actually our shop repaired a "modified" fuzz pedal last week. the complaint was that the fuzz effect sounded raspy and cut off rather than a smooth transition from fuzz to clean as the note sustained. the problem? one of the silicon diodes was replaced with a 1N270 germanium (apparently one of the "mods" found on the net). the imbalance in conduction voltages caused a lot of squirrely things to happen in the pedal. also take notice that germanium conduction curves are much more nonlinear than silicon, and this may have a lot to do with unpredictable types of distortion (the conduction voltage drop changes with the current a lot more than in silicon). i call germanium a "sloppy" diode (also because of it's high reverse leakage current which increases with age)
This reminds me of a recent discussion on a native forum I took part in.
In principle, there is no reason why germanium would be any worse, and in fact there may be good reason why it could be better, IF (and this is a very BIG if) modern manufacturing techniques were used. Germanium has better conductivity and even in the old days when semiconductor 'metallurgy' was in it's infancy, formed pretty high gain and high current devices.
Even then, there remains one big problem, and that is the very considerable difficulty of passivating germanium dies. In silicon, we just make Sio2, which is stable and a very good passivator. Germanium does not form any stable surface oxides. In general, this was the reason for introducing metal and glass cases - the substrate was soldered (as germanium solders quite well), and being 'on the other side' of the surface with the actual junctions, gave little chance of diffusion (in effect junction contamination) during it's expected lifetime. but on the top, the die was covered with aneutral agent (silicone grease...) and hermetically sealed. No such thing as plastic case germanium!
Today we have SiGe processes, but these are mostly FETs.
The crux of the matter is, most if not all of the assertions above do not apply. Germanium technology died a quiet death in the early 70s and even though some of the parts are still available (and made!) today, and germanium prices are at an all-time low, the fact remains it's early semiconductor technology, a sort of NOS. Not fair comparing it with, say, NOS tubes, as these were already mature enough even for WW2 era NOS - it's an equivalent of what germanium semis would be if they continued to be developed into the late 80s and early 90s.
Could an amp be made with germanium tansistors? Certainly - keeping in mind the limited available selection. Could it sound good? Again, i have no doubts - if you ahve enough money, there are plenty of techniques which largely circumvent the part's 'failings' compared to silicon. The question is, though, would it be better, and at what price, if any?
In the meantime, here's a link to browse, to see what was available back then - the more intrepid can go a few levels up and see the rest of the characteristics for many germanium types:
http://www.ginko.de/user/franz.hamberger/infobase/charts/semi/geltmax.html
In principle, there is no reason why germanium would be any worse, and in fact there may be good reason why it could be better, IF (and this is a very BIG if) modern manufacturing techniques were used. Germanium has better conductivity and even in the old days when semiconductor 'metallurgy' was in it's infancy, formed pretty high gain and high current devices.
Even then, there remains one big problem, and that is the very considerable difficulty of passivating germanium dies. In silicon, we just make Sio2, which is stable and a very good passivator. Germanium does not form any stable surface oxides. In general, this was the reason for introducing metal and glass cases - the substrate was soldered (as germanium solders quite well), and being 'on the other side' of the surface with the actual junctions, gave little chance of diffusion (in effect junction contamination) during it's expected lifetime. but on the top, the die was covered with aneutral agent (silicone grease...) and hermetically sealed. No such thing as plastic case germanium!
Today we have SiGe processes, but these are mostly FETs.
The crux of the matter is, most if not all of the assertions above do not apply. Germanium technology died a quiet death in the early 70s and even though some of the parts are still available (and made!) today, and germanium prices are at an all-time low, the fact remains it's early semiconductor technology, a sort of NOS. Not fair comparing it with, say, NOS tubes, as these were already mature enough even for WW2 era NOS - it's an equivalent of what germanium semis would be if they continued to be developed into the late 80s and early 90s.
Could an amp be made with germanium tansistors? Certainly - keeping in mind the limited available selection. Could it sound good? Again, i have no doubts - if you ahve enough money, there are plenty of techniques which largely circumvent the part's 'failings' compared to silicon. The question is, though, would it be better, and at what price, if any?
