“Correct Leakage Numbers”

Big Monk

Well-known member
I’ve seen this said quite a few times lately.

I’m not sure I understand. I’ve never considered leakage desirable enough for those to be a “correct” value for leakage.

My thought has always been that leakage should be the lowest possible value to bias a “no-bias” or “leakage bias” stage and that’s literally the only place it’s desirable.

Obviously, you put up with certain high leakage devices, along the lines of the OC75 for instance, because of the frequency content they may have. In that case, you do t mess with the vintage circuit because they bias up properly with leaky units.

Other than that, I would want low leakage devices and to then tweak the circuit to accommodate them.
 
  • Like
Reactions: fig
you do t mess with the vintage circuit because they bias up properly with leaky units.
I’m pretty sure you answered your question. Where else are people using them? I can’t think of a time someone wanted them for anything other than a vintage circuit. Although, people may be seeking certain qualities such as gating too.

Of course, you can always make a non-leaky Q leaky but not the other way around. There’s a lot of mystique at play here. I think most people see published leakage specs and want those exact specs in their devices. Additionally, the internet is riddled with dubious part-specific tonal quality claims that send people on quests for exact part numbers. I think that’s a bigger issue than not simulating leakage.
 
I’m pretty sure you answered your question. Where else are people using them? I can’t think of a time someone wanted them for anything other than a vintage circuit. Although, people may be seeking certain qualities such as gating too.

Of course, you can always make a non-leaky Q leaky but not the other way around. There’s a lot of mystique at play here. I think most people see published leakage specs and want those exact specs in their devices. Additionally, the internet is riddled with dubious part-specific tonal quality claims that send people on quests for exact part numbers. I think that’s a bigger issue than not simulating leakage.

I was also thinking about leakage ONLY being important if:

a.) You NEED the vintage transistor

b.) You feel the need to stick to the vintage circuit values.
 
When I said it earlier today I meant generally high or low. I found one today with an hFE of 88 and leakage of 1.2 mA 😲

Some of my GE 2N169s have been < 10 micro amps!

I could not get them to bias in a MK II. I’ve found 15 micro amps as a sort of lower cutoff.
 
There are quite a few modern circuits which use Ge - the Hudson Broadcast being a notable example. I guess it comes down to how we define "modern"!
 
  • Like
Reactions: fig
Does anyone feel like having the “golden” transistor label is just a dick measuring contest?

Instagram is full of these really cool builds with these rare expensive parts and I love looking at them but really wonder how they sound.

I personally could give a shit less if its vintage period, oc75, guaranteed to make your dick bigger transistors, its about how the circuit sounds.

A friend loaned me there nkt Analogman Sunface and it sounded really nice. Do you now what sounded better? My Monsterpiece GE fuzz and he is just using whatever and it sounds glorious.
 
Does anyone feel like having the “golden” transistor label is just a dick measuring contest?

Instagram is full of these really cool builds with these rare expensive parts and I love looking at them but really wonder how they sound.

I personally could give a shit less if its vintage period, oc75, guaranteed to make your dick bigger transistors, its about how the circuit sounds.

A friend loaned me there nkt Analogman Sunface and it sounded really nice. Do you now what sounded better? My Monsterpiece GE fuzz and he is just using whatever and it sounds glorious.
I mean if it's a guarantee......
 
I believe it more to be a Iceo limit in most cases, but I am a non-trained novice. When I trade or offer transistors I want that person to get the best piece or set for their build, so if there is a known or preferential threshold for leakage, because Ge is leaky, so if it is a circuit with Ge, there simply will be leakage (unless you buy in lots of 1000 and only save 0 Iceo samples), sooooooo......how much leakage are you willing to accept?
 
Does anyone feel like having the “golden” transistor label is just a dick measuring contest?

Instagram is full of these really cool builds with these rare expensive parts and I love looking at them but really wonder how they sound.

I personally could give a shit less if its vintage period, oc75, guaranteed to make your dick bigger transistors, its about how the circuit sounds.

A friend loaned me there nkt Analogman Sunface and it sounded really nice. Do you now what sounded better? My Monsterpiece GE fuzz and he is just using whatever and it sounds glorious.
If I were buying a pedal I would agree. For building a pedal....I like my mojo...because it's my way of adorning my work (at least for the moment). Seeing that capacitor I tracked down and waited for just for this build is likened to tying flies, building a model, et al. It's not about any better sound (necessarilly) or magic....or is it?
 
If I were buying a pedal I would agree. For building a pedal....I like my mojo...because it's my way of adorning my work (at least for the moment). Seeing that capacitor I tracked down and waited for just for this build is likened to tying flies, building a model, et al. It's not about any better sound (necessarilly) or magic....or is it?
Very interesting perspective, I would have never thought of it that way. My caveman brain just plugs in, turns up and bends out of tune until i get that chill.

That is the way i measure leakage, if its coming out my d
 
Very interesting perspective, I would have never thought of it that way. My caveman brain just plugs in, turns up and bends out of tune until i get that chill.

That is the way i measure leakage, if its coming out my d
That's where the rubber meets the road. When you rock it, none of the pretty matters!


Wait, we ARE still talking guitar pedals...right? 😎
 
It’s definitely become a thing to produce these fuzzes with pulled parts and NOS transistors. So much so that you see people building with whatever is available from the old suppliers and charging premium.

Many times they don’t sound like the originals. For instance, if I hear a Fuzz Face or MK II that sounds like a splatty, spluttery cacoaphoby, I know that person does not know what they are doing. This units in particular, along with the 1.5 and Supafuzz, are what I refer to as “refined” fuzzes.

Frankly, all the NOS Mojo fuzz builders now seem to be going for the sputtery, glitched out thing. I understand that some fuzzes definitely fall in that category from the standpoint of vintage units but it does lead to ear fatigue and has on more than one occasion led me to question whether my builds sound right.

In the end, I think the market is driven right now by people who don’t really know what fuzz is supposed to sound like and the makers are simply providing that.

I could be totally off base as well.
 
I don’t think fuzz has a single description. The Fuzz-Rite and the FY-6 are both fuzzes, yet sound nothing alike. I think the current fascination with the glitch-tone fuzz comes from the influence of the Fuzz Factory…a pedal which is very popular, yet impossible to get. So now everyone is doing glitch monsters to compensate.
 
Back to my OP.

Let’s take a circuit I know pretty well: The Classic Tonebender MK II.

If you take 3 middle of the road OC75s with medium to high leakage, that stock circuit will bias up to vintage spec almost as an afterthought. The OC75 does have a desirable and definite upper frequency thing going on. Let’s bookmark that.

Now you can pretty much take any low leakage Germanium transistor in a reasonable gain range (60-100 hFE) and tweak the circuit for bias and get most of the way there. What you can them do is tweak the input, treble bleed and emitter caps to get back some of that OC75 “zing”.

What you get with the tweaked circuit that you don’t with the vintage spec is temperature stability. You can make the tweaked circuit sound like the Vintage spec but the Vintage spec will never have the bias stability that the tweaked circuit will.

That’s my whole beef with leakage in a nutshell.
 
Some of my GE 2N169s have been < 10 micro amps!

I could not get them to bias in a MK II. I’ve found 15 micro amps as a sort of lower cutoff.
There you go, you answered your own question. Sometimes a leaky transistor needs to be used in order to bias the transistor correctly in the circuit so the transistor "turns on" and conducts in order to pass signal. MK2 is an example of this. Cathode-base resistors can be used but I feel it's a "cheat" and a deviation of the original circuit.
 
Back
Top