Advice on a wireless fencing scoring system?

yoseph1998:
Well in Epee the system needs to be able to lock out an opponent in 50ms. So I'll say that this needs to happen in non more than 1ms. Personally I don't mind if the master box buzzes 5ms too late after a touch as long as the master box knows it's buzzing 5ms too late. If I can do some extra work with time stamping and verification and making sure everything is very well synced then I don't mind the box being 5ms late on buzzing. Ideally not.

That, together with the helpful comments by @dave-in-nj, suggests to me that this is a very challenging project that will need a lot of time and experimentation to develop - if it can be developed.

IMHO it would be quite impossible to achieve that level of real-time accuracy relying only on the "immediate" transmission of a message by an nRF24 (or other comprehensive wireless system).
As an aside, you might be able to do it by using something to detect an LED being switched on (light is a form of wireless) but that would probably not be practical where there are several fencing matches at the same time. And the detecting would probably need to be done with a camera and sophisticated image recognition software - not a job for an Arduino

That means that each slave must record the time of the event and send that to the master. But can you synchronize the time on two separate slaves with sufficient accuracy. Your first experiments will need to determine that. In general I don't think the Arduino's 16MHz clock is sufficiently stable for that sort of use.

If you can synchronize the time so that events are recorded properly you are then faced with the issue of how long it takes that information to get to the master from both slaves in a pair (in case each slave has recorded an event). More experiments.

It seems to me that a quick response is important in case one competitor believes s/he has scored a winning stroke and consequently stops playing rather than continuing for another 5 or 10 seconds until the score is announced. However there is a big difference (in computing terms) between needing a decision within (say) 1 second and needing it within 50 millisecs. So maybe some careful thought needs to be given to the details of the requirement.

And after all that, there is the question of whether, and how, you would do that for multiple fencing matches at the same time.

Interesting problem. :slight_smile:

...R