Transistor testing (FuzzFace) Still messy in 2024

[brett]

RG's transistor tester (Geofex, leakage compensating) uses a tiny base current of 4uA, similar to Q1 of a Fuzz Face (base current of 1 to 3 uA), so it's a much more realistic test of Hfe for a FF. I've got RG testers for both PNP and NPN transistors.

I just measured nine silicon transistors (2N2222) with an average Hfe and around 290 on my DMM (range 265 to 326) that averaged just 155 on RG's tester (range 130 to 187).  There was no relationship between the results from the two testers.  So the DMM is a useless tester, given that RG's tester has a good track record.

Question 1: Does anyone know how the preferred/golden/high-priced Hfes in a fuzzface of 70 and 120 for Q1 and Q2 were measured?  In a test rig or in-circuit?

In the one device, a BD139, where I've measured all three Hfes, it was 150 on my DMM, 112 on RG's tester and 55 in-circuit.  It sounded quite good as Q1, and from memory it gave plenty of gain.

Continuing with only RG's tester, I had some 2N6517s that tested 108, 110, 112, 113 and 121.  Using pair 108 (Q1) and 121 (Q2) in a FF gave fairly soft, lower-than-I-expected fuzz.

My next task is to replace the 108/121 combo with a pair of 2N2222s at 130 and 150.

Thanks for reading.  Any comments?  Maybe I'm missing something quote important?

I'll leave you with a final observation/mystery.  I tested some BC337s.  These are all-purpose Si NPN BJTs.   They were all high (eg 155), but I accidentally put one in backwards (collector for emitter and emitter for collector)... and it tested better/lower (140).
Then ... in the DMM it tested 180 the right way around and only 19 the wrong way.  What's going on?  Testing for leakage in RG's tester showed that the wrong way around, it's 140 "Hfe" of leakage, and only 15 of real gain.  I had no idea silicon transistors can leak like this.  A trap for a lazy player like me, not checking pinout and testing leakage on Si devices.

[mozz]

Every brand of meter is going to give you different results.  Do not believe ( me anyway) the 70-120 hype.  I have found higher gain first often sounds better to me.  Also depends on the guitar. I want cleanup on 6-8 on a strat with audio taper vol. I use a DCA55 when searching for a certain gain.

[Rob Strand]

I just measured nine silicon transistors (2N2222) with an average Hfe and around 290 on my DMM (range 265 to 326) that averaged just 155 on RG's tester (range 130 to 187).  There was no relationship between the results from the two testers.  So the DMM is a useless tester, given that RG's tester has a good track record.

Usually there's no problem with RG's test method but since the measurements are low you might want to check the battery is OK on your RG tester.

I'd at least expect a correllation between the results.   The DMM usually test at variable currents, higher collector current for higher hFE, no different to RG's method.   *However*, if the DMM test current is low it might be measuring the low current hFE which can be quite variable on some devices.   The true low current hFE is lower than the high current hFE *but* when the DMM is calibrated it can be made to display a higher value (perhaps even trying to show a 1mA hFE by scaling up a 100uA hFE measurement - that would be all over the place.)

I'll leave you with a final observation/mystery.  I tested some BC337s.  These are all-purpose Si NPN BJTs.   They were all high (eg 155), but I accidentally put one in backwards (collector for emitter and emitter for collector)... and it tested better/lower (140).
Then ... in the DMM it tested 180 the right way around and only 19 the wrong way.  What's going on?  Testing for leakage in RG's tester showed that the wrong way around, it's 140 "Hfe" of leakage, and only 15 of real gain.  I had no idea silicon transistors can leak like this.  A trap for a lazy player like me, not checking pinout and testing leakage on Si devices.

It's absolutely normal for transistors to work with the C and E reversed.    You might see reverse gains between 1 and 20.  If you see leakage on RG's tester it could be because the BE junction is breaking down a bit as RG's tester uses a 9V test voltage.   DMMs often test at some lower voltage like 2.5V.

FYI: You can use different base resistors on RG's test setup to test transistors at different collector currents.   However, you need to redo the conversion formulas to account for the different scaling.    If you used a 22M base resistor and 10 times the collector resistor it would show a low-current hFE value.

In some cases you can get RF problems on RG's unit and you can try putting caps across the base and emitter of the transistor (only for hFE, not recommended for leakage.).

Try not to touch the leads or transistor while testing.  Even heat from your hands can change the result, it could even make a DMM measurement rise as the VBE drops with temperature and that would increase the base current.