| Quote: Originally Posted by Daemon | | | | |
| Short answer, using the current 2.4Ghz XBee modules. No.
Because it is two way, the satellite receivers are at best, dumb, and at worst
a liability. They can't transmit ACKs without interfering with the main Rx's ACKs
(you'd actually be creating multipath interference at the source) so if the main Rx antenna
is blocked, and the satellite receiver antennas are not, the Tx will still think it has lost
the link, even though the Rx is still getting signals from its satellites.
That'd be ok, if it was meant to always be a one way system, but it's not. | |
| | |
People forget that we use only the XBEE hardware. The firmware is our own, and if you look at the chipset used with the XBEE, you can clearly see that multiple receivers can be used and the ACKs will not collide.
| Quote: Originally Posted by Daemon | | | | |
| As for hopping, it can be done, but it's always going to be *slow*, again
because the two way nature gets in the way and it's tough for both
ends of a two way link to know for sure if the link is 100% broken or only 50%
broken, and you risk more by hopping when it's 50% broken than you would by
staying put. | |
| | |
That is completely wrong. We can support 5ms frames, complete with ACKs and change frequencies every single frame. Frames are typically every 22ms. Think that is slow, well I have some news for you. Spektrum frequency hops every frame between two frequencies. They don't "lock on" to two channels as they claim. Every 22ms, the data is transmitted to the other frequency. Interesting too, since their FCC ID clearly shows their system to be a DSSS configuration. Part 15.247 clearly states that any device changing frequencies faster than once every 400ms/dwell is required to occupy no fewer than 15 channels. Hopping every frame is not legal by these rules. When I added the "hop on saturate to foil the bench testers code", I asked for clarification from the FCC on the hopping speed and we either have to transmit with a psuedo-random 15 frequency algorithm, or we can not hop faster than roughly once every 300ms, which is actually faster than the 450ms spacing we had in the original hopping code. So, it was a benefit to look into this.