so ive been messing around with this idea for a while and finally started actually building it out. basically i have four antennas - a 40m dipole, a 6m yagi, a 2m/70cm vertical, and a spare random wire i throw up sometimes - and im tired of manually going over to the relay box every time i switch bands. its not a huge deal but during contests it gets annoying fast.

i picked up an arduino mega because i wanted enough digital pins to handle the relay outputs without doing any multiplexing nonsense. the plan is to read band data from my radio over the accessory port (its a TS-590SG so it outputs band voltage on one of the pins) and then use that to fire the right relay automatically. ive seen a few projects online that do something similar but they all seem to use different radios so the voltage levels are all over the place.

my question is mainly around the relay driver side. im using a ULN2803 darlington array to drive the coils so i dont fry the arduino outputs, but im second guessing myself on the flyback diode situation. the relay board i got off amazon already has diodes on it but i wasnt sure if i still need them on the arduino side too since the ULN2803 has internal suppression. anyone dealt with this or have a similar setup running?

the ULN2803 has built-in clamp diodes that connect to the COM pin - you tie that to your relay coil supply voltage and it handles the flyback internally. so if your relay board already has diodes too youre double protected which honestly isnt a bad thing. i wouldnt lose sleep over it.

i did something similar with a Pi Pico actually, not an arduino, but same general idea. the band voltage approach with kenwood radios is pretty well documented, think theres even a table floating around on the kenwood reflector somewhere that maps the voltage levels to bands. just watch out because the voltages can be a little off from what the manual says depending on firmware version, so id recommend reading them with the arduino ADC and doing some averaging rather than hard thresholding.

yeah i ran into the exact same thing with my IC-7300, different radio but same band data headache. ended up just using the CI-V serial data instead of trying to decode analog voltages, way more reliable and you get the actual frequency not just a band approximation. the 590 does serial too if i remember right so might be worth looking at that route. takes a little more code to parse but its not bad at all and you know exactly what band youre on down to the hz which is nice if you ever want to do sub-band switching later or something.