Low Volume, Low Gain G2

[ac30irons]

Hey all, I know bits around this have already been asked but I’m struggling for a definitive answer.

I have just built a G2 kit from a popular and very helpful uk supplier. There was a bit of mucking about with the diodes and a couple of resistors around forward voltage but it’s built. The kit was supplied with Russian D9Ks.

On power up volume has to be a least 2/3s up to do anything and the gain all up or there is little to no sound. Through a clean amp it sounds ok but there is no where near the gain there should be. I would describe it as a bit crunchy.

I have heard of something called leakage current. I’m not very technical at all and don’t really get it all.

Are D9Ks generally leaky ? Is there an Ge alternative I should just be able to drop in ?

Bloody headache this diode thing.

 

[ac30irons]

1N4148 Si Diodes have forward voltage around 0.70 Volts.
1N34 Ge & D9K's Diodes have forward voltage around 0.30 Volts.
So basically when used as a pair of Clipping Diodes the Volume is decreased by Half when Ge diodes are used.
The Circuit has to be designed to compensate for this correctly & the G2 has not hence the Aion mod to try & improve it.
YourD9K's maybe as low as 0.25v & this will kill more volume also but it doesn't explain your Low Gain issue, that is why we need to know that your Resistor values are correct & Pictures are the Best option for us to verify this.

Here is the build. You can see I have Si in D1 and D2 to get it doing something and I had to make an 18k for the mod but all is testing correct for placement.

Looks like I’m going to have to put an order in to the US for some mojo ITT 1N34A.

 

[cdwillis]

Here's my unsolicited advice, but hopefully helpful advice.

The germanium diodes only pass half or so the volume that a silicon diode would. The sound is probably just as distorted, but much lower as far as actual volume. Using a magic 1n34a isn't going to change anything because it's going to have about the same clipping threshold as what you've got now. This is about forward voltage, not leakage. A germanium device can have leakage, but people talking about leakage in germaniums are probably talking about germanium transistors in vintage fuzz circuits, not clipping diodes.

I'd change the 8.2k in the last transistor stage to 10k or 12k and you should get considerably more output volume. That or try lowering the value of the 620 resistor. Also if you're using a logarithmic volume pot switch to linear. It will seem louder lower on the dial because of the taper of the pot.

 

[music6000]

Can you confirm the Colour Bands on the 2 Resistors that make up 18K?

 

[Laundryroom David]

EDIT: THE BELOW IS WRONG PLEASE DISREGARD

You could pop some bat41s in there. That’ll help you determine for your own ears whether the leakage issue is an issue or not.

That said, if you aren’t getting enough signal to get the diodes clipping, this is all academic.

 

[ac30irons]

You could pop some bat41s in there. That’ll help you determine for your own ears whether the leakage issue is an issue or not.

That said, if you aren’t getting enough signal to get the diodes clipping, this is all academic.

Thanks for the reply.

I can certainly get these. Please excuse my ignorance, how would these determine if it’s a leak issue? I believe they are silicone also and the thing that makes the G2 a G2 is the germanium diodes.

 

[ac30irons]

Can you confirm the Colour Bands on the 2 Resistors that make up 18K?

From what I can make out with my dodgy peepers. I wired them in series and measured with the DMM and it was bang on 18k.

1 - brown red black green brown
2 - brown brown black black gold

I have always struggled with the bloody colours.

 

[Laundryroom David]

Thanks for the reply.

I can certainly get these. Please excuse my ignorance, how would these determine if it’s a leak issue? I believe they are silicone also and the thing that makes the G2 a G2 is the germanium diodes.

EDIT: THE BELOW IS WRONG

No problem.

If it sounds good with bat41s, leakage is not the key. If it sounds shitty with bat 41s, then yeah maybe leakage is important to the circuit. Designing a circuit around leakage of germanium diodes sounds bonkers to me, but I am not an engineer, so take my thoughts with a grain of salt! But even the si diodes won’t clip if the signal lacks sufficient amplitude to “trigger” the diodes.

