As for frequency parameters of some Ge power transistors, such ones as 1T901A and 1T910A have Ft >= 30MHz.
thanks, Vladimir.
Parallelling transistors with AC coupling gets around the emitter resistor problem, and it is (was?) possible to get high frequency germaniums. Some AU types for TV in the main were common.
If parallelling, then Si devices can get to theoretical values (below 25 milliohms) largely by reducing the effective base resistance.
What I remember of Ge devices (long time since I looked at any) is that they still had high current effects and linear (resistive) base-emitter characteristics at higher currents. Perhaps that explains low distortion rather than low Vbe, which at lower currents is the classic exponential.
Parallelling transistors with AC coupling gets around the emitter resistor problem, and it is (was?) possible to get high frequency germaniums. Some AU types for TV in the main were common.
If parallelling, then Si devices can get to theoretical values (below 25 milliohms) largely by reducing the effective base resistance.
What I remember of Ge devices (long time since I looked at any) is that they still had high current effects and linear (resistive) base-emitter characteristics at higher currents. Perhaps that explains low distortion rather than low Vbe, which at lower currents is the classic exponential.
Traditional transistors paralleling implies quite essential emitter resistors. The way of paralleling that I have mentioned above, does not imply using emitter resistors for correct currents sharing
Yes, that's what I said!
"Parallelling transistors with AC coupling gets around the emitter resistor problem"
"Parallelling transistors with AC coupling gets around the emitter resistor problem"
Too many answers in the thread. Cannot read them all now. THE THREAD IS FOR FINDING GERMANIUM TRANSISTORS. As far as Ge vs. Si is concerned, for any Si transistor, there is a much better Ge transistor for audio in the same class. Therefore, Ge transistors are better for audio. All of their audio parameters are better. Most importantly, their Ube is more than half of this of Si. The only parameters, where, Si has some, unimportant, advantage is maximum temperature ( and, therefore, maximum power dissipation ). Everything else is better with Ge. The ONLY reason for not manufacturing them is PRICE. Read Wikipedia and other threads on the topic. And, yes, their datasheet parameters are better.
My oh my, WHY are we WASTING time on you and your whims?
Help yourself, the NET is wide and deep.
Do your own homework.
Even so, there are whole Op Amps, infinitely superior in all parameters as full gain blocks. guaranteed working from as low as 2.7V, so ....
So you ASK but can´t be bothered to read ANSWERS to YOUR questions?
Help yourself, the NET is wide and deep.
Do your own homework.
Nonsense.
Nonsense.
That is the ONLY "advantage", usually irrelevant and only matters some in MARGINAL circuits (such as those powered from 1.5V to 6V
Even so, there are whole Op Amps, infinitely superior in all parameters as full gain blocks. guaranteed working from as low as 2.7V, so ....
AND Gain AND infinitely less leaks AND higher Voltage AND higher current AND higher dissipation AND less noise AND higher bandwidth , etc.
Nice slogan .......... for a 50´s ad selling soap.
Maybe there´s a couple other reasons too, just read above.
PLEASE post links to those Wikipedia articles,would LOVE to read them.
Again, SHOW those inspiring datasheets.
Why not? It's only 10kHz at full power, next, you only need 3kHz for audio. You are just too spoiled! 🙂
Most answers, such as this one, are useless and untrue. I have, specifically, asked for quality Ge transistors and, only a few posts, have provided information of such. This is why I am not happy to read garbage.
In regards to NET, no much information there either.
In regards to Ge transistors, the NET is clear they are of higher audio quality.
They are faster, too, yet, people compare Ge transistors made in 1950's with Si transistors made in 2020's. Even then, the Ge is superior for sound in every shape and form : faster, low Ube, quieter, faster moving carriers, etcetera. Lower temperature, but, this may not be the most important point always.
Articles :
1.
https://en.wikipedia.org/wiki/History_of_the_transistor
"
Up until the late 1950s, however, germanium remained the dominant semiconductor material for transistors and other semiconductor devices. Germanium was initially considered the more effective semiconductor material, as it was able to demonstrate better performance due to higher carrier mobility.
"
Si is much more inexpensive than Ge, yet, lousier in quality. Even MOSFET lovers talk of Si Ge or pure Ge as the technology now allows more inexpensive production. https://en.wikipedia.org/wiki/Germanium#Electronics
2.
https://spectrum.ieee.org/germanium-can-take-transistors-where-silicon-cant
"
But now, remarkably, the material is poised for a comeback. The world’s leading-edge chipmakers are contemplating a change to the component at the very heart of the transistor—the current-carrying channel. The idea is to replace the silicon there with a material that can move current at greater rates.
"
3.
https://www.easytechjunkie.com/what-is-a-germanium-transistor.htm
"
germanium alloy is commonly used to increase transmission speed of electrical signals
"
"
Their throughput speed or lower cut off voltage is superior to silicon
"
"
Germanium alloys offer enhanced transmission rates in high-speed circuitry over silicon
"
4. Because I am sure you would not believe these, then, please, read the study of Germanium transistor published in this forum. Not everything is true, but, some is :
https://www.diyaudio.com/forums/pass-labs/348948-germanium-investigations.html
"
The harder you drive the Germanium devices, the more the linear region expands.
"
and many others.
