Hacker News Reader

Comment by CarolineW | Hacker News Reader

It's also worth mentioning that transistors are basically back-to-back diodes, and that understanding a diode, really understanding a diode, is the first step towards understanding semi-conductor physics, and electronics in general.

I remember well making crystal radios, firstly with store-bought diodes and capacitors, subsequently making my own capacitors, and then diodes. Taught me so much.

The diode - gateway drug to semi-conductor physics and electronics.

View on HN

So true, one of the coolest moments in college was taking a final where the problem described a PN junction and asked a series of questions about it. That required that you analyze the junction and derive all of the properties of the resulting diode (as I recall it was a crappy diode). But being able to think about computers from a quantum mechanical analysis, up through register transfer logic (RTL) was an amazing feeling for me.

If you want to understand computing as a fundamental level, starting with diodes is a good first step.

of course no one makes bipolar computers .... understanding FETs without understanding the quantum stuff is pretty easy

Tell that to the PDP-8 on my dresser (it's all Bipolar TTL logic, but you are not wrong). FETs are a lot easier to understand in many ways, there is a reason they took over as the dominant semiconductor switching element :-).

Although my favorite FET demo is a piece of surgical tube connected to a faucet, and a water balloon sitting on the tube as the "gate". You can demonstrate all sorts of things like avalanche voltage (water flow regardless of the gate state), gate leakage (poke a small hole in your balloon), pinch off potential etc. Makes a great science fair project for you parents out there.

A transistor is not two diodes back-to-back. It may be literally true that a transistor is two P-N junctions, but if you think of a transistor this way, you'll never understand them. The magic of the transistor is that if the P-N junctions are close enough together, you get transconductance -- when either junction is forward-biased, current can flow between the collector and emitter. Diodes don't do this!

See my comments here: https://news.ycombinator.com/item?id=11522126

You are right about the difference, you are wrong when you say that thinking about them like this will mean I'll never understand them. It's the merging of the N (or P) bands that makes the magic of a transistor happen, but not thinking of them as diode-like junctions means you're only thinking about them one way, not as an integrated concept.

We are probably in violent agreement, but I suggest that if you only understand diodes as diodes and transistors as transistors, you are limiting your points of view, and potentially not seeing the whole picture.

In fact looking at them as a diodes can let you use some bjts to get a fast low leakage diode in a pinch.

http://electronics.stackexchange.com/questions/16966/transis...

I don't think that's true. If you out two diodes back to back you won't get a functioning transistor. A transistor works by letting current run across because electrons are filling holes in the middle region, this smooths the potential across the entire component, allowing for travel. If you just out diodes back to back, the intervening wire segments disrupt the analogous structure and also the individual diodes probably don't have their potentials calibrated to provide efficient switching action.

You can't take separate two diodes and connect them back-to-back, because wire is in the way. When you connect two diodes in a circuit, that's not connecting the junctions back-to-back.

But if you take the PN junction of one diode, the NP junction of another, and then merge the N of one with the other, you have a PNP transistor. The level of doping in each layer matters, and that then (in part) controls what type of transistor[0] you get, but it is, honestly, two diode junctions. What you have done is to think about connecting diodes in a circuit, which is not the same thing at all.

[0] There are switching transistors and amplifying transistors. Then there are field-effect transistors which are another animal entirely. And more. But in essence, a PNP transistor or an NPN transistor, are each two diode junctions.

I think we just have different definitions of "back to back". But I don't think you could even take two diodes and put them together the way you suggest (without a wire) and make a transistor. Correct me if I'm wrong, but to simplify: part of the reason why the transistor works is that the "meat of the sandwich" is extremely thin. My understanding of transistors could be deficient since I only fully understood them recently, and for a rather... Unconventional context:

https://docs.google.com/document/d/1yeiOE1Gy0YoPXlLdMo9y8XSS...

Wouldn't it be awesome if two diodes connected back-to-back would form a transistor? But alas, [1]

[1] http://electronics.stackexchange.com/questions/78366/why-can...

This is somewhat offtopic, but I've tried building crystal radios over and over again in my life, and have never gotten them to work. I've built them from kits. I've studied them online and built them from scratch, in various designs. I'm in a city so the signal should be strong enough. I wish I could figure out what went wrong. I am able to build any other type of circuit that I've attempted.

Have you checked for absolute certain that there is an analog AM transmitter in your area? Simply being in a city is no longer enough. You might want to check.

https://youtube.com/watch?v=HvLmJGM_T3M

Try listening at night maybe?

Did you connect the ground to an actual earth-connected cold water pipe? I made that mistake when I was a kid, thinking I could ground it to any old metal pipe.

Either not getting a good ground, or not having a long enough antenna are two issues I have commonly seen with some of the more basic kits.

I suggest that you join one of the Crystal Set Forums.

eg "The Radio Board" http://theradioboard.com/rb/index.php

There are many patient Guru's there who will help you resolve any problems.

"gateway drug to semi-conductor physics and electronics." => So true. But I will wait a little bit more than 9 years before teaching semi-conductor physics to my little boy ;-)

But I will wait a little bit more than 9 years before teaching

Why? When I was that age, I would tear apart all of the electronics in my possession trying to figure out how they worked. I never could figure it out, until much, much later. Not because I couldn't have understood it, but because I didn't have anyone to help guide me with regards to electronics.

edit:

I would have killed for someone to have gotten me something like http://www.amazon.com/Snap-Circuits-SC-300-Electronics-Disco... which is listed as appropriate for age 8

because he is 3 months old now :-)

Semi-conductors physics may be too much for a 9-year-old, but an introduction to transistors or doing a H bridge to reverse the rotation direction of a motor could be pretty impressive.

Ah, I see. Totally misunderstood what you said, sorry!

Even Horowitz and Hill [1] don't explain the transistor in terms of semiconductor physics. Yet they provide useful operational insights.

[1] http://artofelectronics.net/

I was exactly the same; once, just once, one of my friends told me his Dad was into electronics; he showed me a room full of his Heathkit kits in various stages of construction (I remember the full-size organ). But we moved that summer before I could ever meet him - he may or may not wanted to mentor someone in grade school, but I was so very eager to learn.

Reply on news.ycombinator.com