Transistor confusion

[sweatybetty]

Hello all!
Im hoping someone can help me understand a pretty basic thing about transistors. My confusion lies here.
When testing a transistor (basic BJT NPN) you test it like you would a diode     with current flowing from base to emitter or base to collector. So, how can current flow from the collector to the emitter as shown here?    
Maybe I am just over-thinking this?

Thank you all for your patience

[TimFox]

I never look at these videos.
For a BJT, there is a diode between base and emitter, and another between base and collector.
The I-V curve for each is that of a silicon diode.
With the base open-circuit, you should get almost no current between collector and emitter at reasonable voltages.
Be careful not to apply more than, say, 5 V to the base-emitter diode, since it has a much lower reverse breakdown voltage than the base-collector diode.
The circuit in the still is a normal transistor (not diodes) circuit, where positive (forward-bias) voltage applied to the base-emitter  causes current to flow from collector to emitter  and less current to flow from base to emitter.

[Benta]

Love how the battery is upside down in the 2nd video image. That guy's a real expert!
Never watch that stuff either. Weeding out the 99% junk to find the 1% nuggets is just too much effort.

[IanB]

Love how the battery is upside down in the 2nd video image. That guy's a real expert!

What's wrong with that, actually? In many respects a PNP transistor is the mirror of an NPN transistor, with voltages and currents reversed. So showing the diagram mirrored (inverted) makes sense for intuitive understanding.

[Benta]

Geddit?

[IanB]

Geddit?

Yes. Although I looked at the + and − annotations and ignored the symbol.

[rsjsouza]

with current flowing from base to emitter or base to collector. So, how can current flow from the collector to the emitter as shown here?   

Although non-intuitive at first glance, the physical construction of the  doped silicon "sandwich" makes a difference: the base is usually much thinner than the other two regions. Overly simplifying things, when a direct bias between Emitter and Base is applied, the potential barrier thickens and, at a certain voltage, it overflows to the collector region. This allows charges to travel from emitter to collector that would otherwise be repelled by the reverse bias between the base and collector.

[ledtester]

For good visuals try first watching this video:

Transistors, How do they work? -- Lesics
https://youtu.be/7ukDKVHnac4

and then watch Ben Eater's video:

How a transistor works -- Ben Eater
https://youtu.be/DXvAlwMAxiA

I think the first video does a good job of explaining doping and how diodes work but Ben's video spends more time explaining transistor operation.

[golden_labels]

with current flowing from base to emitter or base to collector. So, how can current flow from the collector to the emitter as shown here?

The answer is: because a transistor is not two diodes connected back to back.

It’s a single block of material, areas of which have different properties. When current is present in some of those areas, it affects the entire device in a way that makes some other areas become more conductive.

Since those areas are n-doped or p-doped, across their boundary they form p-n junctions. Semiconductor diodes are nothing more than p-n junctions, so you can use the diode meter of a multimeter to perform some rudimentary checks.

But the detailed explaination of how BJTs work is much, much more complicated⁽¹⁾ and involves quantum physics. And is not necessary to use transistors in lumped-element model, which is what the second video does. Also note that while that second video is technically correct, ignoring that battery symbol mistake, it does not explain transistor’s operation at quantum level, at the conceptual level, at lumped element model level; neither it does provide explanation of why would anybody use a transistor.

⁽¹⁾ Note that Wikipedia provides an extremely simplified view of the idea, not an actual construction. No commercial transistor ever was produced by simply sandwitching layers. Only simple and old designs use this layout; even that layout is already out of proportions regarding sizes and misses three-dimensionality. The actual, modern transistors are much more complicated: see the attachment (from Phillip Alen’s materials, ECE 4430 Analog Integrated Circuits (fall 2004)).

[MikeK]

While it can be said that there is a diode between the base and emitter and between the base and collect...It is *not exactly* like that.  Two discrete diodes aren't sharing the middle P region (in an NPN).  An NPN transistor has a P region sandwiched between two N regions.  So, it's *like* two diodes in that you can test each separately.  But the shared P region is what allows current to flow from collector to emitter.  The P region is controlling the current from N to N.  You can't do that with two discrete diodes.

[Doctorandus_P]

As said (written) before, you can measure the individual pn junctions as diodes in a BJT, but it is only a part of the story.
In a BJT, the diodes share a layer, which is the base of the transistor, (So it only has 3 layers and not 4 as two separate diodes would have).

The base is also a very thin layer, and when current flows through the BE junction, a lot of the electrons "shoot through" the base because it's so thin, and they end up in the collector. The proportion of electrons that end up in the collector depends on the transistor geometry, and is a somewhat constant value called the Hfe of the transistor. Small transistors can have a Hfe of a few hundred, wile big power transisors can have an Hfe as low as 10.

Measuring both junctions of an transistor with a diode tester is a quick and convenient way to check if a transistor still works. If you can't measure the diodes, then the transistor is broken, while if you can measure the diodes, then there is a quite high probablility that the transistor works.

