Germanium Transistors ...Measuring HFE

[copperheadroads]

Been playing with pedals in my spare time & built & half dozen or more .
My question is are Multimeters that have the Hfe setting with the hpn/pnp pins accurate ?
I have 2 meters the give me almost exact same HFE readings .
Should I use this as an accurate HFE reading or should i use Smallbears scamatic & use there method of measuring Hfe
https://www.smallbearelec.com/HowTos...eFAQ/FFFAQ.htm
The mail reason for the post is I purchased 50 Russian Ge transistors GT313V 's that gain should be from 30 to 170 hfe & they are mostly are in the 30's +/-
only 2 are high as 50 hfe ....

[52 Bill]

My question is are Multimeters that have the Hfe setting with the hpn/pnp pins accurate ?

I'm sure that that they are, but the question is what are the parameters that they are testing at?

Should I use this as an accurate HFE reading or should i use Smallbears scamatic & use there method of measuring Hfe

If you are trying to compare the gain ratings of your new Russian transistors to the gain figures that Small Bear suggests, you should probably test them with the circuit that they use. That way you are comparing apples and apples, and it will also show you if the built-in hfe meter readings are similar or vastly different. Does your meter read for leakage as well?

[J M Fahey]

What Smallbear is literally saying is that if you inject 9uA into that base (9V through a 1M resistor) you should measure between 0.72mA and 0.99mA through those collectors.

Measuring in uA that gives us between 720 and 990 uA .

If we consider that Smallbear considers acceptable a Germanium transistor with leaks/losses "up to" 300uA, meaning a full 1/2 to 1/3 the "useful" current, we see how crappy Germanium is and why it was abandoned to never return.
So test the Smallbear way and post results.

To measure leakage current, you ground base and measure what current they are still passing with 9V on collectors.

[copperheadroads]

I had a chance to test & compare a couple transistors using the Hfe on my meters against the Smallbear circuit & they are very close numbers ,but i was not sure exactly what setting to use ,I did get my results with the 20m uA setting
then the Smallbear circuit crapped out ...What a waste of time

[Mick Bailey]

In general measuring germanium transistor hfe on a multimeter is inaccurate because the measurement techniques don't factor in the relatively high leakage compared to silicon. The leakage increases the apparent hfe reading shown on the meter, but the actual gain 'in circuit' is very much lower.

BUT

Post-1970s Russian Ge transistors have almost zero leakage and can be measured using a multimeter just like silicon. Ge continued to be developed and manufactured in Russia long after being viewed as obsolete elsewhere. Just watch temperature rise through handling as this skews the reading. The transistors you've got are probably from a low gain batch. This is a fairly common occurrence nowadays, because anything with a higher gain is getting skimmed and sold at a premium price. I've had good batches where the average hfe is around 120, with about 5% of those no lower than 80.

Note that some circuits rely on leakage to bias correctly - for example, the Maestro FZ-1 won't work with Russian transistors unless biasing resistors are added.

[km6xz]

Mick is right, high Hfe and low leak transistors are available but they are selected out and used for replacements in industrial equipment that is still serviceable for a price that matches their rarity. I go to an electronic flea market here and get a handful of transistors for a low price and test them on my curve tracer and find a few that are in the 50-60 range, most lower but can go to a wholesale industrial electronics store and pay 10 times as much, about 150 Rubles, $5 and get the same part number with close to silicon specs, very low leakage and 100-160 Hfe. If it is on eBay or sold in batches, it is 100% guaranteed to be culls that work but fail QC tests.

I find that the static Hfe range on a DVM tells you good information if testing the same part numbers against other samples of the same number to get a relative ranking of your samples but since the conditions of test are not published, they can't be expected to match the the numbers from the spec sheet. So the meter is fine for matching a box of power transistors but not usable in predicting performance in the circuit in a new design. Use a curve tracer for that.

[Mick Bailey]

The impossibly low leakage figures suggests maybe different construction techniques for the Russian transistors compared with their western 'equivalents'. Where the leakage is high with a particular sample I've found that there are other problems, mainly very low Hfe and odd distortion or device breakdown. Sellers have latched onto the pedal mojo market and prices have increased dramatically, as well as spurious claims for obscure devices being equivalent to OC44, AC128 or whatever.

[J M Fahey]

Ditto on the Industrial Electronics angle .... and on anything "non consumer".
A friend of mine used to service Medical Equipment at Public Hospitals, which was obviously very high quality (none other acceptable in Life Saving equipment) but many times bought long ago.
Administrators don't accept buying new equipment if "they already have it" , being a Government supplier is very tedious, but there *is* money available for repairs, so old equipment is kept working for ages, sometimes having new add-ons.
So my friend told me about equipment (say a gas-in-blood analyzer) which at the same time had tube sections (including pencil tubes) in the PH meter head, germanium transistors in the converter used to power said tubes, silicon transistors elsewhere, and he replaced a digital readout module based on 74xx logic with a microprocessor based one !!!!!!!!
ALL generations under one roof !!!!!
But they apply the "if it ain't broke don't fix it" and they feel comfortable with the old but trusty equipment they've been using for ages.