AC125 Ge Transistors

[motterpaul]

I got a big batch of AC125 Ge Transistors. They are marked AC125 Uz VI. So, I have done some research and from what I have read & seen on YouTube these are supposed to be decent for making fuzztones. There is one guy on YouTube who does a fuzz face transistor shootout and he seems to love these AC125s. https://www.youtube.com/watch?v=GYOMjPmtMi4 - his TB name is DIY Guitar Pedals

In any case, I have tried a couple of circuits, and so far I get some fuzz, but not that I would say I am in awe of... Its not bad, but also not great.

He seems to like the model VI (like I have) of the AC125s the best. So, I am wondering what else I need to know.

1) How does biasing work with GEs. I haven't really seen it discussed. I know I put in a 10k trimpot, and no matter what, it sounded best on 10k, so I replaced with a 20k pot, and that sounded even better at 20k. It just makes me wonder what I'm looking for...

In any case - I am looking for people who know these transistors, where have you had success (what builds?, what bias, what tricks?).

Thanks in advance.

[motterpaul]

I also had a question about power. I thought it was a simple matter to just make a power inverter, but the one I made gave out the correct voltage at first, but then I did something wrong with it, and the 7660s started frying, so... Now I am on a battery. It really is just a matter of using the black lead for -9v, and the red lead for ground, right? (from a battery), and when you ground the rest of the circuit (jacks, etc) it is to the red lead, right?

If you use a DC adapter, can you just flip the two "poles" - like on a breadboard, just use what was the - as the + now, and use what was + as the ground. No other tricks to it. (My power inverter seems to have fried, but I will build another one, but for testing I want to be able to know what I am doing so I don't fry any more).

This guys videos are really good - by the way: https://www.youtube.com/user/chromespherecom/videos

I believe he is pretty well-known around the interwebs. And he also said he has fried quite a few power inverters.

[motterpaul]

Well Paul (yes, I am posting to myself)...

Just FYI: I made a modification to my Ge tester, because I heard it is very important not to apply any stray voltage to the Ge before it is correctly inserted, so I put a momentary switch on the -9v so no voltage goes to the sockets while I am changing transistors and I can't test either range unless I am pressing the momentary switch.

I am now finding most of these AC125s have HFEs in the 120 - 180 range. I have seen as high as 201, and in general they average 150. Leakage almost always goes up with higher HFE. Some are still below 100 but quite a few are above 150.

I built a Ge Overdrive....



And I find that with the highest HFEs you get more fuzz, and as you increase the bias you tend to get more bass and also more attack on each note (a "pop") - if you roll back the bias you get a smoother sound, but this circuit still sounds more like a fuzz than an overdrive. On full bias on a high HFE the bass is a little too much (especially because it is fuzzy).

But you get more range of tone with a higher HFE - because if you roll back the bias it mostly just sounds like a lower HFE Ge AC125.

[Beaker]

So which transistors did you use in the EHX - and did you build it stock i.e. with no mods or alterations?

You are in uncharted territory there, as there are no comments on the layout.

[motterpaul]

Hmm, you must be looking at a different version, the one I used has 26 comments:
http://tagboardeffects.blogspot.com/2014/09/ehx-germanium-overdrive.html

I used an AC125 with an HFE 180 first, and it did sound more like a fuzz than an OD. Then I tried one with about 110 HFE and it is less hairy. For the NPNs I used 2n5088s (I don't have 5087s, I have 5089s which tend to be even more powerful than my 5088s).

I just realized I did NOT use a reverse pot on the bias (I used a standard linear), so when he (in the demo) turns the bias from full CW to full CCW you can hear how the bass comes out. That's what mine does, except my pot works the opposite way.

The higher the HFE of the Ge the more you get this bass effect on the bias. Use a lower HFE and it sounds about the same as this pedal does with the bias at 50% or less.

[motterpaul]

By the way - Mirosol left a great spreadsheet for calculating HFE and leakage in this post under "Germanium Transistor tester":  I have been using this (the xls) and it makes it really easy.

mirosol 28 August 2012 at 15:55

Anyway. Here's the spreadsheet for the hFE calculations. Simple, quick and dirty. But saved me a lot of hazzle.

http://mirosol.kapsi.fi/varasto/boxes_other/Germ_Calc.ods
http://mirosol.kapsi.fi/varasto/boxes_other/Germ_Calc.xls
+m

[Beaker]

2N5088s are NPN, 2N5087 are PNP.

[motterpaul]

Wow... I didn't notice. In the comments I saw it is a positive voltage effect, so I assumed the non-germs were NPN, but you are right.