 

[jubal81]

No problem.

If it sounds good with bat41s, leakage is not the key. If it sounds shitty with bat 41s, then yeah maybe leakage is important to the circuit. Designing a circuit around leakage of germanium diodes sounds bonkers to me, but I am not an engineer, so take my thoughts with a grain of salt! But even the si diodes won’t clip if the signal lacks sufficient amplitude to “trigger” the diodes.

This is backwards.

Bat41s have a similar fV and almost no leakage, so if it sounds good with those, leave them in there and call it done.

The feedback resistor in Muff gain stages (usually 470K) sets the amount of negative feedback in that stage, reducing gain. A leaky pair of diodes in that feedback position is like adding a much smaller value resistor in parallel, adding a lot more negative feedback and vastly dropping the gain.

In this day and age, trying to find 4 low-leakage germanium diodes is a fool's errand, especially considering the non-Russian makes usually sell for $1 or more each. It's entirely plausible you could spend $400 on diodes and still not have 4 with low-enough leakage.

Schottkey diodes will sound extremely similar anyway. However, be aware that they also exhibit leakage. For example. the 1N5817 used for polarity protection is about as leaky as Russian Germanium diodes.

 

[Laundryroom David]

This is backwards.

Bat41s have a similar fV and almost no leakage, so if it sounds good with those, leave them in there and call it done.

The feedback resistor in Muff gain stages (usually 470K) sets the amount of negative feedback in that stage, reducing gain. A leaky pair of diodes in that feedback position is like adding a much smaller value resistor in parallel, adding a lot more negative feedback and vastly dropping the gain.

In this day and age, trying to find 4 low-leakage germanium diodes is a fool's errand, especially considering the non-Russian makes usually sell for $1 or more each. It's entirely plausible you could spend $400 on diodes and still not have 4 with low-enough leakage.

Schottkey diodes will sound extremely similar anyway. However, be aware that they also exhibit leakage. For example. the 1N5817 used for polarity protection is about as leaky as Russian Germanium diodes.

How is it backwards? I was saying the same thing just in a different way.

 

[jubal81]

How is it backwards? I was saying the same thing just in a different way.

If replacing the diodes with low-leakage diodes and it solves the problem, then it is a leakage issue in his build.

That's why this is backward:

If it sounds good with bat41s, leakage is not the key.

 

[Laundryroom David]

If replacing the diodes with low-leakage diodes and it solves the problem, then it is a leakage issue in his build.

That's why this is backward:

What typically has more leakage and more variable leakage among them? Bat 41s or 1n34a? I am legit curious and don’t know the answer.

 

[jubal81]

What typically has more leakage and more variable leakage among them? Bat 41s or 1n34a? I am legit curious and don’t know the answer.

Bat 41 has almost no leakage for our purposes. 1N34a is a number applied to a wide variety of Ge diodes and will all have significantly more leakage.

Bat46 and Bat54s are also excellent low-leakage Ge replacements.

 

[ac30irons]

Here's my unsolicited advice, but hopefully helpful advice.

The germanium diodes only pass half or so the volume that a silicon diode would. The sound is probably just as distorted, but much lower as far as actual volume. Using a magic 1n34a isn't going to change anything because it's going to have about the same clipping threshold as what you've got now. This is about forward voltage, not leakage. A germanium device can have leakage, but people talking about leakage in germaniums are probably talking about germanium transistors in vintage fuzz circuits, not clipping diodes.

I'd change the 8.2k in the last transistor stage to 10k or 12k and you should get considerably more output volume. That or try lowering the value of the 620 resistor. Also if you're using a logarithmic volume pot switch to linear. It will seem louder lower on the dial because of the taper of the pot.

View attachment 49460

Thanks bro, the reason I’m banging on about leaks is because it’s the same reported problems that have previously been reported with that as the issue.

