1n34a --> 1n4148?

[Badman]

A while ago i build a tone machine clone using specs from the Fuzz central. I couldnt find the prescribed 1n34a diodes, or any of their germanium substitutes so i tired it with AA133 (also germanium). Turns out they didn't do the trick, as there is no octaving when i kick the switch, but the fuzz sounds pretty decent, though. I've poked around the PCB with an audio probe, took some readings, checked for cold solders, traced and retraced the whole circuit and came up with nothing. So I came to the conclusion that either the diodes are a bad pick or i fried them while soldering. So my question is: Is it possible to substitute 1n34 germanium diodes with 1n4148 silicons?

Original debugging post: http://www.diystompboxes.com/smfforum/index.php?topic=90367.0

[tiges_ tendres]

I think the diodes need to be matched to get the octave going in this circuit.

As for where to get 1n34a diodes from, you can get them from the store on this very site:

http://diystompboxes.com/zencart/index.php?main_page=product_info&cPath=13&products_id=6

Very affordable too!

[Badman]

so 1n4148 is a no go?

[tiges_ tendres]

so 1n4148 is a no go?

I don't think that those will work. 

Click the link and spend $3, support the forum.

If you don't have the means to test the diodes, you can socket them and test that way.

[Badman]

3$ are not the issue here. I would gladly pay 10$ if it would make the damn thing work. But the trouble is I don't have a paypal account and shipping to Croatia would probably take some time so i was kinda hoping that the 1n4148 would work since they are readily available in local stores. Thanks for the advice though, I appreciate it.

[Mark Hammer]

Diodes are used in two capacities in the FTM.  One pair is used to rectify the signal from the phase splitter (Q2), and another pair is used to provide some extra clipping between Q3 and Q4, just ahead of the tone control.  That second pair can easily be replaced with Si type (1N914, 1N4148, etc.), and may even be an improvement.  I made one with a switch to use either a Ge or Si pair, or no diodes at all.  The no-diode position provided far more volume (as expected) but didn't really suffer from lack of distortion.  Not as much sizzle as with diodes, but a very pleasing distortion nonetheless.

The first pair, can be matched, but that may not help much if the 4k7 emitter and collector resistors on the preceding transistor are mismatched.  The key to getting robust octaving lies in having matched signals.  Matched diodes help, but only if the collector and emitter signals are, themselves, matched for output amplitude, and that depends on having the two resistors be close (although Bernie Hutchins noted in an old Electronotes that the emitter resistor can be a little lower....a little....than the collector resistor).

Changing the diodes to Si type will result in greater crossover distortion - that is, clipping of the sides of the waveform - in addition to rectifying the signal.  Assuming that one ends up combining equal-amplitude versions of the rectified signal at the junction of the two diodes, you WILL get octaving, though the sound may be a tiny bit thinner and sound like there is a tiny touch of delayed onset.  Removing/lifting that second pair of clipping diodes will fix that in a hurry.

Where I suspect you'd notice the difference between Ge and Si diodes most would be when you cancel the octaving.  By relying on only the half cycle that comes through one of the rectifying diodes, a higher forward voltage will tend to thin out the sound a little bit.  On mine, I use a 3-position toggle to select between octaving, non-octave with crossover distortion, and non-octave with the remaining diode bypassed.  I described it in another FTM-related thread you can search for.  Personally, I prefer it, though I understand others may like the rather distinctive sound of the crossover distortion.

[egasimus]

You could probably replace them with 1n60 or 1n270. I'm sure there's a source of those near you, I managed to find them in Bulgaria and we're much worse off than you guys You could try dismantling some old piece of electronics for the diodes, transistors, and film caps.

[pinkjimiphoton]

in the dano frenchtoast octave up, the original diodes were silicon. i changed them to 1n34a's, it made absolutely no difference in the performance of the pedal, tho to my ear the ge's had a little more and better midrange tonality.

peeps say the french toast is a ftm, so i would guess from my experience the diodes shouldn't matter as much in this circuit. it probably has more to do with matching the transistors i'd imagine.

[jasperoosthoek]

If you don't get the octave to work it could mean that you reversed one of the first two diodes.

[Paul Marossy]

If you don't get the octave to work it could mean that you reversed one of the first two diodes.

+1

[Mark Hammer]

Or that, as so often happens, the functioning of the octave switch was impaired during the process of soldering leads to it.

[amptramp]

If you have any germanium transistors with characteristics that are unsuited for amplifier use, you can connect base to collector and use them as diodes.