Germanium in my attic

[RuptorGB]

I have Germanium transistors knocking about as in the picture and some AC128 that I can't find at present. In my hand there 10 red dot, just the 2 white dot and a roll of OA47 diodes maybe 500 or more. Are these transistors NKT275 as they were bought here in GB back in the early 1970s? This site goes on & on about Fuzz Face diodes and about 10 "page downs" they mention red and white dots on canned transistors.
https://www.analogman.com/fuzzface.htm
I might have more somewhere as I have a lot of older components like Fetlingtons and M68010 cpus in many boxes. I must find them AC128s but although this site has done a big analysis on them in the Fuzz Face
https://www.electrosmash.com/fuzz-face
others, like this one that has seen just about every Fuzz Face
https://fuzzboxes.org/fuzzface#part-1-history
say they never fitted AC128s in any version of Fuzz Face.
Oh! found OC45 or OCxx types are glass but what is red or white dot and what OC number are they?

[modman]

If you want fuzz, use a good and cheap silicon transistor...
If you want a lot of fuss, keep messing with germanium.

Germanium is a sport to itself:
Don't build a germanium fuzz for you first build, we would like you to have a positive experience and not give up the hobby
Germanium transistor type number don't mean much -- all that matter is gain and leakage: http://www.geofex.com/article_folders/ffselect.htm
All transistors used in original 60's fuzzes were put in to get rid of that old germanium crap - by 1965-6 we have reliable silicon transistors.

You probably received a number of pm's trying to buy those off you?
Dots on germanium transitors: viewtopic.php?t=24469
If you still want to give it a go: start with the Goefex page for measuring gain and leakage.

Analogman kinda hates my guts and called me an "internet terrorist" for starting this free speech stompbox forum. Maybe not everything he says is true...

[RuptorGB]

Analogman kinda hates my guts and called me an "internet terrorist" for starting this free speech stompbox forum. Maybe not everything he says is true...

Unfortunately people that chase money and create multiple theories to give more opportunities to sell more don't like others telling the truth scuppering their profit avenues. From what I have assimilated silicon transistors can be arranged to give the required gain for each stage but it is the characteristic of how the transistor clips that counts. It should be possible to add networks to make the silicon transistors work like Germanium just like making an Opamp simulate an inductor. I guess nobody has bothered because it can be done in digital processors now allowing all sorts of variations in parameters.

[modman]

It should be possible to add networks to make the silicon transistors work like Germanium

Piggybacking SI to simulate GE: viewtopic.php?t=2674

just like making an Opamp simulate an inductor.

I guess nobody has bothered because it can be done in digital processors now allowing all sorts of variations in parameters.

Above thread about piggybacking SI to simulate the lower gain of GE is from 2008...

But really, if you are after a specific sound: use band EQ before (and after) you fuzzbox

[drewl]

There's just something about the warm blooming fuzz of good Ge transistors, IMO.
I'll admit, I've never tried that stacking technique with Si ones, so I don't know how close they sound, but that sounds like a fun experiment.

Years ago I found a few dozen PNP and NPN Ge's in an old piece of test equipment and built several different fuzz boxes, the Tonebender MKII Professional being my fave, it just nails that Jimmy Page fuzz sound.

[jrod]

You might consider building Joe Gagan's Easy Face that uses a germanium/silicon combo. Search the internet and you should find the schematic and probably many forum threads.

[RuptorGB]

Years ago I found a few dozen PNP and NPN Ge's in an old piece of test equipment and built several different fuzz boxes, the Tonebender MKII Professional being my fave, it just nails that Jimmy Page fuzz sound.

Out of curiosity did you test and pick the Germanium transistor gain for each circuit position or just fit at random?

[drewl]

Most of the fuzzes I built have bias trim pots for tweaking
And most have the transistors socketed for swapping around to see what sounds best.

[RuptorGB]

It bugged me that in my batch of Germanium transistors there are red and white dots. Although I don't know the source of the transistors I bought them as surplus so it is most likely they are from the same manufacturer. Therefore the dots must mean something as well as marking the collector. Measuring the diode drops it is clear there is a difference and looking on the Web red are better than white dots. The red dots seem to measure 0.19V to 0.2V and white dots are over 0.22V. If the diode drop can identify the better transistor then can it identify the gain? I notice that on each transistor the BE & BC diode drops can be different and on the white dot ones the gaps are larger. Logically the closer the diode drops are the better the transistor is and the perfect transistor is probably when drops are equal meaning the manufacturing process was perfect.

[mauman]

When the transistors were made, they tested each one for beta (hFE), then tossed them into a bin based on the beta range. The transistors in each bin got a different color dot, so different color dots meant different beta ranges. Another way manufacturers indicated the bins was with letter suffixes on the part numbers (BC-549A, BC-549B, etc.)

Here's an example from the datasheet for a Toshiba 2SC1815, where orange, yellow, green and blue signify the bins. I don't know what bins your particular white and red colors mean, but if you test them, the beta will likely be grouped by color.

