AC 125k/z V transistors, FF Ge Geofex test, fluctuation.

[Thecomedian]

I used 6 resistors to get 2.2m and 2.472k for each section of the tester http://www.geofex.com/Article_Folders/ffselect.htm not sure if that makes a difference, but when I put any device under test, there's usually a large fluctuation. AC 125s start off with 1.2v showing on test without the base connected, which would mean 500uA leakage or more. Switching on the base turns that into 2.3-3.0v, but if I leave the thing alone for a while it will eventually fall to 0.64v and switching it on turns to 1.4v.

Is the true gain taken from waiting until it gets near to settling down? How come it takes so long to settle down? Is it related to using multiple resistors? measuring the resistance behaves the same way, with the meter moving around until a settle, which is slower for series resistances compared to a single resistance?

[LucifersTrip]

there's too many threads here to reference. do a bunch of searches.  ge's are temperature sensitive. if you don't want to wait as long for them to settle to get the true reading (at your room temperature), don't handle them with bare hands.

edit:
it's not the multiple resistors, but you can use a trimmer

[Thecomedian]

I know about the temperature issue when handling. When trying to account for and minimize all the variables, the thing still takes about 4-5 minutes to settle. I don't recall any of the other threads going beyond temp sensitivity to examine the meter instability.

[PRR]

These parts are old. And not too clean when new.

Maybe it's like an old movie projector. You have to burn the dust off the lamp.

I'm curious. Leave one in the tester overnight. Or maybe a week. Does it stabilize?

[Thecomedian]

Im running one for 5 minutes. it went down from 1.2v with only leakage, to eventually down to .445v, and now it's climbing slowly again. currently at .468v. turn on the base and get 1.035 (55-56 gain), turn off again, and it's at .474, holding steady. Another cycle, 1.045v and climbing, off .483v and holding steady.

I found some test probe sinks, so I can just shove the probes in them and read the circuit all day long.

[LucifersTrip]

5 minutes is not really that long of a time if they're leaky, especially if they were handled or were moved from a location with a different temperature.  the low leak ones will usually take shorter to stabilize.  

when I finish breadboarding a circuit with ge's (which I think may be my final choice of transistors before soldering), I always play thru it for at least 30 minutes, before I finalize and take voltages

edit:
if you've set up the tester correctly and have no temperature changes and still can't get it to settle in a reasonable time, then probably not a good enough transistor to use in a circuit

[duck_arse]

put some sort of "windshield" around your test socket, at the very least. it will stop the breeze from the open window cooling yr parts. I just wait till I get a reading I like, write it down and yank the transistor.

[Thecomedian]

5 minutes is not really that long of a time if they're leaky

well like I said, the change will slow greatly at a certain point, and even start climbing again.

[Arcane Analog]

Put them in a circuit and play them for a while. If they sound good, they are good.

[mac]

When you connect the base you are heating the guts of the transistor.
If the transistor is leaky, you'll notice a proportional drift immediatelly after you connect or disconnect the base.

Take the values you measure as an estimation, in six months, tomorrow or in an hour you'll get very different readings.
Some good sounding AC187 I bought have variations of 100% through the year, and when tested do what you described.

What is really bad is when the voltage never reaches a plateau. Coincidentally, I noted that this ones are very noisy.

mac

[italianguy63]

I just did this lately, and tested my find of about 75 NOS TO-5's.... They took several minutes each to stabalize.  It's not just you!!  MC