How to tell if a transistor is NPN or PNP

[zoobyshoe]

Today I learned how to tell if a transistor is NPN or PNP:

 

[Hornbein]

Today I learned how to tell if a transistor is NPN or PNP:

But the method for determining the sex of chicks remains secret.

 

[jim hardy]

Today I learned how to tell if a transistor is NPN or PNP:

be sure you know for certain, by actual test, that the red lead indeed applies positive when ohms is selected.

In analog world,
for US made meters red lead is positive for measuring voltage and it applies positive for measuring ohms.
But oriental made meters swap polarity when selected to ohms and apply negative via red lead. At least every one I've ever owned and that's a lot of them.

In digital world-
The only digital meters i own are US made, a Fluke and a Beckman.
When measuring ohms both apply positive via red lead.

So i can neither challenge nor confirm that fellow's video.
I can say for sure
were he using one of my old Japanese analog meters,
he'd have got the wrong polarity for his transistors.
Ditto for one of my newer Chinese analog meters.

So - check your meter before making that test.

old jim

 

[zoobyshoe]

be sure you know for certain, by actual test, that the red lead indeed applies positive when ohms is selected.

The setting he says to use is "diode". As far as I know, that's a setting only available on digital meters. At least, the old Radio Shack analog meter I have has no such setting.

In addition to that meter, I do also have a digital meter with the "diode" setting, and that's what I used to test some unknown transistors I'd pulled off some random board.

Anyway, it seems you're saying they can also be tested by setting the meter to test resistance (provided you're sure you know which lead applies positive voltage)?

 

[collinsmark]

Today I learned how to tell if a transistor is NPN or PNP:

Also, the author of the video seems to make the potentially grave assumption that the base of the transistor is the middle lead. I can assure you with certainty that this is not always the case.

http://www.vetco.net/blog/wp-content/uploads/2013/04/TransistorPinouts.png

What that means is that you'll just have to test a few more combinations of leads to find the particular leads that have the diode-like properties; just not assuming from the beginning what any particular lead happens to be.

The author of the video is correct though that if your multi-meter has a "diode" function, it might make the process a little easier. But you could use the Ohm reading on most multi-meters. But like @jim hardy brings up, if you use the Ohms reading, you might want to verify which multimeter input is which.

Once you determine the base lead and find out if it's an NPN or PNP, finding out which is the emitter and which is collector is a different story. It's difficult to do with with 100% reliability without testing it in a real circuit. But usually the connection with the higher [forward] voltage drop (in reference to the base -- comparing the two diode candidates) corresponds to the emitter.

This link probably explains it better than I can.
http://www.allaboutcircuits.com/textbook/semiconductors/chpt-4/meter-check-transistor-bjt/
[Edit: Caution: the author of the link I provided above uses arrows to show the direction of the "electron flow" rather than the direction of the "current" (which is opposite). 'Just something to keep in mind when reading the link.]

 

[zoobyshoe]

Also, the author of the video seems to make the potentially grave assumption that the base of the transistor is the middle lead. I can assure you with certainty that this is not always the case.

Good to know, and it's the first time I've ever heard this. Everyone seems to say the middle lead is the base. But this may explain why one of the transistors I tested seemed to be "broken." Unfortunately, I tossed it out and can't recheck it.

But you could use the Ohm reading on most multi-meters. But like @jim hardy brings up, if you use the Ohms reading, you might want to verify which multimeter input is which.

Also very good to know, and it means I can use either meter. What's a convenient way of testing which lead is positive?

Once you determine the base lead and find out if it's an NPN or PNP, finding out which is the emitter and which is collector is a different story. It's difficult to do with with 100% reliability without testing it in a real circuit. But usually the connection with the higher voltage drop (in reference to the base -- comparing the two diode candidates) corresponds to the emitter.

Good, thanks!

 

[jim hardy]

The setting he says to use is "diode". As far as I know, that's a setting only available on digital meters. At least, the old Radio Shack analog meter I have has no such setting.

In addition to that meter, I do also have a digital meter with the "diode" setting, and that's what I used to test some unknown transistors I'd pulled off some random board.

