Germanium Transistors! Help!

[wg_ski]

I want to used them as buffers for amplifier only because of their low Ube. Thus, I need a typical for Ge Ube of, say, 0.2 to 0.4V.

1. Germanium.
2. Typical, low, Germanium Ube.
3. Audio ( low noise ).
4. >=30V. 40V preferable.
5. Any power >= 0.6W. Lower may be OK too.
6. Current of >= 100mA. Lower OK.

If you are using a 30+ volt supply, what’s a couple tenths of a volt of signal swing? I could see the need for Ge’s if running off a 3 to 9 volt battery and you conserve every tenth. If you want even less vbe drop from input to output you could use a diamond buffer. 4 Si transistors and no effective vbe drop because the two stages cancel.

If you want to play with germanium for germanium’s sake you take what transistors you can reasonably obtain in your neck of the woods, and design around what they can do.

 

[StevenStanleyBayes]

Forget it.
What you're asking is not good design practice.
If it were, germaniums would have been used in commerial products for that reason.

Yes, Germanium transistors must be used. They have much lower Ube and may be faster. Their capacitances may be slightly higher, but, this depends on the transistor. Modern ones may be faster.

Also, their Ueb ( reversed Ube ) is much higher than in Silicon. Double or more.

They only reason they do not make them in large numbers is Germanium is more expensive than Silicon.

 

[StevenStanleyBayes]

I have MP37B and MP42A germanium, both HFe 40 to 80.

I even have some MP111a (which are Si, and even lower gain like Hfe 20...)

Thanks. This is one of a few helpful posts. MP37B is 30V. However, MP42A is 15V.

Do you know any complementary to MP37B? Do you have any other numbers?

 

[StevenStanleyBayes]

If you are using a 30+ volt supply, what’s a couple tenths of a volt of signal swing? I could see the need for Ge’s if running off a 3 to 9 volt battery and you conserve every tenth. If you want even less vbe drop from input to output you could use a diamond buffer. 4 Si transistors and no effective vbe drop because the two stages cancel.

If you want to play with germanium for germanium’s sake you take what transistors you can reasonably obtain in your neck of the woods, and design around what they can do.

I am interested in low Ube, as low as possible. Some Ge transistors have similar Ube as Silicon. I am not interested in these.

Typical Ube of 0.2V ( even at low current ) is OK. This changes with temperature. For Silicon transistors, the change is - 2mV/C. All, please, inform what this is for Germanium. Looks similar or the same.

 

[StevenStanleyBayes]

Please, be helpful. Post names and numbers of good or popular Germanium transistors.

I am not interested in opinions as they do not help. I am interested in numbers.

 

[StevenStanleyBayes]

I have, just, cross referenced GT402V or G. They are equivalent to AC128 ( 32V, though ).

GT404I is equivalent to 2SD72 ( 25V ), 2SD72K ( 25V ), 2SD128 ( 30V, 32V ) and 2SD128A ( 32V ).

So far, the selections are :

1. GT402G and GT404I
2. AC128 and 2SD128A

( and some NTE's for those who want to use them )

Please, provide more numbers.

 

[StevenStanleyBayes]

In regards to NTE102A (PNP) & NTE103A (NPN). These are not preferable, because, these are very slow. The cutoff frequency of a common emitter amplifier of Beta ( f ) is only 10KHz. Therefore, assuming a slope of 20dB/decade, Beta decreases by, approximately, 6dB at 20KHz. This means, Beta at 20KHz is nearly half of the nominal. In case, however, the slope is steeper than 20dB/decade, which, is possible in case of a multiple order filter equivalent effect, then, the situation becomes tragic.

 

[StevenStanleyBayes]

I have, just, cross referenced GT402V or G. They are equivalent to AC128 ( 32V, though ).

GT404I is equivalent to 2SD72 ( 25V ), 2SD72K ( 25V ), 2SD128 ( 30V, 32V ) and 2SD128A ( 32V ).

So far, the selections are :

1. GT402G and GT404I
2. AC128 and 2SD128A

( and some NTE's for those who want to use them )

Please, provide more numbers.

 

[mondogenerator]

Thanks. This is one of a few helpful posts. MP37B is 30V. However, MP42A is 15V.

Do you know any complementary to MP37B? Do you have any other numbers?

MP37A is higher gain group if I recall.

As for complimentary PNP types, I do not know if they exist in this series of "modern" Ge transistors.

There see very few NPN types, and MP37 is the only one I know of.

 

[StevenStanleyBayes]

I have MP37B and MP42A germanium, both HFe 40 to 80.

I even have some MP111a (which are Si, and even lower gain like Hfe 20...)

Thank you very much. Your post is one of a few helpful.

Yes, MP37B is 30V and, most importantly, is NPN. Usually, PNP Germanium transistors are easier to find.

I will add this to the summary of found transistors. Thanks. In case you or someone else knows of a >= 30V PNP, please, inform.

 

[StevenStanleyBayes]

GT402V or G is equivalent to AC128 ( 32V ).

GT404I is equivalent to 2SD72 ( 25V ), 2SD72K ( 25V ), 2SD128 ( 30V, 32V ) and 2SD128A ( 32V ).

So far, the selections are :

1. GT402G and GT404I
2. AC128 and 2SD128A
3. MP37B ( 30V, NPN )

( and some NTE's for those who want to use them )

Please, provide more numbers.

