Silicon or Germanium tester?

[LucifersTrip]

I believe the Peak analyzer can tell you if a transistor is Silicon or Germanium (and a thousand other things), but is there
a circuit/schematic that was posted here that does that one simple test? ...or isn't it simple?


thanx

[R.G.]

I don't know of an instrument or circuit that does ONLY that. But a DMM, used on the low ohms or diode test range tells you directly something of the voltage drop across a diode tested there. That tells you whether it's silicon or germanium directly. The meters vary a little in what silicon does versus germanium does in terms of exact reading, so it's simplest to try your meter on a few silicon junctions and germanium junctions to see what reading your meter does when presented with the problem. After that, it's easy to tell them apart.

[LucifersTrip]

I have to admit I'm a little surprised. With all the psychosis over germaniums, I  thought someone would've designed the circuit to
test if a transistor was ge or si.  I can do it with a DMM, but it's inconvenient to pull out the clip leads when your hunting. It'd be nice to
just slip it into a socket.

thanx much for the answer.

[darron]

i have the peak tester and i really do love it... it also give you a current leakage readout and pinout...


other than being able to just see that a transistor is probably germanium or silicon, or know by where it came from, it seems that RGs method is so simple that all you'd have to do is make a diode board with a socket and alligator (crocodile?) clips.

i don't think there would be much of a market for something that does so little when you can buy a complete readout device. for example, most people buy a complete digital multimeter these days and not a voltmeter, ohmmeter, capacitance meter, continuity tester etc. etc. in individual components.


again, i love my peak

[PRR]

I've had mixed results using DMMs. I have not used that many. Some never expose more than 200mV at the probes. This won't read Si and is borderline for Ge. Best try your DMM against some known parts.

A classic ANALOG resistance meter will tell the diode drop.

I like VTVMs with 1.5V battery for resistance testing. The diode drop reads directly on the 1.5VDC scale.

The passive multimeters with 1.5V battery also work, but read backward. That means mental math, bah.

Pick the range which would read 1K mid-scale. The internal resistor is then 1K. Assuming 1.5V battery, the maximum current is 1.5V/1K = 1.5mA. This is safe.

You could just use a 1.5V battery, 1K resistor, and any volt meter.

Read the diode (or transistor). If infinite, swap the leads. Still infinite? Pick another transistor leg (you got the C-E connection which is always infinite).

Silicon will read about 0.6V. Germanium will read about 0.25V.

Nice thing about an ohm meter: run the range higher, 10K and 100K mid-scale. A diode's drop will change very little: 0.6V to 0.54V. A resistor's drop will change a lot. Using this fact, you can diode-check in-circuit, as long as shunt resistances are not too comparable to diode impedances. (In cheap/simple pedals, the resistors often are comparable to Ge leakage, so truth is unclear.)

The actual current is 1.5V minus the indicated diode drop, divided by the mid-scale resistance. So 0.3V indicated is 1.5V-0.3V = 1.2V, 1.2V/1K is 1.2mA.

For power transistors, you can go down to the 1 ohm mid-scale range, flow a large part of an Ampere of current. (Most C-cell ohm meters won't do this.) Don't go there with vintage Ge parts!! They were sometimes rated as low as 20mA max, we rarely run them much over 1mA, so a near-1mA reading is valid and safe.

[tubelectron]

Hi,
This tester exists, but of course, it does more complete tests both on Ge and Si transistors, based on the classical transistor measurement circuits : Hfe, Ice0, Icb0, Ices. You can DIY yours : the schematic is on the 2nd picture. Nothing's really new here, finally...





A+!

[.Mike]

Correct me if I'm wrong, but I think that R.G.'s tester would work just fine for determining silicon vs. germanium.

1. Put the transistor in the socket with the switch on.
2. Grasp the transistor with your fingers.

If the gain rises, it's germanium. If it doesn't, it's silicon.

Mike

[tubelectron]

Hi,

Assuming that the condition of the transistor is correct,

If the gain rises, it's germanium

Yes - Ge is very "gain-touch-sensitive"

If it doesn't, it's silicon

Yes and No - the gain raises also, but usually much less than on a Ge device.

It's a way to determine if the Q is Ge or Si, but I don't know if it is 100% reliable. Nonetheless, it doesn't give you reliable information on the transistor's condition or performance.

A+!

[LucifersTrip]

Thanx much for all the ideas and especially Paul for the detailed rundown. It gives me some ideas.
I can pretty much figure it out with my DMM, but I didn't want to carry that around with me, make measurements
and bring clip leads.

I guess what I was hoping for was something simple like the transistor tester I built:
http://www.aronnelson.com/gallery/main.php?g2_view=core.DownloadItem&g2_itemId=2724&g2_serialNumber=2

...where you just stick it in a socket, but in this case the lights to tells you if it's Si or Ge.  

Remember, it doesn't have to light up red for silicon or green for germanium. I only need a test to tell if it's EITHER Si or Ge.
If it's simpler to test for Si, it could light up red for Si and not light up at all if it's not. By process of elimination it would be Ge.
(There will mostly be only 2 cases since I most likely will not be using it for any fets, darlingtons or anything other than old 60's - 70's metal case Si and Ge transistors.)

thanx again

[Rob Strand]

Use a circuit like this,

http://circuitspot.com/wp-content/uploads/2010/03/pDiode_Blog_1.png

with V=1.5V, R=2.7k

NPN Transistor, connect C to B then to Anode on test circuit, and E to Cathode on test ckt
PNP Transistor, connect C to B then to Cathode on test circuit, and E to Anode on test ckt

The trick is to use the right amount of current.  Too high and the ohmic drops in the Ge
devices can artificially raise the measure voltage, too low and the leakage affects the result.
I'd say something around 500uA would be good.

The above checks voltage drops only it wont tell you the transistor is working, if it's blown
the ckt could measure zero.

[snarblinge]

or buy the peak, I couldn't be happier with mine, for mining the random transistors at the surplus places. it is truely awesome.

just wish I'd bought the cap one at the same time, shipping is killer to NZ...