In the meantime, here's a link to browse, to see what was available back then - the more intrepid can go a few levels up and see the rest of the characteristics for many germanium types:
http://www.ginko.de/user/franz.hamberger/infobase/charts/semi/geltmax.html
For the reasons you've mentioned (oxides!!) pure germanium is not very likely to come back in "new shell" and modern technology.
It's unlike modern compound semis like SiGe or GaAs and others...
But keep in mind that many posters in this thread seem to believe, that technology is like wine or violin, the older the better.
regards
Adam
It's unlike modern compound semis like SiGe or GaAs and others...
But keep in mind that many posters in this thread seem to believe, that technology is like wine or violin, the older the better.
regards
Adam
Interesting.
I built and had (until recently) a germanium type amplifier I built when I attended Ryerson in Toronto. The complimentary outputs were ACxxx I think. I may still have some of those pairs. It sounded okay, better than some other cheapie amps.
I think a worthwhile project would be a headphone amplifier. The reason it that it is a real amplifier circuit, but the voltages and currents are low enough so that other parts may be used.
Carlos, this sounds like something you can play with if you are so inclined. Base on "Blameless" or SymAsym with low voltages. Rails at 9 VDC ~ 15 VDC might work out well.
-Chris
I built and had (until recently) a germanium type amplifier I built when I attended Ryerson in Toronto. The complimentary outputs were ACxxx I think. I may still have some of those pairs. It sounded okay, better than some other cheapie amps.
I think a worthwhile project would be a headphone amplifier. The reason it that it is a real amplifier circuit, but the voltages and currents are low enough so that other parts may be used.
Carlos, this sounds like something you can play with if you are so inclined. Base on "Blameless" or SymAsym with low voltages. Rails at 9 VDC ~ 15 VDC might work out well.
-Chris
Friends
Anyone get good results using a Germanium transistor in a output stage ?
Aldovan
I used them in various output stages. The Hi-Fi quality DIN45500 standard was 6W RMS with 1% THD and the output pair for this was AD161/AD162. This is the monster of Hi-Fi if used properly (Grundig RTV380 Stereo) but it is the end of their power (22.5VDC power supply and 6 Ohms load). The more normal conditions are on 8 Ohms but there is no Hi-Fi norm satisfied, 5W RMS only. The distortion is really low, the tone is warm and has the solid bottom due to good damping factor on 8 Ohms. The heat sink is huge and must be for TO-66 case. The must is to find the as exactly matched pair as possible and the Io stabilizaton circuit between bases of the output transistors must be the germanium transistor as AC188K. In original they used various NTC resistors with temperature characteristics matched to germanium. The quiet current is quite low (5-10mA) and the complete stage should run cool (not above 40 Celsius on the heat sink). This all story is valid for AC187/188 for the maximum voltage of 16VDC and the output power of about 3W with 4 Ohms load.
There are germanium transistors as AC151r (rauscharm - low noise) for microphone pre-amps used in professional devices as UHER 4400 Report Stereo reel tape recorder (these one has the AC187/188K output as well). These transistors are the most suitable for preamplifiers. When it comes to fuzz circuit imitation for the valve sound, there is a must to obtain total symmetrical clipping of the sine wave both for on and off condition, by the circuit design. To avoid this they used diode limiter instead (matched pair of 1N60 from the old radios FM detector). This starts on 150mVpp with similar effect and it is not important anymore which transistors or ICs are used in trhe preamp (it is a linear area for them).
The main difference of germanium and sillizium is the way of getting saturated and turned off (the thresholds are softer and wider with germanium so the square pulses produced from the sine wave are getting smoothed a bit).
Regards, Toza.
That link doesn't work any more. Is it available elsewhere??
P.S. I am quite impressed with the AD161/162 output pair, in terms of Ic/Ib linearity (hFE pretty darned constant with current compared with virtually any BDxxx alternatives)... but the cost of these rare animals today is shocking. I noticed Philips and Mullard kept using these Ge output devices in published designs in the late 1960s/early 70s when complementary silicon devices were available; the ease of driving the germanium versions (i.e. low total count of transistors for a given distortion/sensitivity requirement) was probably a factor. Nowadays I wonder if there is an argument in their favour knowing what we do about high-order distortion and listening tests?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.