||[Jan. 15th, 2013|06:30 pm]
In other hobbies-I-can-get-away-with-pursuing-for-hours-at-a-time-since-I-am-unemployed news,
here's a trivial transistor circuit I finally got working just to convince myself (a) I understand transistors correctly, and (b) I can understand the !@#$ing relationship between abstract transistor terminals and where they actually are in the real world correctly. (Answer: I can only do so after like six tries, apparently. I keep thinking Base should be between Collector and Emitter on the package, and I constantly confuse Collector with Emitter) Stupidly basic thing to accomplish, but still feels fun. I can verify that the PNP bipolar transistor does in fact conduct from emitter to collector when I pull the base low, and doesn't when I pull it high.
Also been playing around with a little DC motor. I have a stepper motor in my possession, but I think I need to hook up some more parts before I can drive it correctly.
Transistor packages are infuriating. The base is where ever the heck the person who made the transistor damn well pleases. Various transistors with the same package will not be consistent about where they put it.
At least this TIP42C that I got my hands on seems to conform to its datasheet at least.
Hey, stupid question while I'm at it: what in the manufacture of a BJT actually distinguishes collector from emitter? Regardless of whether it's PNP or NPN, they're made of the same material (as one another, that is, for a fixed type of transistor)... so is it something about the geometric configuration of the doped regions that breaks the symmetry?
Edited at 2013-01-15 11:50 pm (UTC)
nice. I recently discovered Hans Camenzind's book, "Designing Analog Chips
" (free PDF). It serves as a nice introduction to all this transistory goodness.
The student manual to Art of Electronics is also good. If you want a preview, there's a copy on scribd.
Psh you should use a FET.
Well I am no expert but FETs are good for switching applications. They can handle large load current with essentially no gate current (for MOSFETs anyway) and you think of the channel as a resistor rather than a diode. Attainable values for typical FETs can be 4A and 0.1Ohm. But they are more expensive and easier to damage.
Not that you are switching more than about 24mA here. ;)
Huhhh. How can one damage them, if they are more robust in the face of large currents?
Also where do you get the number 24mA from? Surely it is true that it is indeed about that piddling amount, but Ohm's law says 30mA, obviously, if I just look at the resistor and voltage supply. Is there some rule of thumb about resistance across collector-emitter? Or the resistance intrinsic to the LED?
Edited at 2013-01-17 03:00 pm (UTC)
So in a BJT you apply current at the base and the collector-emitter junction acts like a diode, IIRC. In a FET you have a gate, source, drain, and (optionally) base; when you apply voltage to the gate, the S-D junction acts like a resistor. You can apply a lot of current through the S-D channel but the gate terminal is susceptible to damage. It's essentially one side a little tiny capacitor insulated from the channel; when you put charge into it, the electric field causes a change in the channel, but charge is not supposed to move from the gate to the channel, hence field effect transistor. But too much voltage will cause the insulator to fail.
Actually 24mA is too much, I think it's more like 15mA. The LED is a diode and the C-E junction on the BJT is a diode. Diodes have a nonlinear response; if you graph current vs voltage drop, it is a hockey stick. I'm assuming the bend in the hockey stick is around 0.6-0.7V for the components you're using, thus, assume 1.5V is dropped across the diodes and the remaining voltage drop across the resistor determines the current. (I think 20mA is about the max you want to drive typical LEDs with to not burn them out, BTW.)
Hm on further investigation it looks like when the BJT is saturated, the C-E drop is as little as 0.2V, but there is still a minimum drop. Anyway this acting like a diode means you are dissipating significant energy in the transistor if you want to drive a large current, which is one reason they're less preferable for switching. The other is that since you must send some nontrivial amount of current through the base to turn it on, you have to be careful about e.g. sourcing current from a microcontroller output pin, and probably have to daisy chain BJTs to switch large currents.
This is the "Darlington transistor" yeah?