Everything seems to be working. I could put some PNPs in there and see if it sounds different - Miro says most any PNP will do ( I just noticed). I don't have any transistors turned around.

I would be very surprised if the only transistor working is the Ge but I guess it is possible.

[Beaker]

I would just throw any PNP transistor you have (minding the pin-out) in place of the 2N5087, and see what happens.

[motterpaul]

Okay - it does look like only the germ was working before because it sounded a LOT more like a fuzz before. I would actually say it was comparable to the demo (which people in the comments say is not the case).

Now it sounds largely like just a boost with some hair on it. I used 2n5401s

[motterpaul]

Wait - I forgot the bias is backwards, when I go full CCW I get a lot more hair, so I guess it is working correctly. But I would honestly say the sound the other way was not too bad, either (which I guess was just the Ge alone).

EDIT: also - it had a lot more fuzz before than it has now (which is logical if it was just the Ge before).

[Beaker]

2N5401 are much lower gain than 2N5087, so it might well change character with higher gain ones.

[motterpaul]

I could find the few 3906s I have but I would really have to dig around.

BTW: thanks for the input. I will get around to finding some higher HFE PNPs and see how it sounds.

[Beaker]

3906s should be ideal.

[motterpaul]

I looked at my spare trans stash - can't find them, but I tried some 2n4403s and what you get is more gain out of the Ge -  back to a fuzzier sound. Not super fuzzy, but a decent sound that I could see working in a grunge rhythm guitar (smells like teen spirit).

Thanks for your input, I learned a lot that I probably would have missed.

[Beaker]

Watch the Pro Guitar Shop demo on Youtube, as it explains the controls really well, and clearly demonstrates how interactive the controls are.

Judging by the demo, I would put this in the "Bluesy" overdrive camp - not really fuzz, but not Tubescreamer type overdrive either. If you are expecting "Rock" Tubescreamer type tones, then you are going to be dissapointed.

It strikes me as a good test-bed pedal for a little experimentation with different transistors - you could try a Russian germanium GT307 or MP39 for example. It would also be an easy conversion to run it using NPN transistors too. That would give you a whole other set of transistor options, e.g. 2N5088 for the silicons and a MP38 for the germ.

If you want a more "classic" sounding overdrive circuit to try your transistors on, then try the Nick Greer Green Giant. Super low parts count, but awesome sounding. The best thing about it is that you can easily tune it to your requirements, by changing the transistor and diode choice.

I have built several - both "original" silicon, PNP germanium versions, and NPN germanium (Germanium Giant) versions, as well as hybrid germ transistor/silicon diode versions.

They ALL sound slightly different, and they all sound great! It's also a REALLY good way to learn the different characteristics of some of these transistors and diodes.

[motterpaul]

Beaker thanks for your advice. I already had some 3906s on order from a few days ago so they should come soon.

What I hope to understand from this circuit is:

1) why is this not a neg voltage circuit like most PNP layouts?

2) I believe the idea here is to use the Ge for an overdrive sound (as opposed to pure gain) and the Si trannys are there to add more gain to whatever you get out of the Ge. I found that the first time out (when I mistakenly had npns) I got the thickest Ge sound. When I changed to the 5401s (low gain PNPs) I got more  boost but a lot less fuzz. When I put in some heavier gain pnps I have (4401s I think) I got more boost overall, so more gain but also more fuzz. I am hoping to confirm what I see happening here.

3) as for Nick Greer, I have heard it is good. My general goal is not to build out my pedal board but really just to build circuits and see what I like about them. I just ventured in Ge because our good friend Rockett got me interested in what they can do. Plus I heard "Nirvana" and he gets a decent fuzz guitar sound. But my goal right now is to really find the best overdrive (I like amp emulation) and see what circuit combinations make for the best tones. If that includes a Ge then cool, but so far I like FET circuits best.

I did listen to the EHX demo, and I get much the same results, when I lower the voltage and gain I get that sputtering gated effect. I really don't see why anyone would want that for more than a momentary effect, but that's just my opinion. When I raise both I get a lot more hair and boost in the bass. I also get compression and I think Ge trannies tend to compress a lot more than Si. They almost sound like real compressors (put up a Ge fuzz circuit, then put up a compressor circuit feeding an overdrive and the sound is pretty similar in a lot of ways).

4) once again, I need to study this circuit and see how it is possible to mix PNP with positive voltage instead of negative ground (I think I am saying that right). It's really kind of new territory for me.