Germanium Transistors, Noise and Leakage

[squirrels]

Tried out 49 hFe(.121 mA) and 59 hFe(.117 mA) in a Sunface. Not bad, but not great. Guess I'll order some more. Any advice on something higher gain?

Also was I doing something wrong with the GT404s?

[Travis]

Go higher on the hfe. Don't count on the first pair sounding good. It may take a while to figure out what's working for you

When you're using lower hfe transistors you might like it where Q1 is higher hfe than Q2. I just made a really cool sounding one where Q1 was like 90 and Q2 was about 45. I couldn't really tell you exactly what you want though, you need to try for yourself

In the original Ge FFs that sounded good and have been measured, the hfe is usually relatively matched for both transistors, and they usually range from the 50s-70s, with exceptions. This stuff starts to go out the window sometimes when you start substituting different transistors in place of the NKT275 (which is of course impossible to find). Some don't sound dirty enough with those specs, some have an unflattering freq response, etc. that's why experimentation is always key. With the MP21A I know that RG Keen's Fuzz Face prescription works well.

[Marbles]

I tried to make a sunface as well today, but think my hfe is not high enough?

I had some Gt308B and Gt308V. I know I should buy the Gt308V cause the hfe is higher than the Gt308B, but because the 308V has a russian B on the transistor I got confused and bought the B. Measured with the peak, they're around 50 all of them, some in the 40, some 60.

The Gt308V I have are a bit higher, but not that high. Out of a 100 pieces, most of them are between 60 and 80, a lot between 80 and 90 and the max is 95 (about 4 of them).

i learned my lesson, on ebay, when they say the range is between 80 and 200, you are getting 80 haha.

Anyways... I built the sunface with those transistors and used this to measure the resistors: http://www.diystompboxes.com/analogalchemy/emh/emh.html

They were all sort of the same, but since I needed a 8.2 k resistor judging the calculation, I used a 3,9k instead of the 2.2k (cause I thought: 5k for the sundial + 3,9k will get me to the 8.2k. that could be a wrong thought though haha).

The hfe for q1 is 69, and for q2 I have 82.

It sounds... fine I guess. I'm not sure. The volume cleaning up is amazing, but it's not ' screaming'. It definitely has the fuzz face character I know and love, but it's just.. a bit bassy and the volume maxed is about unity. I don't have a germanium fuzz face to compare it too though.

I just hoped I would have that sounds where your notes struggle to fight themselves through a thick nasty fuzz to sustain beautifully if that makes sense.


Could my gain just be too low? I ordered another batch hoping to have higher hfe, but even when buying a 100 batch, elena does not ship them in an unsealed box, so the highest are probably picked out already.

I noticed some european stores started selling the 308s as well too, for 2,50 euro each. They don't have higher hfe either.

[squirrels]

Mine didn't sound too bad with transistors in the 50 to 60 range either but not great. I found a few sellers on eBay who are selling measured and tested transistors individually so I ordered an AC125 with 133 hFe to try in my Oracle. I also got impatient and picked up an NTE102 from my local shop. Leakage was higher like my previous one (around .210 mA) but the hFe measured at 177. Unfortunately I snapped the leg on my higher gain MP21A so I put one that measured at 69 in Q1 to pair with the NTE102. Sounds much more aggressive now, can't wait to pop the AC125 in it.

My local shop also had a bag of old GE transistors being sold for a dollar each. I didn't get any because I didn't recognize them. 2N523A ring a bell? Data sheet says minimum hFe was 250... Worth trying?


Sent from my iPhone

[Travis]

Do you guys have any PNP silicon? You guys might prefer a hybrid. If you pop a 2N3906 or something in Q1, then try some different Ge in Q2, you will see the transistor selection becomes easier and you will have more gain and temp stability

If you have hfe 50-95 that should be plenty to work with. Maybe try something like 70 for Q1 and 90 for Q2 and go from there.

Squirrels the AC125 tends to be pretty leaky. It might sound good in the FF, if not I think it would be better suited as Q3 in a tonebender MKIII or Buzzaround

[squirrels]

It was guarantees at 133 hFe and .10 mA leakage. I didn't think that was too bad. I do have some PNP silicons, I'll give it a shot. Thanks.

I may actually keep the Oracle as is. It's pretty aggressive sounding and old school sounding at the same time. I have more of the MP's in the 60 range, I may try it out in the FF.

[Travis]

Yeah that's pretty low leakage. Good luck.

What is it about the FF that sounds bad?

[squirrels]

It's not bad, just a little weak for my tastes. Maybe I'm just used to silicon fuzz.

[Travis]

You have the volume on your guitar all the way up right? Are you making any bias adjustments or using the stock circuit?

[squirrels]

Ha, yeah the volume knob is all the way up. I actually checked but I don't really play around with them much.

I'm using an Analogman Sunface with the stock layout except for the clean trimmer being off board. I also just built the Creepy Fingers Si Fuzz Face (stock layout) to try out the few NPNs I have. I used a 2N1304 (49 hFe) in Q1 and a GT404B (hFe 88) in Q2. Not bad actually. A little bass heavy but I can dial it out with my guitar volume.

[Travis]

Make sure the clean pot is all the way up or omit it altogether

[squirrels]

It's all the way up but kind of unnecessary, I'll probably ditch it. I still want to try out some different transistors.

[Travis]

Yeah you gotta just have at it with a bunch of stuff until you're happy. The same transistor type with different hfe can make a world of a difference.

Keep in mind most people use the fuzz knob all the way up, volume all the way up, guitar volume all the way up, etc when getting the most fuzz out of this thing

[squirrels]

I loaded Creepy Fingers FF with a 2N1304 and GT404 to a set of BC108 and BC109. Honestly it seemed to have just as much gain but the GE's were a bit warmer and more splatty sounding than the silicons. I liked them much more plus I can daisy chain it with the rest of my stuff. The 2N1304 has a gain of 51 and the GT404 measured at 88. I may look into other NPN to try out.

It also fit nicely in a 1590A.