Germanium Transistor Testing Figures

[zombiwoof]

In regard to Ebay sellers that sell tested germanium transistors, I have a couple of questions.

Some of them show Hfe results at both 500 micro amps and 1milliamp collector current.  Which am I supposed to be most interested in, for use in a germanium fuzz?

Also, they give another figure, "Icbo (uA)".  Is this related to leakage?  I know from the GEO info that anything over 300uA is not advised, but the figures I see here are like 1.0-1.9 etc.  What exactly does that mean?

I've tested trannies before, but not having a great electronics background, I'm trying to understand these figures.

Thanks in advance,
Al

[CynicalMan]

From http://www.geofex.com/Article_Folders/mandt.htm

"The Reverse Collector Saturation Current Icbo
The collector current when the emitter current is zero is designated by the symbol Icbo. Two factors cooperate to make Icbo larger than Ico. First, there exists a leakage current which flows not through the junction but around it and across the surfaces. The leakage current is proportional to the voltage across the junction. The second reason why Icbo exceeds Ico is that new carriers may be generated by collision in the junction transition region, leading to avalanche multiplication of current and eventual breakdown, as discussed in Sec. 6-9. But even before breakdown is approached, this multiplication component current may attain considerable proportions.

At 25C, Icbo for a germanium transistor, whose power dissipation is the range of some hundreds of milliwatts, is of the order of microamperes. Under similar conditions a silicon transistor has an Icbo in the range of nanoamperes. "


I'm not sure, but I believe the parameter you want is Iceo.

[zombiwoof]

Thanks for the link, but I have to admit that most of that language is over my head. 

Here's a link to one of the auctions (in this case for SFT353 trannies).  These trannies both have an Hfe of 122, and the Ucbo (uA) is given as 3.32 and 3.43.  So what exactly does that mean to me about this pair of transistors, in terns of their usability in a fuzz?

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=350217526381&ssPageName=STRK:MEWAX:IT

I'm trying to find out if that figure will tell me anything about how much leakage any trannies may have that guys like this are selling.  For instance, I bought this pair of AC128K's:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=380207759185&ssPageName=STRK:MEWAX:IT

As you can see, the Hfe measurements given (at 500uA and 1mA) were 73 and 84, and the Icbo (uA) figure in both cases was 1.9.  I haven't done my own testing for leakage yet, as I just got them, but I did stick them in my meter, and after waiting for them to stabilize (about 10 minutes before they stopped fluctuating), I got Hfe 130 for each of them.  Without testing the leakage, that just seems like a big difference from the seller's true Hfe measurements.  In the past, when I've tested my own trannies, I don't think I ever had a difference between total Hfe and true Hfe as that.  So I'm wondering if that Icbo figure might have been a tip-off that these were pretty leaky (which it appears they are), or is it really not related.

Maybe this is a dumb question, but as I said, I don't really have an electronics background, and have just had a hands-on approach to this stuff over the years, and learned a little here and there and I went along.  I've waded through a lot of stuff at GEO, and other similar sources.  Maybe someone could point me to an online reference that would explain some of this in an easy-to-understand way (in reference to transistors)?.

Thanks,
Al

[jrod]

Hey man,
I may not be able to help you much, but I had these same questions regarding the sellers.

I think what it boiled down to was the fact that we don't know how they test them. We use RG's method or a Peak Atlas, but they are probably doing something different. I have bought from bgebus before and after testing them using RG's method the hfe matched very closely to what they listed in their auction at 1mA.  I think they had icbo listed at around .4uA (I can't remember, honestly) and they were under 300uA when I tested them. In fact, IIRC, they were 150 or so uA. I am with you though regarding the the language and details. It's over my head, too.

I have some of those ones you are looking at. They are nothing special.

The AC128's should match what the listing stating at 1mA (84 Hfe), I suspect. But, I am very curious what the leakage is. If they are leaky, they would probably work great in a Tone Bender MKII. I would love to know what your test results are. When you have a chance, please post the results.

Good luck. Hope this helps a little. 

[zombiwoof]

Hey man,
I may not be able to help you much, but I had these same questions regarding the sellers.

I think what it boiled down to was the fact that we don't know how they test them. We use RG's method or a Peak Atlas, but they are probably doing something different. I have bought from bgebus before and after testing them using RG's method the hfe matched very closely to what they listed in their auction at 1mA.  I think they had icbo listed at around .4uA (I can't remember, honestly) and they were under 300uA when I tested them. In fact, IIRC, they were 150 or so uA. I am with you though regarding the the language and details. It's over my head, too.

I have some of those ones you are looking at. They are nothing special.

The AC128's should match what the listing stating at 1mA (84 Hfe), I suspect. But, I am very curious what the leakage is. If they are leaky, they would probably work great in a Tone Bender MKII. I would love to know what your test results are. When you have a chance, please post the results.

Good luck. Hope this helps a little. 

Thanks a lot for your info!  That's what I needed to hear.  I was just suspicious when the total Hfe reading I got was so much higher than the figures he gave.  I'll try to get around to testing those myself, but seeing as the guy has sold a lot of parts with a good feedback rating, along with your findings, I think they'll probably test out the same as he says.  Right now, I'm just trying out some trannies in a modified early Dunlop FF, I've got 2n404's in it now, and they bias up OK, but don't sound great to me.  In the past, I've bought small lots of a few different trannies and tested them, but I get tired of getting a bunch that are either too high or too low in gain, so I thought I'd get some tested pairs to try.  I also got some 2SB176's from him to try out.   I have thought of building a Tone Bender at some point.  If I do, I've got a bunch of trannies around to try in it.