Anyway, it seems you're saying they can also be tested by setting the meter to test resistance (provided you're sure you know which lead applies positive voltage)?

Yes they can be tested on an ohms setting.

On better analog meters that have a RX1 scale, one should use the RX10 setting NOT the RX1 scale.
That's because analog meters can apply substantial current when selected to RX1, like ~150 milliamps, which can wreck a small signal transistor.

A good junction will give about 2/3 scale deflection
a shorted junction will of course give full scale deflection , zero ohms
a Darlington transistor's E-B will give about 1/3 scale deflection because it's two junctions in series.

One should avoid the highest resistance scale, RX10k on the Simpson pictured, because most analog meters when selected to that scale apply in excess of 6 volts which exceeds the reverse E-B limit for most transistors.

We've had threads about how analog ohm-meters work in EE forum.
One can extract a lot more information fromhis meter if he understands its inner workings.
if there's interest , re-open or start another one ?old jim

 

[jim hardy]

Once you determine the base lead and find out if it's an NPN or PNP, finding out which is the emitter and which is collector is a different story. It's difficult to do with with 100% reliability without testing it in a real circuit. But usually the connection with the higher [forward] voltage drop (in reference to the base -- comparing the two diode candidates) corresponds to the emitter.

AMEN ! That is a real challenge.

Here's what i do-
take advantage of the fact that a transistor will operate with E-C swapped but not very well, it has at lower gain than with E-C connected properly.
So - i think of the mystery transistor's three leads as Base, Unknowns U1 and U2

I hook one ohm-meter lead to Base, other meter lead to transistor's U1 with polarity so B-U1 junction doesn't conduct. That means junction is reverse biased as a B-C junction should be,.
Next I moisten a fingertip and touch it to B-U1 leads. That injects a small current into B through my fingertip. I observe how much the meter needle rises.
Then i move meter lead from U1 to U2 and repeat moist finger check, and note which gives bigger needle rise - U1 or U2. Repeat until I'm sure.
Knowing that the transistor works better when correctly biased, i surmise that whichever U gives bigger needle rise is the collector.
Hasn't missed yet for me though sometimes i have to use RX100 scale...

old jim

 

[zoobyshoe]

Yes they can be tested on an ohms setting.

On better analog meters that have a RX1 scale, one should use the RX10 setting NOT the RX1 scale.
That's because analog meters can apply substantial current when selected to RX1, like ~150 milliamps, which can wreck a small signal transistor.

A good junction will give about 2/3 scale deflection
a shorted junction will of course give full scale deflection , zero ohms
a Darlington transistor's E-B will give about 1/3 scale deflection because it's two junctions in series.

One should avoid the highest resistance scale, RX10k on the Simpson pictured, because most analog meters when selected to that scale apply in excess of 6 volts which exceeds the reverse E-B limit for most transistors.
View attachment 93199

We've had threads about how analog ohm-meters work in EE forum.
One can extract a lot more information fromhis meter if he understands its inner workings.
if there's interest , re-open or start another one ?old jim

Thanks, Jim! I think I'll try to remember to stick to the digital meter diode setting, since I'm likely to forget which ohms order of magnitude won't damage the transistor.

 

[zoobyshoe]

AMEN ! That is a real challenge.

Here's what i do-
take advantage of the fact that a transistor will operate with E-C swapped but not very well, it has at lower gain than with E-C connected properly.
So - i think of the mystery transistor's three leads as Base, Unknowns U1 and U2

I hook one ohm-meter lead to Base, other meter lead to transistor's U1 with polarity so B-U1 junction doesn't conduct. That means junction is reverse biased as a B-C junction should be,.
Next I moisten a fingertip and touch it to B-U1 leads. That injects a small current into B through my fingertip. I observe how much the meter needle rises.
Then i move meter lead from U1 to U2 and repeat moist finger check, and note which gives bigger needle rise - U1 or U2. Repeat until I'm sure.
Knowing that the transistor works better when correctly biased, i surmise that whichever U gives bigger needle rise is the collector.
Hasn't missed yet for me though sometimes i have to use RX100 scale...

old jim

It strikes me as odd they never adopted some conventional markings to identify whether it was PNP or NPN, or to identify the base, collector, and emitter.

