Kliche overdrive diodes

Most diodes are electrically equivalent, so trying different ones won’t harm the circuit. In the Kliche they just clip the signal so they will affect the sound but no risk in trying different models. You can even try asymmetric clipping (two diode in series on one side and one on the other) or LEDs.
 
Any Germanium diode can be a decent sub for 1N34A in a Klon. I agree with spi, install sockets and experiment. Many people use Schottky, some without even knowing it. Silicon, like 1N4148, will sound different, some people prefer it. If you use LEDs the last two stages can saturate and that will definitely alter the tone.
 
I've built a few Kliches and the one I put D9K's in is by far my favorite. Next favorite is D9E, then the last one is with cheap 1N34A's i found on amazon 😂 surprised they worked.
Nice to let me know, I’m actually planning on using cheap 1n34a in my next klon-style build. Definitely gonna make them easy replaceable.
 
Hi, I'm gonna resurrect this thread to ask, what makes for a good Ge diode in this circuit? I have a handful of 1N34A's and some 1N270's. I might also get some more from Smallbear just to have some variety in my parts drawers. Other than "socket them and see," which I plan to do anyway, is there any datasheet detective work to determine what datasheet parameters might affect the sound, or what DMM measurements might help me make an informed design decision?
 
Nope, the datasheet won't tell you the stuff that affects how they perform in a Klon or any of the other hard clipper circuits. There won't be all that much difference between the various Ge diodes.
So if they sound the same, is it even worth it to use sockets? Well, I guess it would to protect the diodes from the soldering iron...

What should I look for in a circuit to determine that the stats on the datasheet would really matter? Any examples where it might actually make a noticeable difference?
 
So if they sound the same, is it even worth it to use sockets? Well, I guess it would to protect the diodes from the soldering iron...

What should I look for in a circuit to determine that the stats on the datasheet would really matter? Any examples where it might actually make a noticeable difference?
I put a set of each different germanium diodes I have on a breadboard so I could switch between them easily and did see some pretty dramatic differences. They all still sound like a klon but definitely different flavours. I'd say it's worth experimenting even if it's just to see whether you can hear much difference in them yourself. My one piece of advice is to go for the germaniums you have with the lowest FV from each type you have and try those first
 
Agree with Chuck, but I have a bit of an out-of-the-box suggestion: Try 1N5817 you use for polarity protection. You probably have some around and they're hella leaky like a D9-Russian whatever.
 
IMHO, the datasheet stats are basically useless for pedal applications. The three parameters we care about are Vf, Ilkg and Rd.

For Vf, we need to know the current at which it is measured. Lower Vf provides more compression and lower output.

Ilkg matters if we want to use Ge diodes in a soft clipper. IMO, Ge diodes as a rule do not work well as soft clippers because their leakage is too high. In hard clippers, like the Distortion+, Klon, DS-1, Rat, etc. Ilkg doesn't matter.

Rd is the diode resistance. It can have a strong effect on the Vf and compression if the diodes are driven hard like they are in a Rat or DS-1. I measured some D2Gs that had Vf = 0.66V @ 5mA. You'd think they were Silicon! The Vf is high at that current because their Rd is over 30Ω. Down at 1mA, Vf = 0.35V.
 
Agree with Chuck, but I have a bit of an out-of-the-box suggestion: Try 1N5817 you use for polarity protection. You probably have some around and they're hella leaky like a D9-Russian whatever.
Those have very low Vf (<0.1V @ 1mA) and Rd (5Ω @ 5mA).
 
So if they sound the same, is it even worth it to use sockets? Well, I guess it would to protect the diodes from the soldering iron...

What should I look for in a circuit to determine that the stats on the datasheet would really matter? Any examples where it might actually make a noticeable difference?
I think there are cases where the specs affect the sound. For example the rat has its characteristic filtered sound thanks to the frequency response specific to the LM308 chip. A chip with different specs would sound different.

