RG's method of testing transistors

[jlullo]

hey guys... in this diagram:
http://www.geofex.com/Article_Folders/ffselect.htm

i'm wiring up the top and bottom lugs of the switch right?  does the middle just stay empty?

[mdh]

You're wiring up any two lugs of the switch that can be made to close (be in contact with each other) by throwing the switch/pushing the button, whatever. In a switch that has "top and bottom" and "middle" lugs, you probably want the middle lug and one of the others, because the middle lug is probably the common lug of the switch.

[jlullo]

ok.. thank you, mdh!

[jlullo]

hey guys, i breadboarded this circuit and am having difficulties getting it to work.  whenever i flip my switch, my DMM's display doesn't change.  it keeps showing the voltage of 2.47...

i've checked all of my connections and the connections to the switch, took it apart, re-breadboarded it, and nothing... any ideas?  when i touch the contacts on the switch my voltage changes, but that's all

[mdh]

Since you're doing this on a breadboard, it may be easier to just use a jumper wire in place of that switch, or just connect the "free" end of the 2.2M resistor to the base of the transistor when you want to get your gain measurement.

Other than a miswired switch, you may just have a bad transistor or have the pinout wrong (though I'm not sure if the latter could give you the measurement that you're getting).

[jlullo]

thanks mdh, i'll give that a shot!

[jlullo]

wow... i'm a complete idiot... i just realized that that schem is wired for pnp, and i was testing for npn 2n1304 devices... let me go fix that now that i probably toasted at least 3 trannys

[jlullo]

alright, i think this is working correctly, but i still have one question...

here are the formulas i'm using (found in another thread):
- gain = (Von - Voff) * 100
- leakage (µA) = (Voff / 2472) * 1000000


every time i have Von (switch on), i'm going to have a value of 2.47, correct?  this is what i'm getting every time, and i'm assuming is correct because it is reading the value of the resistor i'm using.  i'm having a problem with the leakage formula though.  let's say that my Voff value is 2.31.  This would give me a gain of 160.   Now, to get my leakage i divide 160 by 2472, then multiply by a million.  this is giving me a leakage of 934.446.  How is this possible?  especially since if the above formula is correct, you're not going to stray that far from that result in leakage ever

[jlullo]

for some reason i'm having a really hard time with this guys.  i took it apart again, and instead of a switch am using a jumper (per mdh's suggestion )



here shows my set up.  the green leads are connected to a 5k pot, that i have dialed in to exactly 2.47k.  that is connected to the 2.2m fresistor, and they are both going to a supply of +9v.

the other lead of the 2.47k is going to the collector of the tranny (i have my tranny oriented backwards in this picture, because i  was trying to see if i was insane and had it backwards).  the emitter is jumped to the ground.  the purple lead is the jumper from the 2m2 resistor.

when i'm taking resistance measurements, i have the black lead on the collector, and the red on the lead oconnecting the 2.47k to the 2.2m.  now, all of my resistance (with or without the jumper) are reading 2.47k

do you guys see any problems with this?

[mdh]

I think something is still wrong.  If you read R.G.'s writeup on the page you linked, it should be clear that Von (that's with the switch closed, current allowed to flow to the base) indicates the gain + leakage.  Multiply that voltage by 100, and you should get roughly the number that a DMM with a gain measurement would indicate.  So if your transistors are really consistent, then maybe you'll get the same Von with different transistors, but this would be sort of surprising.

Voff gives an indication of false "gain" due to leakage.  Imagine that the leakage is really, really small.  Then transistor is effectively a really big resistor, so almost all of the voltage should be dropped across it, and only a very small voltage should appear across the 2.4k resistor.  If you're seeing numbers like Von = 2.47 and Voff = 2.31, that would seem to suggest to me that you have a very leaky transistor, with actual gain of only about 16, not 160.  I suppose it's possible that you might actually be seeing this, but I don't have a lot of experience with germanium transistors.  It seems to me that you could check by plugging in a silicon transistor, and verifying that Voff is really, really, really small, and Von * 100 is close to the gain that you would expect for that part (or you could check it in a DMM with hfe setting, if you have one).

p.s. From the picture it looks like two of the transistor leads might be shorted to each other near the body of the transistor.