 

[OmCheeto]

It strikes me as odd they never adopted some conventional markings to identify whether it was PNP or NPN, or to identify the base, collector, and emitter.

"Conventional"? In the EE field? Ah! Hahahahaha!

http://web.engr.oregonstate.edu/~traylor/ece112/lectures/elect_flow_vs_conv_I.pdf
In 1752, prior to electricity being identified with the electron, Ben Franklin chose a convention regarding the direction of current flow. Franklin assumed that electrons (being assumed positive) flow from positive to negative terminals. We now know this is incorrect.​

And there are still arguments about it.

 

[collinsmark]

It strikes me as odd they never adopted some conventional markings to identify whether it was PNP or NPN, or to identify the base, collector, and emitter.

Perhaps some insight might come from the fact that these same packages are also used for things that are not even similar to BJTs. And not just other transistor technologies like J-FETs and MOSFETS, but also complete integrated circuits. Voltage regulators come to mind; they also have three leads and use the same package types as transistors. Modern day transistors are really, really tiny, and on an integrated circuit you could easily fit thousands (even tens of thousands) of them in that tiny package, as long as the complete circuit itself only required three external leads. So the packaging is used for more things than just BJT transistors.

But even when using them for BJT transistors, based on your desired circuit board layout, you might find one pinout more advantageous than another in terms of routing on the board (one pinout might make your circuit topology more compact than another). So having different pinout options provides flexibility to the circuit designer.

 

[collinsmark]

"Conventional"? In the EE field? Ah! Hahahahaha!

http://web.engr.oregonstate.edu/~traylor/ece112/lectures/elect_flow_vs_conv_I.pdf
In 1752, prior to electricity being identified with the electron, Ben Franklin chose a convention regarding the direction of current flow. Franklin assumed that electrons (being assumed positive) flow from positive to negative terminals. We now know this is incorrect.​

And there are still arguments about it.

Yes, that makes a great trivia question. And, by extension, makes a great post in the spirit of this particular thread.

• Q: Which famous scientist is credited for creating the convention that ultimately results in electrons having negative charge and protons positive?
  
• A: Benjamin Franklin

Most trivia players who are not familiar with physics or electricity might guess Einstein or Newton or some-such, not realizing the actual scientist was one of the United States' founding fathers.

Franklin's choice does confuse students. I agree there. Students initially want to imagine current as something tangible moving around. But Franklin's convention is not really wrong. Separating electron flow from a more abstract idea of "current" sometimes has conceptual advantages, which students learn later down the line.

In his defense though, Franklin's convention came about before the discovery of the electron. He was rubbing glass on fur and such, and had to pick something to be positive and something else to be negative, either way, so he just picked one. Either choice would have worked fine. We could switch conventions today, making the electron positive and the proton negative and all of physics would work just the same, as long as we maintain consistency.

 

[Hornbein]

Franklin was the first to realize that positive and negative charges canceled one another out. They weren't two entirely separate phenomena.

He also used to route lightning strikes through the interior of his home.

 

[jim hardy]

He also used to route lightning strikes through the interior of his home.

He must have been a really fun neighbor.

Franklin's choice does confuse students. I agree there. Students initially want to imagine current as something tangible moving around. But Franklin's convention is not really wrong. Separating electron flow from a more abstract idea of "current" sometimes has conceptual advantages, which students learn later down the line.

In vacuum tubes electrons actually do leave the cathode and arrive at the anode , aka plate.
My 1958 GE Transistor Manual introduces transistors by using parallel to a triode tube

from page 7

Here the emitter emits electrons, the collector collects electrons and the base controls the flow of electrons by controlling the charge concentration in the base region, so in the broadest sense, the function of the three elements in the triode tube and the transistor are similar..."

