I'm designing some educational toys to teach children about electronic and logic circuits. They are quite literally individual electronic components (switch, resistor, transistor, capacitor, ...) and digital logic gates (and, or, not, ...) in 3D-printed encasings that can be easily and safely clicked together.

I'm familiar enough with the theory, but I have near zero experience reading data-sheets and choosing the right physical components. I would appreciate some names / serial numbers, or anything else to point me in the right direction. Ideally they'd be just large enough for me to solder them by hand and they'd work on a AAA battery or similar.

If needed, I could assemble the logic gates myself, but I'd prefer not to. I only need a single gate per physical component (1 or 2 input pins, 1 output pin, 1 pull-up/pull-down pin), though I can imagine they may always combine several in one. And if they're cheap enough, that's fine.


closed as off-topic by Dmitry Grigoryev, uint128_t, Voltage Spike, pipe, DoxyLover May 1 '17 at 5:54

This question appears to be off-topic. The users who voted to close gave this specific reason:

If this question can be reworded to fit the rules in the help center, please edit the question.

  • 3
    \$\begingroup\$ That's what Snap Circuits is. snapcircuits.net \$\endgroup\$ – Brian Carlton Apr 26 '17 at 0:34
  • 1
    \$\begingroup\$ littleBits is also similar-sounding... \$\endgroup\$ – vicatcu Apr 26 '17 at 0:44
  • \$\begingroup\$ Oh, I'm aware similar products already exist. But part of the fun is designing and building them myself. And I do have a few twists in mind. \$\endgroup\$ – mhelvens Apr 26 '17 at 1:26
  • 1
    \$\begingroup\$ @mhelvens I already contacted Snap Circuits about the possibility of simply purchasing empty casings so that I could use my own circuits inside without having to re-invent the wheel. They were actually very supportive, held meetings about the idea, and now seem quite open to the idea. You might contact them. Unless you really want to go with 3D printed elements. I have two 3D printers here in home that I've used to build everything from a 6-speed transmission with reverse, to electronic project boxes. But I still would rather use high pressure molded parts. Just better. \$\endgroup\$ – jonk Apr 26 '17 at 1:37
  • 1
    \$\begingroup\$ @mhelvens I've been taking what you are thinking of doing and going it one further, so that one can build a full microcontroller on a table-top. Connected parts with a single wire are often (unknown to the student) serial buses. It's a 4-bit wide bitslice (to make it some work to get to an 8-bit cpu but still possible to reach for 16-bit with enough parts.) \$\endgroup\$ – jonk Apr 26 '17 at 1:41

Your main difficulty will be protection of the logic components against ESD and overload.

My first thought was to suggest something like LS as being more robust than HC, but it's not immune to ESD, and being TTL uses asymmetric thresholds, has significant bias currents and a tight VCC range, which make it less than ideal for the neophyte.

With a few extra components, HC can be made fairly bomb-proof, and as it's specified from 2v to 6v, it would allow an unregulated 2 or 3 cell alkaline power supply (4 cell even, abs max is 7v) to be used. I'm thinking a diode clamp pair on every module pin exposed to the user, with 200 ohm series to the IC outputs, and 10k series to the IC inputs. That gives ESD protection two bites at the cherry, and allows LEDs to be connected directly to outputs if you want. Don't forget VCC capacitors.

How instructional and how representative do you want these to be? Do you put a 1M pulldown on every input? Yes if you want them to 'just work' with inputs left open, no if you want them to find out the hard way about leaving CMOS inputs floating. I think I've just persuaded myself into recommending pulldowns, as some logic families can legally leave inputs open, CMOS will draw excessive power when oscillating, you're not necessarily teaching them CMOS but generic logic, and you can specify in your book of words that 'this logic defaults to LOW when unconnected' (or HIGH, or active or inactive (think which you need from a didactic point of view)).

To expand on the didactic use of default inputs, if you provided (say) a 4 input AND gate, and defaulted the inputs HIGH, then it would work as a buffer, or a 2, 3 or 4 input gate by just connecting the inputs you needed, while leaving the others open.

Continuing with didactic considerations, what logic functions do you want to provide? While learning logic, I was always thrown by NAND gates, 'why do they have to provide the complement, what's wrong with an AND gate?' Because it's simpler inside, a NAND uses less power and is faster than an AND, which is why you'd choose those for a real design. But is it the right thing to choose for neophytes?

You can get single gate per package CMOS logic in SOT23-5, or 4 gates per package in SO-14 or DIP-14. Fairchild's Tiny Logic (HS family === HC) or the Little Logic from Texas (AHC === HC) give you single gates equivalent to HC, though the abs max is lower at 5.5v.

You might want to consider that if you made a single board that took a quad NAND 74HC00, you could configure it to be a NAND, AND, OR, NOR, Inverter and even a 2:1 multiplexer, by cutting tracks or solder-blobbing links on the board, if you wanted to minimise the number of different parts you used. Unfortunately, a quad NAND won't give you an XOR, you need 5 NANDs for that, so just use an 74HC86. But that's just me being old skool. As CL points out in comments, the Texas 74LVC1G199 'ultra configurable' gate does all of those, with Schmidt inputs, so maybe that's your one part.

Schmidts are useful for delays, essential for clocked logic if you want to go there. This is specified to behave itself when input voltages are not clean logic levels, allowing it to be used after timing RC delays, as an oscillator, and for switch de-bounce. Plain switches give multiple transistions, de-bounce is essential if you want to clock them manually.

  • \$\begingroup\$ TI says that AHC is different from HC. And there are better choices for flexible gates with Schmitt triggers. \$\endgroup\$ – CL. Apr 26 '17 at 7:00
  • \$\begingroup\$ @Neil_UK: Thanks for your detailed answer! --- If it's not an imposition, could I have (say) 15 minutes of your time some day this week to ask some questions? You made me realize I'm missing some fundamentals, and I think a quick back-and-forth via something like Skype would be more efficient than a discussion here. If that's not convenient, no worries. I will figure it out. \$\endgroup\$ – mhelvens Apr 26 '17 at 18:37

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