Depending on the region of operation (active, saturation, etc) or frequency response, different transistors can also sound different. But it all boils down to how that part is supposed to work within the particular circuit. In a buffer, a transistor is merely replicating the same input but making the output impedance much smaller, so a variety of different specs may work more or less equivalently in the same circuit. In a fuzz with a ge transistor in saturation, small differences in specs may produce very different sounds.
 
I agree with all of that. Just bear in mind that the datasheet offers little or no information on how an opamp saturates and recovers from saturation. When Pro Co selected the LM308, it was either by listening tests or that's just what was available at the time. A lot of these fuzz circuits just happened. Pedal builders didn't use germanium because it sounded better than silicon. They used gemanium because that's all there was. Same goes for opamps. JFET opamps were rare and expensive when the Rat was developed.
 
I agree with all of that. Just bear in mind that the datasheet offers little or no information on how an opamp saturates and recovers from saturation. When Pro Co selected the LM308, it was either by listening tests or that's just what was available at the time. A lot of these fuzz circuits just happened. Pedal builders didn't use germanium because it sounded better than silicon. They used gemanium because that's all there was. Same goes for opamps. JFET opamps were rare and expensive when the Rat was developed.
Dunno if this story is less truth and more story but

“A quote from the Book 'Stompbox' by Art Thompson, published by Millar freeman Books. The passage starts 'Scott Burnham's story';

"I had the basic circuit roughed together and I'd found an op-amp I liked, the LM308N, which was an instrumentation amp used for seismic and medical sensors. I was experimenting with an EQ boost for this op-amp in order to pre-boost the treble so I could use just a passive tone control to cut back the highs. i was bypassing the voltage divider that sets the gain when I picked up a resistor, looked at it and thought to myself, "yellow, violet, brown - thats 470ohm." I plugged it in expecting to get about 50dB of gain, but when I picked up my strat and hit a string , it went wooooo. I thought, " Holy shit, this is cool. What did I do?"
I looked real closely and realised that I'd plugged a 47ohm resistor instead of a 470ohm resistor (sound familiar
::)
). That meant it had somewhere around 70db of gain , which, according to it's spec sheet was impossible from that op-amp. Trying to set the gain on this thing I had stumbled across a combination of resistors that produced this really weird high-frequency shelving boost that the op-amp couldn't possibly sustain. It didn't have enough slew rate to produce that much gain at those frequencies so it drove the op-amp into incredible slewing distortion. This usually is very bad , but in this case it gave the Rat it's yeowl - I've never heard any other stompbox make that sound."”

via https://www.diystompboxes.com/smfforum/index.php?topic=55234.0

but yeah happy accidents like that are the exception and not the rule (and I’m sure selecting the LM308 in the first place had as much to do with cost as it did with sound). “Mojo” parts aren’t rare because they’re “better”, they’re rare because newer better stuff replaced them.
 
A great story.

The slew-rate limiting actually happens in the second stage inside the opamp. That 30pF cap connected to pins 1 & 8 limits the slew rate. Any externally compensated opamp can have its slew-rate altered by way of that capacitor. The NE5534 is a low-noise opamp with external compensation. It should be possible to slow its slew-rate down enough for it to mimic the LM308's behavior. Worth a breadboard experiment. If you try that, be aware that the NE5534's compensation pins are 5 & 8.

[edit]

VFE built a hopped-up version of the Rat called The Alpha Dog Glass using an NE5534 "for brighter attack and tighter bass response."
 
Last edited:
A great story.

The slew-rate limiting actually happens in the second stage inside the opamp. That 30pF cap connected to pins 1 & 8 limits the slew rate. Any externally compensated opamp can have its slew-rate altered by way of that capacitor. The NE5534 is a low-noise opamp with external compensation. It should be possible to slow its slew-rate down enough for it to mimic the LM308's behavior. Worth a breadboard experiment. If you try that, be aware that the NE5534's compensation pins are 5 & 8.
Chuck, can I do college over again and have you as my thesis advisor?
 
Back
Top