My high school textbook was the 1962 edition of the GE transistor manual. Found it online here
http://babel.hathitrust.org/cgi/pt?id=mdp.49015000619008;view=1up;seq=24
Observe they speak of electrons flowing... page 10

Since textbooks and indeed the GE manual used conventional current for all formulas we boys became ambidextrous , able to work with either current convention. I won't say ithat was painless.
But we did our basic thinking with the electrons for charge carriers because they're the particles that actually move. That was common in 1960's.

So if one wanted to start a row,
he might say something like 'Slide rules and electron current got us to the moon. '
old jim

 

[dlgoff]

... 'Slide rules and electron current got us to ...

Good possibility I wouldn't be a member here today if this wasn't part of my upbringing.

 

[Hornbein]

"Conventional"? In the EE field? Ah! Hahahahaha!

http://web.engr.oregonstate.edu/~traylor/ece112/lectures/elect_flow_vs_conv_I.pdf
In 1752, prior to electricity being identified with the electron, Ben Franklin chose a convention regarding the direction of current flow. Franklin assumed that electrons (being assumed positive) flow from positive to negative terminals. We now know this is incorrect.​

And there are still arguments about it.

I'll be pedantic. Franklin had no concept of electrons. I think everyone thought of some undetectable liquid. I don't know that he assumed it flowed from positive to negative. The point was that the charges were opposite and canceled, as do positive and negative numbers.

 

[mfb]

Also very good to know, and it means I can use either meter. What's a convenient way of testing which lead is positive?

Use a diode or a known transistor as reference, or use a second multimeter.

To make things worse, some multimeters use very low voltages to test resistances, if the voltage is too low it might read high resistances in both cases.

I think we should make a separate thread out of that discussion as it is beyond the scope of "today I learned".

 

[zoobyshoe]

"Conventional"? In the EE field? Ah! Hahahahaha!

Despite "conventional current," think of all the things in electricity and electronics that are successfully identified by conventional markings. The first one that comes to mind is resistors. Next is batteries. The positive and negative terminals are physically different in shape on most batteries, and if they're not (as with a car battery) they're clearly marked + or -. Diodes are clearly marked. LED's have the positive lead physically longer than the negative. Electrolytic capacitors have the positive and negative sides clearly marked. Speakers, microphones, and motors nearly always have a red wire soldered to the positive terminal. You don't have to get a meter out and laboriously measure any of these components: they're marked one way or another.

Perhaps some insight might come from the fact that these same packages are also used for things that are not even similar to BJTs.

There are many unconventional batteries, phone batteries for example, but that doesn't make it impossible to mark conventional batteries.

I don't see any good reason most transistors couldn't be clearly stamped PNP or NPN, and why the base, at least, couldn't always be identified with something as simple as a notch. And if they could identify the base that way (which they could) they could also identify the collector or emitter with some conventional symbol or mark.

 

[zoobyshoe]

Use a diode or a known transistor as reference, or use a second multimeter.

Of these three, I think diode wins.

To make things worse, some multimeters use very low voltages to test resistances, if the voltage is too low it might read high resistances in both cases.

I don't follow.

I think we should make a separate thread out of that discussion as it is beyond the scope of "today I learned".

This thread survived a 5 page excursion into a peculiar property of factorials unscathed. Things will normalize. Don't worry.

 

[mfb]

I don't follow.

If you apply a voltage of 0.2 V (multimeter in an unfortunate resistance mode), it will be hard to see conductance in any direction.

 

[zoobyshoe]

If you apply a voltage of 0.2 V (multimeter in an unfortunate resistance mode), it will be hard to see conductance in any direction.

Oh, you mean because the needle will hardly deflect.

 

[mfb]

More thinking about digital multimeters here, but the measured resistance value will be high if you are below the threshold voltage.

 

[zoobyshoe]

More thinking about digital multimeters here, but the measured resistance value will be high if you are below the threshold voltage.

The digital meter should have the "diode" setting, so you can use that, as per the video.

 

[Hornbein]

The digital meter should have the "diode" setting, so you can use that, as per the video.

I think he's saying that the voltage must be at least 0.7v, otherwise current will not flow in either direction.