Transistors (by alaric)

So I went and studied IGFET transistors too (commonly, though not always correctly, known as MOSFET). Turns out that what most people mean when they say FET is an enhancement mode FET, although there are also depletion mode FETs.

But enhancement FETs, like bipolars, come in two flavours; N-channel and P-channel. I know from past experience that the resistance of the 'channel' between drain and source (what's the difference?) is controlled by the gate voltage - but what is that voltage measured relative to? Voltage is not an absolute property in electronics. And at what kind of voltage range does the device switch between on and off? How wide is that range? What's it behave like in that range?

Again, I pored over electronics books (I found a very very good one online: Lessons in Electric Circuits) and found out what I needed to know.

It turns out that, contrary to what simplified diagrams tell you, the drain and source in a FET are not identical. I'm not sure what the difference really is in the silicon, but what matters with regards to controlling the transistor is the voltage betwen gate and source. This is helpfully shown for you in the circuit symbol; the source is always the one connected via an arrow to the 'middle' of the transistor. So you can imagine the gate->source voltage difference acting across the center of the channel.

But what's the difference between P and N channel? Well, a P channel enhancement FET switches on as the gate becomes more and more negative compared to the source, while a N channel enhancement FET switches on as the gate becomes more and more positive compared to the source. For both of them the resistance is fairly infinite when the gate->source voltage difference is zero, but they drop off fast as soon as it becomes over a couple of volts; N channels fall to around 25Ω at +3v, while P channels fall to around 10Ω at -3v. Note that P channel FETs have higher 'on' resistances, due to N-doped silicon being a better conductor than P-doped.

So armed with this new knowledge, I then knew that an N-channel enhancement FET connected with its source to 0V and its drain to the base of the power control bipolar transistor (via a 10KΩ resistor, since only a few mA need pass to fully activate the bipolar) will take a 3.6V CMOS logic level on its gate input. When the input is 0V the transistor will have a very high resistance, and no current will flow through the base of the bipolar, so the charger will be disabled. But at 3.6V input, the transistor's resistance will just be tens of ohms, meaning that enough current will flow through it and the 10KΩ resistor in series out of the base of the bipolar to allow it to supply a few tens of mA to the charger, thus switching it on.

I'm now putting together some nice A4 sheets explaining all of this with diagrams, which I plan to print and laminate and pin up where I do circuit design.

Check out their current status at the Quick Reference Cards Project Page.

As the licence on the ref cards suggests, I want to make them publicly available, but I wonder if there would be a market in mail-order of nice laminated printed sets? Hmmm...

Pages: 1 2

3 Comments

• 

  By Ben, Tue 13th Sep 2005 @ 8:14 pm

  Just two things...

  1) You might want to actually upload your cards.

  2) When's my board going to be ready? 🙂

• 

  By andyjpb, Wed 14th Sep 2005 @ 7:43 pm

  I wish someone had told it to me like this when I was studying 2nd year electronics...

• 

  By Alaric Snell-Pym, Thu 15th Sep 2005 @ 6:34 pm

  I DID upload the cards... yes, they're there! Works For Me.

  Boards will be ready as soon as I can build them all, now the circuits seem to work on breadboard 😉

  And yes... I also wish transistors were taught that way. I've learnt about them many times before, but found myself forgetting the beginning before I got to the end. Why should something so simple be handled so verbosely?

Other Links to this Post

RSS feed for comments on this post. TrackBack URI

Leave a comment