[NEWB Q] pls help with R.G.'s transistor hfe measurement circuit

[gaussmarkov]

i could really use some help with R.G.'s hfe (current gain) measurement circuit.

here's a figure of my breadboard of this circuit:


i am testing a 2N3904 and a 2N5089 with this circuit.  but, because of my inexperience, i am worried that i am making some mistake because some of the measurements seem wrong.  the 2N3904 seems to leak and the 2N5089 seems to have an hfe of about 5.  my understanding is that silicon transistors leak negligibly and their hfe should be in the hundreds.  i hope you will bear with me because i have tried to do this really carefully.

i used the diode setting of my dvm to check the forward voltages, following something i read somewhere.

Direction	2N3904	2N5089
C->E	0	0
E->C	0	0
B->C	0.895	0.892
C->B	0	0
B->E	0.886	0.880
E->B	0	0

nothing much different here.  i hope these are about right.  the zeroes are in the right places (as i understand it).

i measured the resistors (1% film) and they are within spec.  (my dvm does not measure resistance above 2M so i used a 1M resistor in parallel to measure it.)

first, i removed the 2.2M resistor and read the voltage across the pair of 1.2K resistors in series.  2N3904 gives 0.93V and 2N5089 gives 0V.  both should be 0V, right?  the 2N3904 is leaking right?

second, i put one lead of the 2.2M resistor back in the +9V rail and stuck the other lead in an unconnected rail.  i read the current from the open lead of the resistor to the base pin for each transistor (A7).  my meter was set at the 2mA scale and i read 0.004 for both.  so the base current is approximately 4uA, as R.G. says it will be.  to double check my understanding, i put the 2.2M resistor back across the supply rail and the base rail and read the voltage across the transistor.  2N3904 gives 8.49V and 2N5089 gives 8.55V.  dividing these by 2.217M (my read on actual value), i get base currents equal to 3.83uA and 3.86uA, respectively.  so i think this checks.

third, i read the voltages across the pair of 1.2K resistors (with the 2.2M resistor still in place).  2N3904 gives 0.94V and 2N5089 gives 0.045V.  using R.G.'s instructions, i subtract 0.94V from 0.93V and get 0.010V net for the 2N3904 implying an hfe abround 10.  for the 2N5089, the measured hfe is around 5.

fourth, i move the lead of the 1.2K resistor at C6 to an open rail and measure the collector currents directly.  2N3904 gives 0.380mA and 2N5089 gives 0.019mA.  so the 2N5089 has a measured current gain of 0.019/0.00386=4.92, right?  this seems to check with the voltage reading above.

i get similar measurements for other 2N3904 and 2N5089 transistors in my stash.  it's not just these two.  i really doubt that the 2N5089 transistors are bad.  they work beautifully in a ts808 clone that i have on my breadboard.  i suspect that the 2N3904 transistors are bad.  they do not work in the ts808 circuit.  but if i am doing this correctly then the hfe on the 2N5089 should be something like 400, right?  more evidence that i am making a mistake is that i just got a pair of germs from small bear with measured hfe, 190 and 173 respectively.  i get numbers like 6 using this method.  same wrong order of magnitude.  so surely i am making one of those mistakes you just cannot see from having stared at the thing for a long time.
but the leaking 2N3904s seems inescapable.

as always, i'd appreciate any feedback.

[R.G.]

One thing that strikes me is that if I remember the pinout of the 2N3904 and 2N5089, the transistor you have drawn in is backwards.

The C, B and E letters are in the right places on the breadboard, but the body is facing the wrong direction if my memory isn't failing like my eyesight.

[gaussmarkov]

     you're right!!!  aaaaagggggghhhhhh!   another backwards transistor!!!    well, at least i drew the picture accurately.

oh man.  i feel so much better.  i have been hitting my head against this wall for a couple of days.  beers for everybody on me!!!  yippee!!!!

it's great to be a newb.     thank you, thank you, thank you.  what a twit i am.   i was about to go out and buy a dmm with transistor gain measurement (and for measuring caps, too).  i was getting that desperate.

now i have to go see if this fixes my bread board of joe davisson's aniquity fuzz. ... and see whether putting transistors into a ts808 clone backwards really still "works."  maybe i put them into that 2 x backwards = correctly!?!

[MartyMart]

That DMM doesn't sound like a "bad" idea .... VERY useful if not "INVALUABLE"   

Marty.