5. Read the translated datasheets. Only 1MHz, but, there are other, much faster.
Last edited:
It's not that you can't have high frequency germanium transistors, af239 works up to 900 Mhz, but its thermal regime is a bit complicated to follow as its thermal runaway is damn fast .Its thermal coupling to dissipate the heat wasn't great either so maybe new technologies would alow it to be easier to track , tame, dissipate.There's a debate over its maximum speed in audio transistors where you need higher speeds for feedback based circuits to use lower compensation, but, here being utterly speculative cause I don't know for sure the difference, it seems to me that the gm doubling effect that asks to be corrected by the feedback in silicon class b or ab amplifiers in a timely manner isn't that obvious in germanium transistors even though with germanium you have significant differences between pnp and npn transistors .First of all the emitter resistors in germanium transistors is at least three times more effective for the same value than it is with silicon in reducing distortions locally which lowers the need for a faster global feedback.Hfe factor in germanium medium power transistors(1...10w) looks fantastically high at 3...500 compared with similar dissipating power silicon transistors yet they have leakage which is 3...6 times higher in npn transistors than in pnp transistors and its dissipation power is 3 times lower than in silicon transistors so it's a bit difficult to keep them cool unless perfect temp coeficient matching is done and that looks way more difficult with germanium when run at high powers than with silicon .Seeing on youtube a 4 transistor amplifier with current feedback that was showed to be completely linear up to 100khz while using 1 mhz transistors was a revelation to me honestly cause you can hardly need something more from an audio amplifier.I don't know for sure the intricacies of germanium operation, but it looks like you don't need germanium transistors to be as fast as silicon ones for the same duties.
You might wanna hang around this guys here for more discoveries on germanium transistors : https://www.diyaudio.com/community/threads/old-soul.370744/
a) nonsense, circuit is what matters the most, etc.
b) nonsense, of coutse Ge is way slower
c) you only need to compare Ge of the 50's with Motorola of the 60's.
d) nonsense
e) nonsense, not faster
f) nonsense, low Vbe has nothing to do with anything, it is a major disadvantage when it changes so much with temperature, to the point of becoming full conductor at high temperatuires, or full insulator at low temperatures.
g) nonsense, it always depends on transistor structure, like you said, why compare transistor with a base of a diameter of 0.5mm with one with 5mm diameter? It's only 100x difference in area...
h) what were you smoking?
Show us your circuit. It's that simple.
Recently i measured about 100 ac 180/181/187/188 and in every single one the hfe increased sharply with temperature with the base current relatively stable.Wouldn't it be the case that the base width is virtually the same and its current very stable up to a point? If it's going to melt than maybe this would happen over half the maximum collector current where I usually stopped taking my measurements.I usually stopped taking measurements when the case became mildly hot to the touch with the collector current being about 1/3rd of the max current in the datasheet as i used no heatsink .For ac187, the hfe range would change from 250 to 600(over the max hfe in the datasheet which was 500) in about 30...40 seconds with temperature rise, but the base current would have very small changes in that empirical range from cold to hot to the touch.Am i interpreting my results in a wrong way? If not than maybe just keeping the transistor below 35...40 degrees celsius would be perfectly ok.
Power amplifiers that had germanium power transistors had sophisticated thermistor-based work point compensation, and even then they burned out frequently.
There is so much work required around germaniums! No wonder colleague Peas called them geraniums! They needed to be potted in some moisture-absorbing soil and taken care of. What happened is that the Vbe becomes zero at some mildly hot temperatures (as far as electronics goes) and below that, the Vbe can be, say, 100mV instead of, say, 500mV. Your AC signal would be suddenly much larger, compared to the cold working point, and the self-leakage between collector and base would be also increased. But of course the hfe was greater as you were coming close to the semiconductor becoming a full metal state. With the Vbe going closer to zero, any current injected into the base was edging on it becoming a metal, or, rather, conductor between collector and emittor, you can see this in the planar silicon transistors which had thermal breakdown, the silicon between some part of emitter and collector completely remelts and flows out, that is where the base stopped existing at about 400 degrees Celsius. Once the base stops existing, all current can flow trough.
This wild change of hfe was usually managed trough emitter degeneration, that is, resistor added between emittor and common voltage point in case of common emittor circuit. Once silicon transistors became common, this practice was quickly abandoned as unnecessary, as it also increases the noise of the circuit.
The ac187 datasheet shows that both the Vbe and hfe go up with the collector current not down so the measuring technique for germanium hfe implies a higher Vbe for higher collector currents which require a lower value resistor to bias the BE junction....I didn't see the emitter resistors going anywhere once silicon entered the market.It's everywhere.
Last edited:
"current" does not equal "temperature"!!
One of the old, old silicon 2N2222A had like hfe=200 at room temp, and hfe=320 at 150 degrees Celsius. Not a big difference for a huge increase in temperature. Germanium has the same difference in much smaller change of temperature.
With germanium transistors, the emitter degeneration, with an added capacitor over the emitter resistors were like in 100% of cases. how many do you see in common emitter connection today? I'd like to see the examples, please.
Unless there's a translation error somewhere...what are you even talking about? The only circuits where i don't see emitter resistors in silicon trz are in switching circuits and motor drivers maybe...and i see a lot of germanium circuits with no emiter resistor at all in a lot of old books with old commercial equipment that i have around me though...
Probably.
They were also painfully under-heatsinked.
Were easy to get into thermal avalanche.
Didn´t have short protection.
Didn´t use Zobels.
Then Engineers learned better.
ac187/188 and ad161/162 are nice transistors for audio, and still to be found. 30 volts or so max, in those days it was very hard to make high voltage + high power devices.
Maybe they don't need Zobels...being capacitevely coupled you wouldn't need much protection...heatsinks were expensive, active cooling even more, short protection to a short circuit ending in a capacitor was useless, transistors were too expensive to be used in great numbers in one circuit...engineers knew less, but were smarter cause they would read more and had no computers to make symulations...