Author Topic: How to measure silicon transistor HFE (sorry this may have been done to death)  (Read 7641 times)

Ghost Planet

Again, sorry if this has been done before, but...... I was wondering how you measure HFE in silicon transistors. I have a digital multimeter, but it doesn't have a socket for testing HFE. Thanks!

slacker

You can do it using R.G.'s  transistor checker here http://www.geofex.com/Article_Folders/fuzzface/fftech.htm#sound this works for silicon as well as germanium transistors. For silicon you don't really need to do the leakage test because there should be virtually no leakage. Switch the battery leads to measure NPN transistors.

Seljer

And for testing NPNs, tie the emitter to ground, connect the collector to your positive supply through a small value resistor (lets say 1kiloohm) and the base to the positive supply through a high value resistor (lets say 1Megaoohm),

A current will flow into the base over the base resistor. Measure the voltage over the resistor and use Ohm's law the determine the current thats flowing. It should be a couple of microamperes. If you haven't got any funny device, with silicon you can assume that the base-collector junction will drop about 0.6V, so with a 9V supply you should have 8.3V over the 1Mohm resistor = 8.3microamps

Then measure the current flowing into the collector by measuring the voltage drop over the other resistor and again with Ohm's law.

Current gain is just the ratio between the base current and collector current :)



The values chosen in the geofex tester do the math in advance so you can just read it directly from the meter (and basically it'll work for NPNs too if you just hook up the battery in reverse).

Ghost Planet

Cool, sounds easy. Thanks for the help.

Arcane Analog

Many multimeters have slots to test a variet of transistors with different pin configurations. If your multimeter has an "hFE" setting you should be able to test silicon transistors easily.

GibsonGM

Yes, but many don't....mine doesn't, altho it's a very useful Extech.   I rarely "need to know" (data sheets are enough for 99.9%), but when I do, I use the R.G./ Seljer method described.   
1M on base and a 1K on the collector, easy as pie.    B = base voltage/collector voltage.
MXR Dist +, TS9/808, Easyvibe, Big Muff Pi, Blues Breaker, Guv'nor.  MOSFace, MOS Boost,  BJT boosts - LPB-2, buffers, Phuncgnosis, FF, Orange Sunshine & others, Bazz Fuss, Tonemender, Little Gem, Orange Squeezer, Ruby Tuby, filters, octaves, trems...

Arcane Analog

The point is that he may have a multimeter that performs this test and not know it. Would that not be easier than pie?

For $5-$10 you can have a transistor tester that also measures DC, diodes, capacitance, resistance, etc, etc....

merlinb

The point is that he may have a multimeter that performs this test and not know it.
Read the OP's post in full.

GibsonGM

The point is that he may have a multimeter that performs this test and not know it.
Read the OP's post in full.
+1

Plus, for the low sum of $5 to $10, you will likely have a meter that sucks, to be honest with you.   OP also said his doesn't have the transistor test function....nice of you to try to help him out, tho, Analog!   Many good meters don't do hfe, and many do; depends on what you settle on.  If it does, then yes, that would be easier than pie...

The reason for doing this with NPN Si transistors is for the 'old school' ability to do things for yourself.  If you don't know how the meter does it, you have another hole in your knowledge, altho admittedly not everyone is interested in knowing more than just how to sling parts together.  This method is for those who do have that interest.

The reason you do this for Germaniums is to get gain AND leakage, which are VERY important in fuzz face circuits.  Your hFe meter doesn't give leakage...and may just roll all over the place, rendering a legible reading unlikely.  That's why R.G. wrote his original piece on testing Ge's - there was a need.

The OP's question was a good one...
MXR Dist +, TS9/808, Easyvibe, Big Muff Pi, Blues Breaker, Guv'nor.  MOSFace, MOS Boost,  BJT boosts - LPB-2, buffers, Phuncgnosis, FF, Orange Sunshine & others, Bazz Fuss, Tonemender, Little Gem, Orange Squeezer, Ruby Tuby, filters, octaves, trems...

