Logic Gate Board Game 01: Project Planning -- Episode 282

For magnetic connections to work you need to ensure that the contacts are clean for a reliable connection.

I'd just use small clusters of machine pin sockets and 24AWG single core/strand insulated wire for interconnection.

The wire is cheap and is easily replaced.

The dexterity required to plug them in is good for the development of kids skills.

PSoC 4 Rocks for $1 you get 36 I/O pins and some CPLD capabilities (including Shift registers and LUTs) !

(Why not cascade two for 70 user I/Os)

Using buffered LED drivers is the way to go if the price point permits it.

The finished product must be Pb (Lead) Free and devoid of any dangerous residues and chemicals.

Staying with the magnet idea...

• simply press fit magnets into a PCB at every input and output node
• they can be glued on the bottom as long as electrical contact is retained
• the SMT ICs can be on the underside, leaving the top clear for graphic gates to be silk screened
• to interconnect, simply use steel wire, preferably insulated except for small segments at each end (paper clips come to mind)
• this allows multiple low cost connections to each node
• I like Gord Goebel's suggestion of current limiting resistors on outputs
• the logic trainers I've seen in the past have had LEDs at all inputs and outputs, this could work if LED current is kept at .5 mA
• the system will need some switches and pushbuttons to initiate or select sequences

Consider a shift in register instead of a high pin count micro (such as CD4021B).

Using shift out and shift in registers with a big pcb means adding things like extra LEDs or detecting which ports have plugs wouldn't be that expensive.  These might be nice additions to help users find bugs or get them started (just connect the 2 flashing ports).

You could also consider making this a headless IoT device using something like esp8266 or nrf52x.  There are some advantages to using a person's phone rather than having your own screen on the device.  But the obvious big downside is you need to write the mobile app.

Hi  just seen the video and got a suggestion

use a I2C controller for the LCD display than you have to use only 2 I/O ports for controlling the display.

Sorry I thought you were referring to the argument between Ben and Karen over buses for single signals going to multiple gate inputs.

So each input and each output would have an I2C module monitoring it or driving it? Or just a small/inexpensive MCU with I2C I/O expanders instead of a huge, powerful and expensive MCU just because it has lots of native I/Os?

PS could you tell me more about learning logic with cards, scissors and needles?

The buses I mention are not for the participant of the game but a way to wire up the components.

I use buses often to reduce wire count.

I2C, 1-Wire, Ethernet are schemes that only require two wires instead of a rats nest.

You made your own McBee cards? That's very cool.

I don't suppose you have any pictures, do you?

You don't need buses for complicated circuits, just two connectors at each input will let you daisy-chain sources to any number of inputs.

The jumpers should be directional (an input end and an output end) and the connectors at each gate input should be one input and one output (pass-through) so you can't connect two sources to the same input (and thus to each other) to prevent damage to outputs. If you go with stackable connectors then each cable should only be stackable at one end. The end that goes to an input should only be able to connect to an input and only accept an output connector on top. The end that goes to an output should only be able to connect to an output or the back of an input connector.

e.g. The output contact of each gate could be a magnet in a recess and the input contacts could be a bit of iron that sticks out. Then the output connector could have an iron bump that goes into the recess and touches output contact and the input connector could be a doughnut with a magnet in the middle of the hole. So an input connector would snap onto and input contact (either way) and leave a magnetic recess on top that the output connector could snap into to daisy-chain the signal.

I was thinking of a pegboard cross-point switch but it has some potential problems and I realized your idea of slide switches has a lot of potential, especially if there is a way to make it look like a circuit diagram.

The way I see it, there would be a slide switch for each input to select the signal source.  Sources could be one of the four data inputs or any of the gates outputs.

Of course the four output lights would each have a slide (or rotary) switch to select its source.

Multiple gate inputs could connect to the same source by setting them to the same position, and it's impossible to connect outputs to each other.

The slide switches could be purchased but it would probably be cheaper to make 3D printed bits that slide across pads on the circuit board (built into the case) and the number of positions wouldn't be limited to what's available in stores. Of course you could use a bunch of slide pots but you would need a bunch of analog inputs. Or, digressing even further from the visceral jumper wire concept, you could have a bunch of LEDs (charlie-plexed of course) to indicate the position of virtual slide switches which could all be adjusted with just two rotary encoders.

If space is not a constraint then you could use premade modular connectors as used for Ethernet and phone wiring.

It's a great upcycle/reuse idea.

Also Springs as connectors also spring to mind.