2N280 Question

[usurp]

Have a quick question regarding 2N280 PNP transistors that I recently purchased for a hudson broadcast clone.

I picked up these bad boys from small bear (https://smallbear-electronics.mybigcommerce.com/european-2n280/) and they list it in their "bulk lot" pile of transistors. I ended up purchasing these because during my research I read that these were pretty similar to an OC71 (while also easier to find and way cheaper).

On the smallbear website in the description it mentions that they have a gain range of 30-50. However, while testing the hfe of them on my multimeter (this one: https://www.harborfreight.com/7-function-digital-multimeter-63759.html) I'm getting readings of 106 and 110 respectively. This isn't necessarily a bad thing because from my reading I saw that for this pedal any hfe rating between 70-120 should be good for this pedal.

I guess my questions are this; why do you think these transistors are reading much higher than what smallbear was reporting or am i missing something really simple here. I know my multimeter is pretty basic and was wondering if maybe i could be getting some crap readings from it. This thing feels like I'm just missing something really simple here. I am very new to building my own pedals and seem like I'm confused much more often than not 

[mdcmdcmdc]

My total rube guess would be that the leakage from the transistors is probably borking the multimeter's ability to take a proper reading, and it's also hard to say how the multimeter is taking the reading in the first place.

The hfe readings that people refer to for stompbox purposes are usually taken using something like a DCA55, one of the eBay semiconductor testers, or a homebrew rig following the RG Keen method.

Also, germanium transistors are a crapshoot at the best of times.

[PRR]

hFE varies with current. On older devices, hFE varies A LOT with current.

Bear does not specify emitter test current; he puts 9uA in the base and got 270uA to 450uA in the emitter.

That meter (5 versions!) does not say what test current. Not a clue. Could be 1uA or 10mA. If higher than Bear we would expect a higher hFE reading.

[idy]

Fuzz fans use this to measure gain and leakage...and then subtract the leakage...

http://www.geofex.com/article_folders/ffselect.htm

Most who have gotten the Ge "bug" have built one of these.

[duck_arse]

and welcome to the forum. plug unknown transistors into breadboard, listen to results. good? then you don't care about the numbers.

[kaycee]

Socket the transistor and try what you have. I've built a couple of these and they weren't very particular. Mind you I never deliberately tried any of my known duff Ge's in one.

[PaulBass]

Those Harbor Freight multimeters aren't very accurate. You'll get better accuracy with a transistor tester. They aren't that expensive

[Rob Strand]

Those Harbor Freight multimeters aren't very accurate. You'll get better accuracy with a transistor tester. They aren't that expensive

Fuzz fans use this to measure gain and leakage...and then subtract the leakage...

http://www.geofex.com/article_folders/ffselect.htm

Most who have gotten the Ge "bug" have built one of these.

The main source of error when measuring hFE for germaniums is not accounting for the leakage.    If a component tester doesn't report leakage there's a good chance it will report high hFE on a germanium  because it adds the leakage.

An additional source of error is the lower Vbe of germaniums.  That makes old school DMMs read about 20% high and probably makes component testers read high by 10% to 20%.

A third source of error is the test current that PRR mentioned.

For many DMMs you can use the same subraction trick as RG's jig. 
- Measure with hFE normally
- Measure hFE again with the base disconnected to get the leakage value.
  (it is possible to work out the leakage *current* as well, for example on many DMMs a leakage hFE reading of 100
   is the same as 1mA leakage.  Some DMMs will have a different scaling.)
- Subtract the second reading from the first.

For many component testers I doubt the base disconnection trick will work since it won't be able to work out it is measuring a transistor.

Here's an example of a leaky germanium.

Actual value VCE=5V:      leakage ICEO=1.8mA, hFE = 60.2 @ 2.15mA, hFE = 66.5 @ 3.24mA

DMM:
- normal   hFE = 148
- base open hFE = 61
- corrected value with leakage removed  hFE = 148 - 61 = 87
- corrected value to account for DMMs measuring high 20% on germaniums hFE = 87/1.2 = 72.5
[- corrected value because the DMM is not accurate, high 8% (worked out using silicon transistor): hFE = 72.5/1.08 = 67]

The measure hFE = 148 means the DMM transistor test current is about 1.48mA.    We can see from the real hFE values that the hFE decreases as the current decreases.   We might only expect the real hFE value to be 55 to 60 at 1.48mA.
So the accurate value from the DMM is 67 and we should be comparing that to real value of 57.  So despite all the corrections the agreement is still only 17%.   If we didn't apply the corrections and just used the raw values from the DMM we would be comparing 87 to 57 (which looks bad but it's also incorrect).

I didn't bother chasing down where the disagreement is but I suspect it's related to the leakage drifting up.   The numbers increase by a factor of 2 (!!!) as the transistor warms up.   The potential for errors in both the DMM and real measurements is very high.  In fact I did a second run and the numbers varied 20%.   Lower leakage transistors will have less trouble.