Arcane Analog


The reason you do this for Germaniums is to get gain AND leakage, which are VERY important in fuzz face circuits.  Your hFe meter doesn't give leakage...and may just roll all over the place, rendering a legible reading unlikely.  That's why R.G. wrote his original piece on testing Ge's - there was a need.


Haha. Thanks for clearing that up for me. I had no idea. Good thing an expert like you frequents the board to educate. :icon_lol:

The OP himself knows his question has been beat to death. The best answer here is to give the OP a link to the search button.
« Last Edit: May 29, 2014, 07:32:52 PM by Arcane Analog »

Psychophonic

I am trying to test a pair of silicon BC108C NPN transistors for a fuzz face project.

Has anyone posted a YouTube demo or tutorial of the geofex/RG transistor checker in action? If so, I cannot find one. As a noob, I canít seem to make sense of the testing and formulas used in the Geofex document. I have set up the test circuit on a breadboard but I canít get anywhere from there. For me, the use of the sample readings in the document only add to my confusion. Iíd just like to watch someoneís step by step process so I can follow along.

mozz

I'd just pick up a component tester from eBay. He may want to check germanium in the future. My older mk328 is fine and also checks germanium. Some of the newer models may show you gain but do not measure leakage. The RG tester may work fine but it's a pain in the ass and would recommend a DCA55, save up some $$ for it.
  • SUPPORTER

antonis

As a noob, I canít seem to make sense of the testing and formulas used in the Geofex document.

Excuse me but I don't see any "formulas" other than Ohm's Law..
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

Rob Strand

I wouldn't go out and buy a new multimeter because it has hFE measurement.    The hFE test circuits inside of multimeters are fairly crude they are only good for ball-park measurements.   RG's circuit is easy to build and at least you know what it is doing.
Send:     . .- .-. - .... / - --- / --. --- .-. -

pee-j

Yes, but many don't....mine doesn't, altho it's a very useful Extech.   I rarely "need to know" (data sheets are enough for 99.9%), but when I do, I use the R.G./ Seljer method described.   
1M on base and a 1K on the collector, easy as pie.    Beta = base voltage/collector voltage.

Beta = base voltage/collector voltage * 1000  (just a nuance)

.................
PS:
my cheap DMM says 1000 for hfE of a BC550c, while the universal part tester "LCR-T4"  says 580....
so, I really needed to have a third measurement, which says: 600

I'm sure that many FF experiments go stranded cause measurement using a cheap DMM slot -- in unlucky cases -- gives false result
this is why I was really looking for this method :) so thanks!!!
« Last Edit: June 25, 2022, 04:22:10 PM by pee-j »

antonis

Let's shake hands for Beta = (Base voltage / Collector voltage) X (Base resistor / Collector resistor)..

(or else, we have to deal with Alpha which, IMHO, is more appropriate parameter when considering a BJT as current controled current source..)
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

Rob Strand

Quote
my cheap DMM says 1000 for hfE of a BC550c, while the universal part tester "LCR-T4"  says 580....
so, I really needed to have a third measurement, which says: 600
Just use RG's set-up.  There's no mystery about how it works.   You just have to build it right.

(If you want to split hairs, you need to reduce the base resistor or increase the collector resistor a tad to account for the silicon transistor VBE drop.  Vbat=9V, Rb=2.2M and Rc=2.635k, then hFE = 100*Vmeasured;  OK upto about hFE=800.  Vbat =9V is not a really valid a assumption for a battery but it's good enough. FWIW, for technical reasons on how hFE is defined the fact VCE isn't constant means about +/- 3% to 4% error in hFE.)
« Last Edit: June 25, 2022, 09:13:08 PM by Rob Strand »
Send:     . .- .-. - .... / - --- / --. --- .-. -

Rob Strand

Quote
Let's shake hands for Beta = (Base voltage / Collector voltage) X (Base resistor / Collector resistor)..

