect.

Could it be that I just still need to find the right combination of transistors?

[BAARON]

The transistors might have something to do with it.  I think you should aim to match the gain (hFE) of Q2 and Q3 - that's probably why the original design uses 3904 and 3906 for those positions... R.G.'s mu-doubler needs matched FETs and it's a bit similar in some ways, so this would probably benefit from matched (or at least somewhat matched) transistors too.  A 2N2222 has an hFE ranging from as low as 35 to a maximum of 300 (depending on the circumstances it's used in), while a 2N5086 has an hFE range of 150-600!  The 3904 and 3906, on the other hand, have the same gain range as each other, which is probably why they're used in the original circuit.

Are you sure you have the right pinout on the transistors you're using?  They might not match the layout of the PCB you're using.

[Top Top]

Thanks for the response BAARON. A couple of other people had reported having better results with mismatched transistors, but I will buy the ones suggested in the design, and perhaps a 2N5087 as well and give them all a try.

One other question - the .1uF caps I am using are ceramic caps salvaged from some old circuit, as are the 10uF ones (though I am pretty sure those are ok because I know that circuit was working)... could a blown cap somewhere be causing this? I would think that would just cause the signal to not pass through, but I don't know if they can fail in a way where they just pass the signal straight through and not perform properly.

[BAARON]

Any two random 3904 and 3906 transistors are not likely to be identical.  They'll still be mismatched... just not as mismatched as two totally unrelated transistors are.  In order to cancel the fundamental tone (as stated in the schematic you posted), they'll have to be pretty close to each other or else they won't knock that fundamental out sufficiently to create the octave effect, AFAIK.

Not sure what a bad cap would do, but you could meter your caps to see if they're open.

[earthtonesaudio]

Search the forum for "pushme pullyou."  I seem to recall some discussion about purposefully mismatching the transistors of the NPN/PNP pair made a better octave.

[Top Top]

Search the forum for "pushme pullyou."  I seem to recall some discussion about purposefully mismatching the transistors of the NPN/PNP pair made a better octave.

Yea, that is what I was referring to in my original post about the post by RLBJR65s where he reported successfully using the unmatched transistors I mentioned there (Q1 2N5088, Q2 2N2222, and Q3 2N5087)

It could be that my transistors are TOO mismatched though, I suppose.

I don't really understand why mismatched transistors would work better than matched ones personally, but that is what was reported... perhaps the mismatch allows for emphasis of certain pleasant harmonics.

[Top Top]

I did some more experimenting with this plugging in transistors randomly from a bag that I have (I have sorted out PNP and NPN) and I got it a little closer. I know I have a 2N5088 for Q2, don't remember the other two, Q3 is a metal can PNP and Q1 is a strange UFO looking transistor...

On higher notes, it has a sort of fuzzy octave now and there is some ring mod-type sound when you play two notes. On low notes, it still sounds like a thin fuzz. Still haven't really gotten a clean octave out of it, but it is good to know I am on the right track and that it is just a matter of the right combo of transistors.

[Top Top]

Just an update on this build... well stupid me, the reason I wasn't getting a very obvious octave effect was that there was a broken solder trace that I couldn't see by looking (hairline crack).

The problem was that Q2 and Q3 weren't getting the +9v, basically. With it wired correctly, the circuit is pretty lenient after all in terms of the transistors you use.

After a bunch of experimenting, I went with a 2N5088 on Q1, 2N2222 on Q2, and 2N5138 on Q3 (random PNP I had laying around).

I tried different transistors on Q2 and Q3, but the differences there with different transistors there were very subtle. Q1 had the most obvious effect on the sound - mainly because of the amount of gain you could get. Q2 had some effect, but not as drastically. Q3 was the least noticeable when switching them out.

I tried lower gain transistors on Q1, but I liked the wider range that the 2N5088 allowed with the gain control. I don't think I will use full out fuzz that much, but it gives me the option if I want it.

So, now that the circuit is actually wired correctly, it works great! This is actually a really cool circuit, I have never used analog octave up effects before aside from trying a green ringer in a guitar shop once.

It can do a lot of different sounds with different pickup, gain, and guitar tone settings - on the bridge pickup with the gain up only about 1/2 way (I used a 10K pot instead of a 1K, which actually works really well - about 3/4 of the knob's range is useful), it can do some jangly electric sitar-type sounds - adds nice buzzy overtones, less sustain.

On low power chords with the gain up, it makes a thick pulse wave sounding fuzz.

And of course the octave-up effect - it can get a range from a pretty clean smooth octave up (which works best on upper frets), to totally fuzzed out octave fuzz.

And then the fun ring mod type stuff when you play two or more notes on the higher strings - fun to play drones and find harmonic overtones.

Really amazing for such a simple circuit. I am glad I tried it out.

[earthtonesaudio]

Excellent build report, thanks!  Glad to hear you got it working (and sounding good too).   

[Top Top]

Thanks!

I tried playing it today placed after a reverb and that was pretty interesting, as the sound of the note you are playing modulates against the decaying previous notes. There is also a sort of cancellation that happens as the note you are playing reverberates back. Sounds pretty cool with the right amount of decay on the reverb (not too much). Certain settings with the reverb and low gain on the pushme pullyou and the bridge pickup sounded even more sitar-like.

I bet this effect would sound cool after other effects as well - delay and chorus are two that come to mind.