Need help picking out hardware for wireless communication for jeopardy buzzers

[Aurora900]

So, as the title so descriptively says... I'm trying to build a wireless jeopardy buzzer system.  I had initially wanted to use something like the Xbee, which can do point to multipoint...  But it seems to only be able to natively transmit from one device to many receivers, as opposed to many devices transmitting to one receiver.  I need the latter, of course.

So I'm wondering what the best way to go about this is?  I'd prefer to use something I can integrate with an arduino... but I can't seem to find much info on a network setup like this.  The only major requirements for this are that the transmissions are reliable and that the receiver can determine which transmitter the signal came from.

Of course, you can always just tell me I'm crazy and I'll go back to designing a wired system instead

[Aurora900]

Maybe something like a CC1101 or one of the many Nordic parts. Steer clear of Hope RF.

One problem you will face is when two people press a buzzer at the same time. If both transmit at the same time they will talk over each other. The simplest option would be to poll each button in sequence. It's not perfectly fair but if you do it fast enough it should be more or less impossible to tell

I thought that might be an issue...  But the arduino likely runs fast enough that you wouldn't even notice, like you said.  I was considering the option of polling in sequence... I was thinking I could probably set it up to poll for a certain amount of time, and then all packets received in that time frame would be ordered based on a timestamp set by the transmitter.  That way, if your buzzer sent its packet at the appropriate time but was skipped over initially, it wouldn't be discounted.  Then we begin an entirely new issue... syncing the time on all the buzzers.

I'm starting to like the idea of going back to wires haha

[Marco]

Like this? (Was on Hack A Day a week ago.)

[daveatol]

If there are issues with reliability of the transmission, you can just have each transmitter transmit multiple times, and include the transmission number and the amount of time since the first transmission. The time between transmissions can be random, or can rely on the ID of the button (e.g. button 1 transmits every 9 ms, button 2 transmits every 11 ms, etc.).

With the nRF24L01 transceivers, one unit can be set to receive/listen constantly.

[z0r]

I've used the programmable xbees in digimesh for multiple -> one.  You get a little HCS08 chip on the xbee itself to do the thinking and only really need to add power and a buzzer button.  You'd need to write a short bit of C but the sdk comes with examples that are very close to what you want.  Dev kit, 3 modules with a programmer, is around $300, mainly because the programmer itself adds a lot, I don't think there is a cheap one

[daveatol]

The nRF24L01+ modules are something like $1.50 delivered and there are Arduino libraries written for them already

[sleemanj]

http://maniacbug.wordpress.com/2012/03/30/rf24network/

[Aurora900]

I'm liking the idea of the nRF24L01+, mainly because of the cost lol.  I'll have to look into it more.

And thanks for the link sleemanj, that looks like it should be a good read for getting acquainted with the transceiver

[lgbeno]

Since it is just one button you could just use an oscillator to turn on or off then space out each button at a slightly different frequency.  On the receiver side, it could use a RTL-SDR to sample the whole bandwidth and find what frequency spikes up first.  You could do something similar with a micro too.