Have you bought any from Orpheus (the first link I gave)?  I've bought other stuff from him in the past and no problem, but was wondering if his transistor testing was good.

Thanks,
Al

[jrod]

I have bought a few items from Orpheus. They were an untested lot, so I am not sure what method they use to test them. Quite a few were duds, but, for the price, I got several useful ones, and made it worth the small about of dough. I really like the Japanese transistors. To me, 2sb175 and 2sb176 sound great in a Fuzz Face. The 176's are higher gain (150-195 hfe). I know Mac from this forum has posted his results with different Japanese transistors and their variation from brand to brand. Anyway, I love my 175's in a Fuzz Face. I have some house numbered 2n404's that sound great in a Fuzz Face and Tone Bender. I am not familiar with the Dunlap circuit, though.

Let know how your testing comes out. I am curious how they compare to their listings. Good luck with your build, too!

[zombiwoof]

I have bought a few items from Orpheus. They were an untested lot, so I am not sure what method they use to test them. Quite a few were duds, but, for the price, I got several useful ones, and made it worth the small about of dough. I really like the Japanese transistors. To me, 2sb175 and 2sb176 sound great in a Fuzz Face. The 176's are higher gain (150-195 hfe). I know Mac from this forum has posted his results with different Japanese transistors and their variation from brand to brand. Anyway, I love my 175's in a Fuzz Face. I have some house numbered 2n404's that sound great in a Fuzz Face and Tone Bender. I am not familiar with the Dunlap circuit, though.

Let know how your testing comes out. I am curious how they compare to their listings. Good luck with your build, too!

The 2SB176's I got were around Hfe 120 according to his figures, as I recall.  I got them because many guys here said they sound good in a FF.  Have to get around to trying these things out!  I've got two FF pedals to try them in, one of them is that reviled red Dunlop Jimi Hendrix fuzz (the little MXR-sized one), I'm convinced that it can be made to sound good with some properly biased trannies, and it's got the built-in buffer to make it work with a wah.  I checked the bias on it stock and it's way off, even with noodling with the internal trimmer that's in there.  I can't believe they didn't try to bias these things at the factory, maybe people wouldn't hate them like they do.  The stock trannies are NTE germs.  The other one is an old red Dunlop full-sized FF, it's an old one with the regular case mounted pots, switch, and jacks, not the one with everything board mounted.  I already modded it with an .01uf output cap instead of the stock .1, that got rid of the excessive bass, I replaced the resistors with carbon comp (yeah, I know...) which warmed it up a bit, and put in 2N404's that biased up right.  It still just doesn't "have it', so I'm going to try different trannies.  I'm also experimenting with a small treble bypass cap on the volume pot, I got that idea from looking at the schematic for the Roger Mayer Classic Fuzz, he's got an .002uf  there.  It helps keep the high end when you turn down the volume pot, like it does on a guitar.  I've also got some Russian trannies to try for Q1.  We'll see.....

Thanks,
Al

[jrod]

Don't hesitate to try some lower gain transistors than the usual FF suggested gains. I have found that 60ish for Q1 and 90ish for Q2 can sound great and clean up better. My personal FF taste is less of the higher gain squishness for more bite. That would be great if you could get both of your Dunlap models to work well. Your right, it may be just getting the right transistor biased well. What Russian ones do you have? I love GT404/402's!

[zombiwoof]

Don't hesitate to try some lower gain transistors than the usual FF suggested gains. I have found that 60ish for Q1 and 90ish for Q2 can sound great and clean up better. My personal FF taste is less of the higher gain squishness for more bite. That would be great if you could get both of your Dunlap models to work well. Your right, it may be just getting the right transistor biased well. What Russian ones do you have? I love GT404/402's!

I've got a bunch of 404's.  I threw a few into my meter and most were around the 70-80 mark, as I recall.  Figured they would be good for Q1.  I think I bought some others, maybe 308b's or something a while back.  I'll have to find those and check them again. And as for those small untested lots of trannies from sellers that also sell tested pairs, I have to think that the "untested" ones could really be the rejects from their testing.  I don't know if this is true, but I don't think they'd toss away the tested ones that don't meet a certain spec.  That doesn't mean they wouldn't be good for something other than a Fuzz Face (which the gains of the pairs are usually shooting for), but I wouldn't think that I'd find a bunch that are in the right gains for a Fuzz Face per se.

Anyway, thanks for your help.

Al

[zombiwoof]

Oops, just realized the GT404's are NPN, both of the Russian ones I have are PNP.  The 70-90 Hfe ones I have are the 308B's.  Going to try them in Q1 on the FF's.  Can't remember what the other ones I got are, they are packed away somewhere.

Anyway, thanks for the info.  Still don't fully understand the Icbo thing, I'll have to find a simple explanation somewhere.

Al

[jrod]

Oh, yeah. The 404's are NPN. They sound really good. I love 'em. I just got a new box 1T402's to test. They are a full box of military spec from 1984. The 308B's I have all tested on average around 50 somthing hfe. Very low leakage though. I also got some P416 that are shown in an old Russian fuzz schematic I found. I think they tested around 70-80's with low leakage.

I don't understand the way Icbo as listed in their auctions either. I know bgebus will send you their testing method if you email them through ebay. May they will explain it you. I have emailed them and they were always very helpful.