Transistors

[Refugee]

When I was a student in the 1970s I took on the challenge of creating a graph like the ones in data sheets.
Collector current was set by switched resistors and a regulated power supply.
VCE was set by a metal 741 (remember those 10 pin bugs) at 5 volts.
Base current was measured when the 741 circuit showed that the collector current was correct that being set by pots and switching and what we now know of as dimwit LEDs.
The thing still works but the resistors have all drifted. This has been mitigated by using one of those Farnell power supplies with a current control and meter.
I got my graph and was able to go through all those bags of random transistors back in the day.
These days it is audiofiles and they will have to pay a premium for me to haul it out to match there output transistors for them all the way up to two amps with the external power supply.

[Nowhere-Man]

Rained off today. I was fitting a shower tray but it's too wet to be clambering on and off a boat.
Suffice it to say, my transistor knowledge is pretty basic. ARRL (as Mark commented) is a bit limited on transistors. They tended to be used as 100 KC calibrators and diodes were gradually replacing valve rectifiers.
I think with transistors it became inevitable circuits would get more complicated and the lifespans of the radios shorter.
On Sunday a KB radio was delivered to me. It may be late 40's. A casual glance indicates there's not a great deal inside. Looks to me like a very enjoyable future project.
The simplest transistor radio I have is a cream, suitcase type Bush. No PCB and chassis mounted.
The Soviet VEF circuit wise is still over my head but then again it's more an old radio than vintage. Later maybe we can look at the circuit to try and deduce the stages.

[jimmc101]

I think Alpha was of more use in connection with point contact transistors which had a common base current gain (= Alpha) significantly greater than unity. (See here for example).
This would make the use of Beta more than a little confusing.
Once junction transistors came on the scene Alpha became (just) less than unity and Beta became the preferred measure of gain.

Jim

[ppppenguin]

What a brilliant article written in the early days of transistors. I didn't realise that 70MHz was possible in 1952. Transistors only got better from there.

Some of his remarks about valves are not correct. As Tommy Flowers knew back in 1940, valves can have a very long life so long as they are powered continuously and used comfortably within their ratings. This fact was used in Colossus, in telephone exchange equipment and from 1956 in undersea telephone cables. In this last application where reliability was paramount the designers used a fairly old fashioned type of valve as its long term reliability was better understood than later valves or early transistors.

[Nowhere-Man]

I didn't set out to overkill as such. No plans to design circuits. I just wanted to visualise how current is amplified to then influence voltage and wattage. In fact, it's really pretty simple. 120 microamps base current times load equals a very small voltage. Amplify that current by 50 electrons to one and you're in business.
Laying out my boat sink over the weekend, I had to make a voltage divider. That is, split the pump flow into two. I needed one outlet for my tap and another for the shower. I thought how similar it was to HT current.
I would like to get better at transistor circuits but there is much to learn. I was better grounded in valves as Chas Miller's book got me to a good start.

[ppppenguin]

One of the most important things to understand about transistors is that they are voltage amplifiers, not current amplifiers. Yes a change in base current will give a bigger change in collector current but base current is actually a nuisance, something that we have to put up with. Designers have to make their circuits work despite base current which will vary a lot with teperature and between different devices.

Many people have written that transistors are current amplifiers. They are wrong. Any article that claims this should be taken with a large pinch of salt as the author clearly doesn't understand how transistors work.

[jimmc101]

One of the most important things to understand about transistors is that they are voltage amplifiers, not current amplifiers. Yes a change in base current will give a bigger change in collector current but base current is actually a nuisance, something that we have to put up with. Designers have to make their circuits work despite base current which will vary a lot with teperature and between different devices.

Many people have written that transistors are current amplifiers. They are wrong. Any article that claims this should be taken with a large pinch of salt as the author clearly doesn't understand how transistors work.

I'm sorry but I have to disagree, the relationship between collector current and base current is reasonably linear over quite a wide range.

On the other hand the relationship between base voltage and collector voltage is not...
The base current is exponentially related to the base voltage (wrt the emitter) and for a linear collector load (e.g. a resistor) the collector voltage will be a linear function of the collector current.
Thus the voltage gain of a transistor is an exponential function.
( Without a collector load the voltage gain would depend on the transistors output resistance and feedback voltage (Early effect).)

I would agree that a FET may be regarded as a voltage amplifier in similar fashion to a valve but not a Bipolar transistor.

Jim

[Nowhere-Man]

They are not free electrons. As desribed in a scene of the 50's Attack Of The Crab Monster, "the electrons jump from atom to atom". Maybe too as in ARRL you could point out the collector base junction of a PNP is huge resistance so the watts are high.

[ppppenguin]

I'm sorry but I have to disagree, the relationship between collector current and base current is reasonably linear over quite a wide range.

And the ratio is very uncertain.Varies widely with temperature and with different devices. I can't imagine how you would design an amplifier or anything else by relying on the current gain.

Whereas the voltage gain is utterly predictable, varies only slightly with temperature and can be approximated as linear for small signals.

[Craig]

They are not free electrons. As desribed in a scene of the 50's Attack Of The Crab Monster, "the electrons jump from atom to atom". 

Nonsense. Electrons (and holes) in semiconductor lattices behave according to Fermi-Dirac statistics. Statistics because there is a large enough density of them. Indeed the conduction process in doped p-n junctions is completely understood in terms of where the Fermi level is in relation to the dopant energy states near the conduction and valence band edges.

Take it further to superconductors, where paired electrons (cooper pairs) condense at low temperatures way below the Fermi energy. They pair over distances of hundreds of nanometers (so thousands of atomic diameters) and move with zero electrical resistance. Any concept of electrons jumping rather than smoothly flowing is clearly wrong here too.

The only process that could conceivably considered as "jumping" is electron tunnelling. If an energy barrier is higher than the electron energy, and also thin enough, electrons can tunnel though because of their quantum mechanical nature. That is the principle behind the Tunnel Diode. But really even there there is no jumping as such.