The knowledge you guys have is crazy, I can’t even read a schematic !!!

 

[music6000]

From what I can make out with my dodgy peepers. I wired them in series and measured with the DMM and it was bang on 18k.

1 - brown red black green brown
2 - brown brown black black gold

I have always struggled with the bloody colours.

1 - brown red black green brown = 15K
2 - brown brown black black ORANGE = 3K if that's what you have to make 18K?
And your Colour Bands you listed are Back the Front for reading resistor values.
Brown, Green, Black, Red, Brown = 15K
Orange, Black, Black, Brown, Brown = 3K so I hope this is what you Have!???

Here is a Resistor Calculator I suggest Members use.
You can do a Test Run by putting the Colours listed above & you should get the Correct Values!
Click on Bands gives you 5 colours:

Resistor Calculator

Resistor Calculator is an App developed by CherryJourney to help the user find the value of a resistor.

resistor.cherryjourney.pt

 

[ac30irons]

This is backwards.

Bat41s have a similar fV and almost no leakage, so if it sounds good with those, leave them in there and call it done.

The feedback resistor in Muff gain stages (usually 470K) sets the amount of negative feedback in that stage, reducing gain. A leaky pair of diodes in that feedback position is like adding a much smaller value resistor in parallel, adding a lot more negative feedback and vastly dropping the gain.

In this day and age, trying to find 4 low-leakage germanium diodes is a fool's errand, especially considering the non-Russian makes usually sell for $1 or more each. It's entirely plausible you could spend $400 on diodes and still not have 4 with low-enough leakage.

Schottkey diodes will sound extremely similar anyway. However, be aware that they also exhibit leakage. For example. the 1N5817 used for polarity protection is about as leaky as Russian Germanium diodes.

Cheers dude, Iv ordered some bat41~they where really cheap.

Been speaking to chuck d bones who recommended some 1N34a by ITT but need to them froM the US. See what they do.

I really want Ge’s to keep it a true G2 but the cost is outweighing the reality. I would be cheeper buying the real deal.

 

[music6000]

Cheers dude, Iv ordered some bat41~they where really cheap.

Been speaking to chuck d bones who recommended some 1N34a by ITT but need to them froM the US. See what they do.

I really want Ge’s to keep it a true G2 but the cost is outweighing the reality. I would be cheeper buying the real deal.

I understand your Frustration & you want the Gilmour gear but the Cornish stuff is way over rated & you may still be disappointed!
He put's a seal that says warranty is void if you open it up to have a look.
That is telling me you might not like what you see & payed for!

 

[ac30irons]

1 - brown red black green brown = 15K
2 - brown brown black black ORANGE = 3K if that's what you have to make 18K?
And your Colour Bands you listed are Back the Front for reading resistor values.
Brown, Green, Black, Red, Brown = 15K
Orange, Black, Black, Brown, Brown = 3K so I hope this is what you Have!???

Here is a Resistor Calculator I suggest Members use.
You can do a Test Run by putting the Colours listed above & you should get the Correct Values!
Click on Bands gives you 5 colours:

Resistor Calculator

Resistor Calculator is an App developed by CherryJourney to help the user find the value of a resistor.

resistor.cherryjourney.pt

That’s what I have. Always get gold and orange mixed up

 

[ac30irons]

I understand your Frustration & you want the Gilmour gear but the Cornish stuff is way over rated & you may still be disappointed!
He put's a seal that says warranty is void if you open it up to have a look.
That is telling me you might not like what you see & payed for!

I get a bit obsessive end up spending more time sorting kit than playing through it !!!

 

[ac30irons]

$20 to ship something as light as feather to UK. Worlds gone mad.

 

[ac30irons]

Quick update - Stuck in some BAT41s and what do you know, it works as it should with the only issue being it pretty much sounds like a muff.

Some AA143s should arrive tomorrow, let’s see what they do !