 

[StevenStanleyBayes]

Someone very helpful has posted these transistors in another thread I have started :

AD161 \ 162, AC187 \ AC188K, AC127 \ 152, AC128 \ AC152, AC128K \ AC176K, 2N656A \ 2N1057,(ГТ402\404 А-И, ГТ703\705 А-Д)

AC127 \ 132, AC141 \ 142, AL103, OC31 \ OC139 or ASY28

We have already discussed some of these, mainly, GT402G or V and GT404I ( ГТ402\404 А - И ). In regards to the others :

AD161, 162 : 20V
AC187, AC188K : 15V
AC127 : 12V
AC152 : 24V
AC128 : 16V
AC176K : 18V
AC176 : 20V
AC128K : 16V
2N656A : Silicon

2N1057 : 45V ! ( also, 2N650, 1, 2 : all 30V ) all of these : PNP

AC127 : 10V
2N132 : 24V
AC141 : 18V

ACY11, 14, 18, 19, 23, 32 : 30V, all PNP !
ACY17 : 32V, PNP !

AC142 : 20V

AL103 : 40V, 30W, 6A, PNP !

OC31 : ?
OC139 : ?

ASY28 : 15V

 

[StevenStanleyBayes]

FOUND SO FAR :

GT402V or G is equivalent to AC128 ( 32V ).

GT404I is equivalent to 2SD72 ( 25V ), 2SD72K ( 25V ), 2SD128 ( 30V, 32V ) and 2SD128A ( 32V ).

So far, the selections are :

1. GT402G and GT404I
2. AC128 and 2SD128A
3. MP37B ( 30V, NPN )
4. 2N1057 : 45V ! ( also, 2N650, 1, 2 : all 30V ) all of these : PNP
5. ACY11, 14, 18, 19, 23, 32 : 30V ; ACY17 : 32V. All these PNP
6. AL103 : 40V, 30W, 6A, PNP : Attention : High Power!

Better choices : see the attached datasheet.

( and some NTE's for those who want to use them )

The datasheets do not provide all parameters and do not provide popularity. Please, provide more numbers.

 

[mondogenerator]

The vast majority of Ge will be PNP.
No?

 

[StevenStanleyBayes]

Yes. This is a problem. Especially, with high frequency and speed transistors. Not very many complementary and not very many NPN.

 

[StevenStanleyBayes]

GT402 and GT404

Attached is the English translation of GT402 and GT404. These transistors are amazing for audio.

They have a high leakage current of <= 25uA. This must be considered for high Beta Sziklai configuration.

Also, with ft = 1MHz, these are very slow for some applications. I forgot to mention, I need extremely high speed transistors.

Anyway, in most cases, Germanium transistors are much better for audio than Silicon as mentioned in Wikipedia and many threads and forums.

Thus, when huge speed is not of concern, these transistors are ideal and, still, available.

Please, inform in case of any mistakes. Scroll down for the original, USSR, documents.

Here are the documents :

Update your browser to use Google Drive, Docs, Sheets, Sites, Slides, and Forms - Google Drive Help

 

[Mark Tillotson]

Anyway, in most cases, Germanium transistors are much better for audio than Silicon as mentioned in Wikipedia and many threads and forums.

Oh dear oh dear, yet another myth based on (as far as I can tell) pure nostalgia. Read the datasheets, do the math and the measurements. Reality doesn't agree with this conceit (yes you can pick a poor Si device and compare with a good Ge device if you want, but cherry-picking isn't a valid argument)

There are some truly excellent Si devices these days that Ge devices haven't a hope in hell of outperforming on every characteristic you can think of. Bandwidth, noise (voltage, current, flicker and popcorn), manufacturing spread, power handling, linearity, leakage, gain, SOA, switch-off speed - all of these are well represented in Si devices because the technology has been in active development for 50 more years than for Ge - so whether or not Ge is better/worse/comparable to Si as a semiconductor the old devices are nothing like state of the art. Where Ge has an advantage is in very low voltage operation where the low Vbe is useful. So if you want to power your device at 1.5V Ge BJTs have a big advantage over Si BJTs. (But perhaps not over Si MOSFET technology!) This is one of the _very few_ cases where Ge is "better for audio" than Si. I believe one of the niche uses for Ge devices was for hearing aids as a result of this.

Don't you think if Ge transistors were "much better for audio" in "most cases" people would be selling Ge amplifiers in the millions? But they don't, Si overtook Ge and never looked back - you simply have deal with it, nostalgia isn't valid audio engineering


BTW I find no claim on wikipedia about Ge transistors being better - where did you read this?

 

[NareshBrd]

He seems a dreamer from his posts, powerful imagination.
But he needs to learn from the experience of others.
Realize why Silicon replaced Germanium, then walk the talk.

 

[wg_ski]

Nostalgia for nostalgia’s sake *is* a valid pursuit in DIY audio - with the understanding that that’s what it is. Not necessarily being driven by ultimate performance. People play around with germanium and tubes. Often fooling themselves into thinking they are the best tool for the job - if the job is to make the best amplifier possible. Sometimes the goal is to get a specific sound, or to do as good of a job as I can given X technology or with these specific Y and Z parts. But make no mistake you would not start out with X,Y, or Z if you start with a set of engineering specs on paper and then take the best route to get there.

 

[MarcelvdG]

Coming back to the original question, h_FE is 63 ... 295 at 50 mA, h_fe at 10 mA is presumably about the same. The cut-off frequency of the common-emitter current gain f_αe is at least 10 kHz. 10 kHz cut-off with an h_fe between 63 and 295 corresponds to an f_T somewhere between 630 kHz and 2.95 MHz. I don't know why someone claimed it to be 200 kHz. Miller effect has nothing to do with it, as the small-signal current gain is measured with the output AC shorted.

StevenStanleyBayes, why does the base-emitter voltage matter so much? Can't you use a PNP emitter follower followed by an NPN emitter follower, so the base-emitter voltages more or less cancel?