Ge Transistors hfe too high

[Elijah]

Hi folks. I've got some GT404's and they show to high hfe around 200. Datasheet shows max on them 150. I just wonder if my multimeter is lying or tranies are dying?

[PRR]

hFE of transistor depends on current. On Ge parts it can depend a LOT on current. I would not worry.

[Elijah]

Thank you!

[Rob Strand]

Don't forget a DMM will measure leakage as well.  No real way to workout how much extra gain your DMM is showing due to leakage.  You should check out RG's page on measuring leakage and hFE of germaniums.
The difference is RG's method separates the leakage part then works out the true gain.

[Bunkey]

I've got some 402B's measured with a DCA55 here,
An 85hfe @91uA, 140hfe @ 88uA and a 205hfe @ 219uA.

You'll probably see an increase in leakage with the higher figures but it's still entirely acceptable by all accounts.
They're pretty consistant devices so I'd bet yours is something similar.

FYI the 140hfe transistor measured with the DCA55 measures ~200hfe in my DMM

[Elijah]

Thank you brother!

[Elijah]

Don't forget a DMM will measure leakage as well.  No real way to workout how much extra gain your DMM is showing due to leakage.  You should check out RG's page on measuring leakage and hFE of germaniums.
The difference is RG's method separates the leakage part then works out the true gain.

Took notes on that. Appreciate all this!

[any]

Just doubling up on what's already advised. Build a simple circuit (perfboard if needed, but I just soldered one up years ago)
as per RG Keen's advice on: http://www.geofex.com/article_folders/ffselect.htm

Also, measure at 'room' temperature... say 18c/64f as a hot or cold room may also change gain.

[mozz]

Doesn't RG's circuit accuracy depend on accurate resistors and exactly 9volts? If your gains are really that high just add emitter resistors.

[Rob Strand]

Doesn't RG's circuit accuracy depend on accurate resistors and exactly 9volts? If your gains are really that high just add emitter resistors.

Yes but it's good enough.  You can trim-up RG's circuit to produce a good enough readings for casual use.

If want more accuracy you can measure all the circuit values and the battery voltage at the start of the test.  Then use a spreadsheet to do the calculations and you should get very good accuracy.  For accurate results Vbe should also be measured and factored in as it affects the base current.

If you want something that doesn't need to be calibrated each time then you need to have a circuit built with stable regulators and have a current source driving the base so the base current doesn't depend on the Vbe.   The circuit becomes more convenient to use but more complicated.

[mozz]

Yes I have a DCA55 so my handbuilt rg tester sits idle. Just never thought it was that accurate and how often do you find a battery at exactly 9v? I've never really gone by the 80-120 gain position thing. Often they sound better with them swapped and have often used trannies with a lot lower gain and sound fine. Higher gain and higher leakage, save them for your next build which after a fuzz will be a tonebender.

[R.G.]

Yes I have a DCA55 so my handbuilt rg tester sits idle. Just never thought it was that accurate and how often do you find a battery at exactly 9v?

The problem with that concept is that you really can't ever get a single number for the current gain of a germanium transistor. Even putting a range on it doesn't do much unless you also specify the temperature and test current ( and possibly humidity, moon phase, planet, etc.) that the measurement was taken at. All bipolars have gains that vary with collector current; in fact, that is the way that most VCAs work; they vary the emitter current of a differential pair to vary gain. Germanium is hugely dependent on temperature, of course.
A digital measurement device like the DCA55 is great - except it probably does a single point test. That only tells you the gain, however precise and accurate it is, at that time, temperature, current, etc.

Given the huge variation, it makes sense to think about whether a battery of exactly 9.0000... volts and resistors of 0.0000001% make a real world difference in what the test tells you about the transistor. Our electronic measurements course harped on the need to do appropriate measurement. Measuring something that inherently varies by 50%-100% down to a precision of 0.01% can spend a lot of money and time telling you not much useful other than - it varies.

[mozz]

Testing a high gain pnp germanium on the DCA55, gain 235/ leak .163. Tested on a Radio shack bench meter, gain 285, tested on a different model Radio shack bench meter, gain 350.

Not sure if the OP knows, just holding them with your fingers to place them in the socket will make the reading higher. The meter does not take into account leakage thus you are getting a reading that is actually higher. Also,most Russian transistors i have tested will read towards the lower end of the advertised gains.

[Elijah]

Thank you guys!

[kaycee]

I have 3 meters that I use. 2 digital multimeters and one of those multi component testers. For resistance and capacitance, they all pretty much agree, for hfe of both germanium and silicon transistors, they all vary.

For my casual usage, it's fine. I generally want to know if a transistor is duff, very low gain, has ridiculous gain, or ridiculous leakage. Most of the time, if I measure on the same device, the comparative measure is enough and I'll use my ears for the rest.

[pinkjimiphoton]

them cheezy 20$ transistor tester kits from china that are all over ebay work great for this.
they seem to give a fairly accurate reading for gain, even with GE's, and are a worthy addition to the ol' bench.
they seem to give accurate results when my other meters will feed ridiculous gain levels that can't possibly be right.

ge transistors will NEVER give ya the same reading twice. its impossible. so "close enough" is usually fine. temp, cycle of the moon, radio frequencies from alpha centauri,  wind, sound, light... all have effects on these suckers. seriously.

i think even google searches can influence them. i'm not kidding.