Need help identifying transistors "SGS 87117 S154"

[jjj]

Hi,

I got some old hifi amp manufactured in 1970 for 3€ and I wanted to scavenge some parts from it, so I opened it and found what looks like transistors in (I think) TO-18 cases, written on them "SGS 87117" on one line and "S 154" below. I searched the whole web for two days and came up with absolutely nothing about these.

As I'm quite a newbie please correct me if I'm wrong : I assume they are germanium transistors, germanium because the amp was manufactured in the year 1970, and transistors because of the shape of it and that there are C B and E markings on the PCB (so it shouldn't be something else than transistors).

Does someone know anything about these How could I test them with no info about them - could I just determine hFE and guess its voltage limit (considering the fact that I won't submit them to more than 10/15volts max) or is there a way to determine their max voltage without burning one

Thanks a lot if you can shine even a dim light onto my newbie path

Here's a picture of those aliens, sorry for the quality (you know that aliens pictures are never sharp hehe)

[Govmnt_Lacky]

Well....

If it is a 2SA154 then it is a Germanium PNP transistor. 15V max. Low-ish Hfe but they would need to be tested.

This is a guess, of course, based on the info provided.

[smallbearelec]

SGS was a fairly well-known Italian maker of semiconductors. The part number likely a "house" number rather than an industry standard desi.

--Carefully unsolder one from the board, heatsink on lead.
--ID Collector, Base and Emitter.
--Use your multimeter's diode test scale on the B-E junction. Germanium will show about 200 mv. in the forward direction, silicon about 600 mv. This test will also tell you whether the parts are NPN or PNP. If the positive side of the meter is on the Base when the junction is conducting, it's an NPN device.

Now you can test for gain.

[jjj]

Thank you both for the replies.

I did the diode test, it shows 1065mv for the B E junction and 1055 for the C E junction (negative side of the meter on the base, so PNP). So they're (double) silicon transistors Did they make silicon transistors in 1970

Another strange thing, I did the gain test like described at geofex (http://www.geofex.com/Article_Folders/ffselect.htm) and when the 2.2M resistor is not connected, it shows 0.000v across the 2K4 resistor. Does it mean that the transistor has no leakage I don't think that's possible is it

For info the DMM shows shows an hFE of around 200, the test at geofex around 190 (so the voltage across the 2k4 with the 2M2 connected, minus 0 as it shows that on the first stage of the test...).

Thanks for helping me.

[R.G.]

I did the diode test, it shows 1065mv for the B E junction and 1055 for the C E junction (negative side of the meter on the base, so PNP). So they're (double) silicon transistors Did they make silicon transistors in 1970

They made silicon as far back as the mid 1960s. They're silicon transistors.

Another strange thing, I did the gain test like described at geofex (http://www.geofex.com/Article_Folders/ffselect.htm) and when the 2.2M resistor is not connected, it shows 0.000v across the 2K4 resistor. Does it mean that the transistor has no leakage I don't think that's possible is it

Zero leakage? No. It's just too small for your meter to read it. One of the reasons the industry went to silicon and didn't look back is that silicon's leakage is on the order of a thousand times smaller than germanium. Your meter reads to a millivolt, right? A millivolt across 2.4K is 416nA. Silicon transistors commonly list maximum leakages in the 100nA and lower.

For info the DMM shows shows an hFE of around 200, the test at geofex around 190 (so the voltage across the 2k4 with the 2M2 connected, minus 0 as it shows that on the first stage of the test...).

That's reasonable.

[PRR]

Hi-Fi in the _1970s_ was almost all Silicon. Ge was well on the way out in the 1960s. Si leaks less, allowing more freedom of design. Anything more complicated than a pocket radio, Si's higher cost per device was moot. And I think even $4.99 5-transistor radios were going Si by 1970.

Really good Si was available before 1965. And Planar, which became much cheaper than the older techniques. High price at first, but TI and others started making them by the millions for mainframe computers. Prices fell, especially for "rejects" which just-missed the stringent specs for specific computer applications but were perfectly good for applications designed around the available parts. The use of Ge in late-1960s guitar pedals was just trailing-edge design and bargain-barrel parts buying.

[PRR]

me> Hi-Fi in the _1970s_ was almost all Silicon.

Not the first and far from the last.