Beta = Collector Current / Base Current

Collector Current = Collector Voltage / Collector Resistor

but

Base Current  = (Supply Voltage - Transistor Base Emitter Voltage) / Base Resistor

So when you write an equation with the same form as those posted you get,

Beta  =   (Collector Voltage / Collector Resistor) * (Base Resistor / (Supply Voltage - Transistor Base Emitter Voltage))

Beta  = (Collector Voltage / (Supply Voltage - Transistor Base Emitter Voltage)) * (Base Resistor / Collector Resistor)

The difference is the denominator of the first term "Base  Voltage" vs "Supply Voltage - Transistor Base Emitter Voltage"


Picking transistors for FF Clones
http://www.geofex.com/Article_Folders/fuzzface/fftech.htm

The way I look at the circuit is to start with a ball-park collector current.  A good test current is 1mA to 2mA, as that's what is often used in datasheets.   A ball-park middle transistor gain is beta  = 300, so the ball-park base current is  5uA.

After that I don't try to write out a whole equation.   Go a step at a time.  That lets you keep your hands
on the wheel about what is going on.

1) Work out what resistor is required for a base current around 5uA.

*** The whole basis of the circuit is to operate at a constant base current. ***

Base Resistor  = (Supply Voltage - Transistor Base Emitter Voltage) / Base Current
                    = (9 - 0.65) / 5e-6
                    = 1.67M

VBE of 0.65V isn't a bad estimate for 1mA to 2mA collector current.  At a squeeze you could argue for 0.62V.
The accuracy of the circuit just isn't there for splitting hairs and different transistors will have different
VBE's even at 1mA.   The circuit doesn't keep the collector at a constant current anyway and that causes
VBE to move with different hFE (assuming the same transistor type).

RG's choice of 2M2 is perfectly acceptable.

2) Once the base resistor is chosen we can work out the actual base current,

Base Current  = (Supply Voltage - Transistor Base Emitter Voltage) / Base Resistor
                   = (9 - 0.65) / 2.2M
                   = 3.7955uA

3)  We want hFE = 100 * collector voltage.
     For hFE = 300 we need 3V out.   We just need to pick a point so we have some concrete numbers.

    With the base current in (2) the collector current is,

    Collector Current = 300 * 3.7955uA = 1.1386mA

4) Now we choose the collector resistor to have 3V drop with this collector current.

    Collector Resistor = 3 / 1.1386mA = 2634.7 ohm

   [If you want to be precise, that's the require final value including the multimeter loading.
    For a 1M ohm input impedance multimeter, the actual resistor value is Rc = 1/(1/2634.7 - 1/1M) = 2341.7 ohm
    For a 10M ohm input impedance multimeter, the actual resistor value is Rc = 1/(1/2634.7 - 1/10M) = 2635.4 ohm

   Don't forget tolerances on the 2M2  :icon_mrgreen:]

5) If the collector voltage rises much above 8V the transistor's VCE will be too low to guarantee
    the transistor isn't saturating.   Since hFE = 100 * collector voltage = 100 * 8V = 800.
    The circuit will measure a maximum hFE of 800.

« Last Edit: June 26, 2022, 01:09:46 AM by Rob Strand »
Send:     . .- .-. - .... / - --- / --. --- .-. -

antonis

I was trying to make a definition between Sir Mike's  & pee-j's posts.. :icon_wink:

To make it more plain, let's replace Base & Collector "voltages" by their respective voltage drops..

So, Beta = [(Vcc - VCollector) / (Vcc - VBase)] X (RBase / RCollector)
(voltage drop across Collector resistor divided by voltage drop across Base resistor times their inverse ratio..)

P.S.1
Obviously, DMM's impedance should be taken into account, especially for high value Base resistors so RBase should be replaced by their Thevenin equivalent..
(unless, of course, their ratio is high enough - say, >10..)

P.S.2
As beta variability is pertained to almost everything (even to the phase of the moon), it should be wise to take measurements as close as possible to particular circuit working conditions..
« Last Edit: June 26, 2022, 06:25:55 AM by antonis »
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..