Great Jumpin Germaniums! (Ge Diode and Transistor Reviews & Discussion)

[iggy]

hopefully this is a good place for this question. Just got a order of supposedly germanium transistors in. And one group testes as silicon, and fairly high gain. with a very wide range of gains. But some also show leakage. Have any  of you ever seen silicon show a leakage on a Atlas testor. And they do change with heat. very odd. not sure what i got. so thought i'd ask the experts here.

[rocket88]

hey buddy. two questions come to my mind.
1: how are you testing them? are you using a DCA55?
2: what are the part number of the transistors?

i've had my DCA55 test some transistors and it says they're Si, when i know they're Ge. i've also had it test some Si transistors, and it says they're darlington, but again i know they're not. i could be wrong, but in my experience i haven't seen or come across Si transistors that have leakage. i'm sure someone with more experience then me would know better, i'm just going off what i've dealt with.

[dbat69]

I would echo what Rocket has said.  THe DCA can give some slightly false readings when testing transistors when it says "Silicon" or "Germanium".  It obviously uses certain electrical properties to determine what type of transistor it is and if they reflect an Si transistor it will say that even though it is actually a Ge transistor.  Even with the same transistor, say a 2N404, it can say Si with one and Ge with another.  

I have not seen a Si showing leakage yet (which is one of the reasons why Ge got phased out).  Si transistors also are very temperature stable unlike Ge (another reason why they were replaced by Si transistors).

[rocket88]

Just one other thing about the dca55 and false readings. I've had it tell me that some Ge transistors are bi-color LEDs. It's probably due to having a very low hfe, but still. So it can be kinda off some times.

[dbat69]

Haa ha, yes, I've had some of my Ge transistors reported as diodes - they are the duds.

And a word of Caution if you are getting, or testing Russian GT313 transistors, these little blighters are very susceptible to static, even the slightest bit of static will kill them.  I found this out when testing them.  Treat them like they have come from Chenobyl

[rocket88]

Haha.

[iggy]

Hey guys thanks for the information. Pretty much matches my experience with the DCA55. So there is a good chance these are germaniums.  Still a bit odd. No part numbers and some are incredibly high gain and low leakage. I have tested units up to 530 hfe with only .04 mA leakage as tested on the Peak. I need to see how they sound. They certainly fit all the traits of a germ. hfe varies with temp. and they show leakage. which also changes with temp. I'm feeling a high gain germ Muff in my future.

[rocket88]

i have some Ge's that are like that, in perticular the GT109's that i have. they are super tiny, and i have some with a gain over 600!!!! when i saw that i was all like . so i'm sure they're are some other Ge's out there that can reach high gain no problem.

[iggy]

Rockett have you built anything yet with those high gain germs? any recommendations? I was thinking about a Muff. But these are PNP. And I think all the Muff circuits I've seen are NPN.

[rocket88]

Mine are pnp as well. I haven't used the really high gain ones yet, but might be good for a pnp muff variant. I used one that's 110hfe and one that's 350hfe in an oracle. Sounds ridiculous. Tones of sustain and gain, a little bitting and has sorta Si vibe to it, when compared to the mp41a's I tried, which are a little darker and warmer sounding.

I think there is a pnp miff variant, the diceworks muff diver is pnp, here.

[iggy]

sweet!! V1 Muff, in pnp. perfect. thanks for that link. I missed that one. I think I know my next project. I'll see what a full GE muff does. Might end up a mix of Si & GE. but time to see what this sounds like as full GE first. woohoo!!!!

[Silver Blues]

If you want to do a PNP Muff then just build a voltage inverter to keep it daisy-chainable. Even if you didn't need that feature a PNP Muff is just as feasible as converting any NPN circuit to PNP, just flip all polarities and go to town.

[iggy]

Silver,

that's a great idea. I wouldn't have thought of that. I've got a ton of PNP's and could try slipping them into more modern circuits. Thanks for that suggestion.

[Travis]

I saw this last night at the Griffith Observatory. Couldn't resist taking a picture! I thought it was neat to see germanium in it's raw form since I'm so in love with it's sonic glory

[rocket88]

That's beyond neat. That's fucking awesome. Now we all can see why germanium makes pedals sound amazing. It just looks it raw.

[dbat69]

It's pure rock man

Thanks Travis, first time I've seen an image of Ge in its natural state.

[alex.s]

Finally got round to testing my small stash, mostly of GT308V (plus a few goodies from everyone's favourite eBay seller..!).

I used the RG Keen germanium tester method, using a four pin socket in a BCBE configuration to accommodate both BCE and CBE pinouts without twisting and/or heatshrinking

First of all I can confirm what I alread read, those transistors are remarkably consistent and with remarkably low leakage.

My 20 GT308Vs all ranged between 0.012 and 0.036 in leakage, which is great, but somewhat disappointingly the consistency extended to their gains, meaning I will need reinforcement to actually get some matching done.

The gains ranged between 46 and 69, with only five of them in the 60 or more territory.

Goodies wise, I had:

two Texas Instrument 2N1307s one hfe:81 leakage:0.057, the other hfe:175 leakage:0.182

three GT2 308 (gains: 108, 113, 159 with leakages between 0.065 and 0.146)

six SFT308 which were quite hit and miss (gains between 40 and 140, leakages between 0.094 and 0.995, plus one wildly off the scale..!)

All in all it's a good starting stash, but will need to source some higher gain PNPs to pair up as I'm stuck with mostly low gain trannies.

Just thought I'd share these finding as some people might be interested in either comparing their result with the same kinds of transistors or looking to place an order.

Sonic findings to be posted at a later date..! :)

Ending with a question for everyone... I noticed a pretty solid trend in terms of gain/leakage ratio, in that the vastest majority of my higher gain trannies also had slightly higher leakage compared to their lower gain counterparts... Any explanation for this? Is this a common thing?

Thanks, and hope someone might benefit from this (small) contribution to the cause of weird and wonderful soviet Germaniums..! :)

[alex.s]

PS: the joys of eBay..!

[rocket88]

Awesome contribution!! To answer your questions. First, I have noticed the same thing as far as gain/leakage with certain transistor models. But, there is definitely a trend in higher gain have higher leakage, which makes having high gain Ge transistors with low leakage such unicorns. If you want higher gain gt308's look for 1t308v's. They're the military version, and have higher gains, I have some in the 90's & 100's, and leakage is pretty much 0, oh and most are 70-80hfe.

Secondly, I think I'm going to have to buy that. I don't know how to explain why, but I feel like I  NEED to have a piece of Ge in my work space. Maybe if is hold it over my builds and say some magic works if will add it's Ge mojo to every pedal I make.

[IvIark]

alex.s wrote

All in all it's a good starting stash, but will need to source some higher gain PNPs to pair up as I'm stuck with mostly low gain trannies.

Try some of these: http://www.ebay.co.uk/itm/251637232034

I bought 150 of these AC125K's from him a couple of years ago for exactly the same reason as you (in fact he has some 150 qty auctions in his shop now) and they're great.  Leakage isn't as low as many of the Russian ones but they're still very low and non-excessive leakage really isn't the most important consideration, especially as some circuits benefit from it.

But don't give up on the GT308s until you've heard them in a circuit.  I've built Fuzz Faces with GT308s around 50hfe which have sounded fantastic.  Gain of 80 and 120 isn't the most important thing either, it's building a pedal which sounds great with your gear irrespective of the numbers.