This explanation is a gross oversimplification, but it's a useful explanation. Such simplifications are quite common in engineering. Quite often the details of how something works is quite complicated and not very relevant for practical use. And then a simplified model is made, that is just sophisticated enough to "get the job done". And when more details are needed, the model is refined. For repairing electronics, measuring the diodes of a BJT is often enough, while this does not even explain the working of a transistor. If you want to understand the working of an electronic circuit, then the first refinement to the "two diodes model" is to add the current amplification factor (Hfe) to the model.

With a simple model you can get quickly to a result that is "good enough" for the task at hand. But you should always be aware that the model you are using is simplified, and those simplifications break at some point, and then you have to add more sophistication to your model.

A further refinement of the model would be to acknowledge that hfe is not a constant, but it varies a bit with Collector - Emitter voltage (the "Early Effect"). This effect is one of the many factors that for example cause distortion when you use BJT's to make an audio amplifier.

[sweatybetty]

Thank you all. Things are quite a bit clearer now. 

[Zero999]

I never look at these videos.
For a BJT, there is a diode between base and emitter, and another between base and collector.
The I-V curve for each is that of a silicon diode.

I haven't watched the video, so I don't know if this is mentioned. An easy way to identify the emitter and collector is the voltage drop of the base-emitter junction will be higher, than the base-collector junction, given the same current, because it's physically smaller.

[Doctorandus_P]

An easy way to identify the emitter and collector is the voltage drop of the base-emitter junction will be higher, than the base-collector junction, given the same current, because it's physically smaller.

Add to that you also have to measure them at the same temperature.
Diodes have a temperature coefficient of about -2mV per degree celcius and this can throw off your measurements if the transistor is still cooling down after desoldering for measurment, or when it's heated up by your fingers.

Also noteworthy to mention is that the PN junctions of BJT's can often be measured in circuit. "Other stuff" on the PCB can only be parallel to the PN junctions and make it measure lower. So if you're measuring in circuit and you measure more then around 700mV for diode measurement, It's quite likely the transistor is broken. If you're measuring below 500mV then some low ohmic resistor may be in parallel, to pull your diode measurement off. Measurements below 100mV or so indicate a likelihood of a short.  If you're measuring between 550mV and 650mV then it's likely you are measuring the PN junction. It's a very coarse method, but it's valuable because it's so simple to do. But when in doubt, you have to desolder the BJT and measure it out of the circuit. This method also has some other limitations. For example it does not work if multiple transistors are put in parallel, which is often done in audio power amplifiers or (less common) in low noise circuits (With a lot of transistors, the noise averages out and the total noise is less than with a single transistor).

[MrAl]

Hello all!
Im hoping someone can help me understand a pretty basic thing about transistors. My confusion lies here.
When testing a transistor (basic BJT NPN) you test it like you would a diode     with current flowing from base to emitter or base to collector. So, how can current flow from the collector to the emitter as shown here?    
Maybe I am just over-thinking this?

Thank you all for your patience

Hi,

The short answer is that is how transistors work and that is why they work.
With a small current through the base emitter diode a larger current can flow through the collector to emitter.  In fact, the ratio of collector current to base current as a maximum is called the Beta of the transistor.  It's an important measure of the quality of the transistor.

To understand this easily you can imagine the transistor is a current controlled current source.  The control current is the base to emitter current, and the output current source is the collector to emitter current.  If you dont know what a current controlled current source (CCCS) is you can look that up on the web and that will make things very clear.  The other thing you have to understand though is that the Beta (current gain) is not constant but can change depending on various other things like collector to emitter voltage, but you can ignore that at first just to get a basic grip on how bipolar transistors work.

[EPAIII]

Another point to remember is the common test with a VOM is only one way to check a transistor. There are many other ways. A simple circuit with a pot for the base resistor and a meter or LED for the collector load can show the actual amplification when the pot is rotated and the meter or LED shows more or less current.

And there are curve plotters that attach to oscilloscopes.

Of course, transistors have multiple specs and different tests are used to measure them.

I have tested thousands of questionable transistors with the VOM/resistance readings. I always read all three pairs of leads: B-E, B-C, and C-E. There have been times when the C-E readings showed a short when the B-E and B-C showed diode characteristics. And there were other infrequent times when all three readings were seemingly OK, but the transistor was out of spec. and a new one repaired the problem.

My point is you must add judgement and sometimes additional tests to the VOM readings. Good VOM readings do not guarantee a good transistor.

[MrAl]

The last post reminded me i forgot to mention something about multimeters, thanks to that poster.

Some multimeters have sockets right on the front where you can plug a small transistor in to test it.  It's a basic test to show if it works or not, and the Beta shows up in the meter reading.  It's the Beta for the particular current the meter uses so it's just a general basic test, but if the transistor is really burnt it will show that something isnt right.

In reality there are a lot of tests for a transistor but usually if you do a simple Beta test you can assume, for a time, that it is ok.
If you really want to get into it then you have to test for switching time and stuff like that.  You need better equipment for that though.