REV touch sensor does not work in SmartIO port

[alan412]

In a bizarre way, the Magnetic Limit switch does work but the REV touch sensor does not. The led on the touch sensor lights up when it is pressed, but the Systemcore doesn't detect it (on either the LCD or the web page)

[ThadHouse]

I'm guessing its because the touch sensor is wired to the N + 1 channel, and the adapters included with Systemcore are forwarding the N channel. Can probably be solved by swapping the white and blue pins on one end of the cables. The end goal would probably be a cable from Rev or someone that does this automatically.

The Magnetic Limit switch wires its outputs to both N and N + 1, which would explain why it shows up.

[alan412]

I just looked it up and the magnetic limit switch puts its value on both n and n+1, and REV has that the touch sensor is on n+1. So maybe the adapters should forward the N + 1 channel instead since they are the only 2 digital input devices (that I know of in FTC - other than handmade switches of course). Then it would work for both of those.

[ThadHouse]

GoBilda's Laser sensor goes into the N port, and it's digital input.

[alan412]

bummer. so much for a simple solution.... (Although I do wonder how many GoBilda Laser sensors are out in the wild. I have never seen one in person.)

[ThadHouse]

Realistically, I think the right long term solution is JST-PH 4 pin to Molex SL 3 pin cables, sold in both N and N + 1 configurations. In fact the wiring doc in this repo has that as one of the cable options.

[alan412]

The best long term solution is to have switches with 3 pin connectors on them. :-)

But for the medium term, we should make sure those cables are different colors because otherwise that would be a nightmare to diagnose.

[qwertychouskie]

How about adding a DIP switch on the adapter board to switch between N and N+1?

Also, having one 4-pin port per SmartIO port may not ever work for some sensors, e.g. this one that uses both N and N+1 for separate things: https://brushlandlabs.com/product/laser-rangefinder/ Perhaps it would be better to just have one 4-pin port per 2 SmartIO ports? Though arguably that makes the experience worse for teams using sensors that only use one channel, which most sensors do. A theoretical "perfect" solution would probably be for every 2 SmartIO to have 2 4-pin ports, each port with both N and N+1 wired up, and a DIP switch per port that swaps N and N+1 on that port. I believe this would satisfy all possible configurations, without requiring teams juggle a bunch of very-similar-looking adapter cables that are functionally different, or manually pin-swapping cables.