[tonyharker]

To my recollection - I'm 79, in the early days in the absence of a type number, a white dot donated an RF transistor - suitable for medium wave radios - and a red dot meant it was suitable for audio frequencies. These were usually rejects from the manufacturers and sold for hobbyists. In all cases the coloured dot donated the collector lead, followed by base then emitter.

[mozz]

That is with UK/EURO transistors right?

[tonyharker]

Yes thats right. Probably they started to appear about 1955 or 1957. There was also a blue spot transistor with higher frequency range up to about 1.5MHz. Red spots were approximately equivalent to Mullard OC71.

[RuptorGB]

The Transistors were itching to be used in a FuzzFace circuit so I had a spare 15 minutes that turned in to hours and stuck them in a plug board to have a fiddle. What a nightmare is my impression of the circuit design that is a classic case of someone making one, deciding it works then putting it into production. Not having any knowledge of a FuzzFace other than reading other opinions on the Web I had the stupid idea that the circuit would pass an undistorted signal when set to minimum fuzz and low volume but in hindsight why would there be a straight through switch. The two pictures show my plug board is giving rounding on low and high levels so is working as described by others using 2 red dot transistors. The white dot ones were useless for this circuit.
My impression of the circuit is it doesn't need the output potentiometer because you have no control of the load so don't add one as the output level can be fixed. I can't see the point in the fuzz level control either because you can't operate it with your foot or alter it easily when playing. My solution is you set the level of fuzz by regulating the input signal level so it is better to use a volume pedal before the FuzzFace. I can't see the point in cutting frequencies on the input and output so I am putting big capacitors of 100uF on the input and output. When I try it with a guitar things might change but I think not cutting the sub 100 Hz frequencies will add to the warm sound.

[mozz]

No need to use a vol pedal, use the guitar volume as a fuzz face was meant to be used. Also, you should build it stock fist before changing those capacitors that drastic. Also, why were the white dots not useable? You post your transistor voltages in circuit you will get lots of help and probably make about any tranny work.

[RuptorGB]

Maybe I didn't explain things well enough or left out things I thought people would assume.

No need to use a vol pedal, use the guitar volume as a fuzz face was meant to be used.

My thinking is a foot volume is an improvement as it makes it easier for us lesser players to do two things at once rather than trying to turn a guitar volume while fingering or plucking. Another advantage is the guitar can be on full volume increasing the signal to noise level at the pedal but a foot volume was only a suggestion.

Also, you should build it stock fist before changing those capacitors that drastic.

As an engineer I am a logical thinker so why assume I would change all the components before setting up a plug board test bed especially when I said I didn't know anything about the FuzzFace?

Also, why were the white dots not useable? You post your transistor voltages in circuit you will get lots of help and probably make about any tranny work.

To get the bias voltages correct as shown on Web sites I plugged different transistors in and based on the junction diode drops honed in on the right match. The worst transistors for the original circuit and biggest diode drops were the white dot transistors.
I can always change the bigger caps to restore the original circuit if the unit sounds crap as I said I haven't heard it yet and they are only affecting the bass. I want to try it on a bass guitar anyway out of curiousity.
Hope these further details explain what I did and my reasoning.

[drewl]

I built this Hoffner fuzz years ago, even with 10uf caps instead of the 50uf and 25uf it was too bassy

[modman]

How to score germanium for cheap:
1. Search local, non specialist, second hand sites for pre 1965 electronics, preferably without brand or model number (they are cheap)
2. Use the picture on Google Lens (image search) and find the brand and model number
3. Put the brand and model into http://radiomuseum.org - all active devices are listed (tubes, transistors)
4. Buy it as cheap as you can - non-working is probably not a problem (and cheaper) and cross fingers
5. Harvest some prime germanium transistors
6. Measure them and get ready to be dissapointed

[RuptorGB]

6. Measure them and get ready to be dissapointed

Yup I am. My impression is the Fuzz Face circuit is flaky. It does seem better as a bass fuzz maybe because the range of frequencies is low. On a bass it gives a warm fuzz and if all the string are played it gives a rolling thunder sound as the frequencies add and subtract. Thin string guitars give spiky high frequencies that all affect the circuit and to me they seem to cut the actual guitar notes. Anyway I have a dislike of the circuit so I am probably biased and shall use as a bass fuzz. The main point people seem to treasure is the curved cut off of the Germanium transistor that is the function of the transistor junctions. The same characteristic is available in Germanium diodes so slapping them in a standard diode fuzz circuit seems an obvious way to go. I tried a diode fuzz with Germanium ones on LTSpice and the output is exactly the same curve. If the asymmetric aspect is required it was easy to implement on the simulation but the proof of the pudding is when you hear the result so I need to build the circuit to compare asymmetric to even waves. Curiously I was playing with a JFET simulation and noticed it gave the same curved roll off as Germanium junctions as well as the ability to be asymmetric if biased right but it would take a lot of time and maybe additional circuitry to make a useable unit. If the simple Germanium diode fuzz doesn't cut the mustard I will take a look at a JFET version.

[jalmonsalmon]

Jimi hated them... yup! I mean you take a Strat and some 4 hole plexis with a fuzz face... sounds like ars!

(facepalm)