6
\$\begingroup\$

I am currently teaching my daughters Digital Electronics, ages 9 and 12. In doing so, I wanted to introduce them to CMOS logic, since I feel that a hands-on approach is always better than just doing theory.

I purchased a few NMOS and PMOS through-hole type transistors intending to demonstrate basic logic gates such as NOT (1 NMOS, 1 PMOS), NOR and NAND (2 NMOS and 2 PMOS). That worked great, and they could reason about why the gate worked the way it does.

Next up is XOR and XNOR, but it stumbled me a bit. I know how to make a circuit using nine transistors, but looking online I've seen 6-transistor and even 3-transistor solutions. However, these assume you can change width and length, so these are solutions meant for ASIC designs, not for through-hole components.

So, my question is - is there a way to demonstrate XOR/XNOR gate logic with off-the-shelf CMOS/PMOS transistors using only a few transistors? Six transistors would be all right, eight slightly worse yet doable.

Perhaps I should just skip XOR/XNOR and move on to latches and flip-flops?

\$\endgroup\$
  • 4
    \$\begingroup\$ At some point, building the actual circuits from discrete transistors becomes more tedious than educational. At that point, you introduce the concept of ICs (SSI/MSI) and use them to build circuits. Show how they work internally -- built up from the simpler gates they already understand -- by looking at the databooks. Confirm their operation on the breadboard. \$\endgroup\$ – Dave Tweed Apr 1 '16 at 23:20
  • \$\begingroup\$ What about possible ESD damage ? Remember that in discrete Mos transistors the substrate is connected to the source .Would you consider BJTs to teach first principles before moving up? \$\endgroup\$ – Autistic Apr 2 '16 at 0:47
  • \$\begingroup\$ The kids have messed with these particular CMOS transistors and they all seem to be working just fine, so perhaps that danger isn't as big as it used to be? \$\endgroup\$ – Pål-Kristian Engstad Apr 2 '16 at 0:54
  • \$\begingroup\$ @PKE , That is so cool. They will appreciate it later on. \$\endgroup\$ – Tim Spriggs Apr 2 '16 at 12:53
  • 1
    \$\begingroup\$ @Autistic I wouldn't worry too much about ESD. I experimented with BJTs and CMOS logic a long time ago when I had zero understanding of ESD. I might have damaged a few components but that was usually not because of ESD but other mistakes I made. In my experience, the old 4000 series CMOS logic is very robust regarding ESD. \$\endgroup\$ – Bimpelrekkie Apr 2 '16 at 13:15
2
\$\begingroup\$

As always, Wikipedia ia a good place to start and here's a CMOS circuitenter image description here and 6 transistors is what you asked for. Of course, it's not the most obvious of circuits. I'd recommend that it's time to transition to CMOS ICs, with the 74HC series my preference. You can make an XOR gate from

4 NAND gates

Since a 74HC00 package conveniently has 4 NAND gates, it's a good place to start. Get them to build up one XOR, and then show that 4 of those fit in a single 74HC86 package, and you're on your way to teaching them about the benefits of integration, starting with SSI (small scale integration).

\$\endgroup\$
  • \$\begingroup\$ Thanks, that's good advice. However, I've seen that circuit. If you build that circuit using through-hole transistors, you will find that the outputs are not quite what you expect. \$\endgroup\$ – Pål-Kristian Engstad Apr 2 '16 at 0:41
  • \$\begingroup\$ Yes, the 6-transistor circuit requires separate connections to all of the transistor substrates in order to work correctly. With each substrate tied to its source, the body diodes mess up the logic. \$\endgroup\$ – Dave Tweed Apr 2 '16 at 11:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.