AF127 -an interesting Ge PNP transistor

[brett]

Hi
the other day I noticed that a local electronics store had a bunch (50?) of these.
Some figures (and my comments) for them are:
AF127  GE-P 32V 10mA 75MHz   (10mA is low current handling, but that's no sweat.  75MHz is VERY high for early Ge transistors)
AF127 pnp TO72 32v 10mA 75mW  75MHz hFE = 50min (hFE looks promising for Fuzzfaces and Tonebenders)

Prices are usually around US$2 or $3, although one well-known British supplier sells them for a tad over $5.
I got 6 for about US$6.

Testing my transistors, I got very low leakage and low hFE.  From lowest to highest hFE for the six transistors I bought, leakage (uA) and hFE were :
(15uA, 35), (20uA, 40), (25uA, 45), (25uA, 55), (35uA, 55), (50uA, 70)

These are really exceptional leakage figures.  But the hFEs are very low, too.  Compare with 3 AC128s I bought from the same store (they only had three):
(220uA, 125), (180uA, 145), (160uA, 155).

What I'm interesting in discussing is the use of AF127s.  I believe that they were used in Tonebenders (MkII).  But these hFE figures look really low (too low).

All of the figures above were measured using RG's tester, which generates 4uA of base current.  I wonder if these babies need a little more base current to get them up and going?  Or are they built for high frequency, and not for gain? (like the 2N2369A?)

thanks for any advice.

PS Tomorrow I'll re-check the price and how many are left at the store.

[Jaicen_solo]

Isn't the AF series similar to the AC127's RS sell?? I think they're higher voltage/lower current devices. I think I had one a while back from a surpluss supplier, I could never get it to sound good in a FF, though it did have some interesting tone in high gain variants.

[brett]

Hi.
AF127s are very different to AC127s, which are early PNP devices that are more like AC128s (but usually leakier and noisier).

AF127s seem to be like (or identical to) AF117s and AF115s: low power audio to radio frequency amplifiers.

Anybody out there used AF127s??

I'll generate an Ic vs hFE graph sometime in the next few days and post it.

[mac]

I have some AF106 pnp ge mesa 18V 10mA
hfe: 20 - 30;  leakage current: < 5uA !!!
gain is not enough but as darlington pair gain goes up to 400 - 600 which is good enough to replace 2N5088/89s in many circuits, for example a Ge BMP

Another transistors I love are Toshibas and Matsushitas. If you find some, buy them.
I have some Toshiba 2SA49 hfe: 60 - 160; leakage: < 60uA  ideal for Fuzz Face and Tone Benders

Lucky me, I have a bunch of different germs all low leakage low noise with the right gain for FF and TB:
2sb172 A , 2sb175 B, 2sb176 R matsushitas (letters indicate hfe classification)
2sa49, 2sa52 toshiba
matsushitas have a darker tone, toshibas are more crystal.

for a rangemaster I like 2sa102 C matsushita, hfe: 60 - 70; leakage: <150uA, or 2sd352 H npn see below

low gain, low leakage germs useful as a darlington:
2sa102 A & B, 2sa101 A & B matsushitas
2sa53 toshiba

NPN power transistors 32V 1A
2sd72k sanyo hfe 60 - 100; leakage: 300 - 500uA
2sd352 H matsushita hfe: 45 - 100 leakage: 100 - 300uA
I tried these in a 3-knob TB and they sound great. Also in a RM. Going to try a deacy amp some day.

generic ACxxx: 127,128,187,188, too much leakage but I guess with imagination I could use them as buffers or as clipping diodes

And they all cost me under a dollar each hehe.

mac.

[MartyMart]

Arn't AC127's the "NPN" version of the AC128 ....??

MM.

[brett]

Hi.
Yep, the AC127 is the complement to the AC128.
But it was more difficult to make NPN than PNP devices (I'm not sure if this was just for Ge, or for early Si devices as well).
Therefore AC127s are leaky and very often noisey compared to 128s.  Good ones CAN be found, but probably only 1 or 2 out of 10 from a typical random batch.
cheers

[Austin73]

Not really technical but you said you bought them for 3dollars have you tried thishttp://www.bardwells.co.uk/subcats.asp. They seem to be quite cheap.

Hope this helps someone

Aus

[Connoisseur of Distortion]

if i'm not mistaken, Ge naturally makes a better pnp, while Si naturally makes a better npn.

[brett]

Hi.
I paid AUD$23  (US$15) for 1 x AC188, 3 x AC128, and 6 x AF127
The AC128s were $5 each, and I think the AC188 was $4, so the AF127s were AUD$0.80 or US$0.60
In pence, that's about 30p.
cheers

[mac]

Are the AC128s from a well known manufacturer? All my generic ACXXXs are too leaky, +500uA.

mac

[brett]

Hi.
Sometimes old germaniums have been "trawled" for good ones.
A guy I built a pedal for wanted to do me a favour, and he bought me about 60 AC128s.  They were all bad.    I didn't have the heart to tell him.
My experience of untouched bulk lots is that about 50% of AC128s are good enough for pedal usage, and about 25% are very good.
cheers

[mac]

I bought 100 2SB176 R matsushitas in their original plastic bag, all written in japanese. Except a few, 3 or 4, they are great. hFE is between 80 and 200, 150 the average, leakage under 200uA. To my ears they sound better than AC128 since they leak less.
The few AFXXX I have are all siemens. All good, low hFE, low leakage.
IMHO the brand makes the difference.

mac

[ilcaccillo]

AF transistors are stated to be Germanium small-signal RF (Radio Frequency) transistor.

I don't see them used in DIY audio frequency related circuits, any reason for that?

Will the fact that they were made for RF have any limitation when used in AF (Audio Frequency) ?

[Rob Strand]

AF127s are very different to AC127s,

Yes, that's what I remember.

IIRC, AF127 were most used for RF in radios and TVs.

The RF devices tended to have lower gains.  (Which often follows from the slightly different structures used to reduce capacitance.)

[ilcaccillo]

AF127s are very different to AC127s,

Yes, that's what I remember.

IIRC, AF127 were most used for RF in radios and TVs.

The RF devices tended to have lower gains.  (Which often follows from the slightly different structures used to reduce capacitance.)

if the gain of an AF transistor falls wothin the ranges needed for a fuzz, would it perform well in this type of circuit?

besides the gain range that might be smaller is there any other limitation when using an AF transistor instead of an AC type?

[Rob Strand]

if the gain of an AF transistor falls wothin the ranges needed for a fuzz, would it perform well in this type of circuit?

besides the gain range that might be smaller is there any other limitation when using an AF transistor instead of an AC type?

The AFxxx devices are smaller size so it might not be suitable for small amplifiers that drive speakers.   In the context of pedals that's not really an issue as they operate at low current.   The high frequency devices will have lower capacitance so they might not sound as smooth as the ACxxx devices.   That being said they will sound more like a germanium than say a silicon transistor.  And you could add small caps between the collector and base to roll-off the highs - that idea does work and people do this to silicons as well.

[Edit:
Cob 
AC126 = 50pF,  AC128=100pF
AF126  = 2.5pF

https://archive.org/details/MullardDataBook1965-66
]

[ilcaccillo]

Thank you so much,
so basically I should be careful with HFE selection in AF types and also mind the lower internal capacitance, besides that for fuzz pedals or boost circuits they are totally usable, is that right?

[Rob Strand]

so basically I should be careful with HFE selection in AF types and also mind the lower internal capacitance, besides that for fuzz pedals or boost circuits they are totally usable, is that right?

Yes they will certainly work.   You have to be a bit careful the hFE and leakage of all germaniums.