endrew wrote:ok, I used this calculator to compute the path losses:
http://www.daycounter.com/Calculators/F ... ator.phtml
output power = 50mW ->17dbm
freq = 900Mhz
distance = 9600m
Gtx = 2dbi
Grx = 2dbi
the result I get:
Prx = -90.2dbm
now, according to the Xbee 900 manual, the Receiver Sensitivity is -100dbm so i should be able to get more than 10Km... which I know cant possibly be true...
any way I decided to use the Xbee XCS (100mW) and a 2dbi dipole antenna:
http://www.sparkfun.com/commerce/produc ... ts_id=9143
I hope it can provide a 2-3Km range. Ill be thankful to anyone who can tell me if I'll really get this range.
-100dBm is the claimed sensitivity to get a certain bit error rate, and in the absence of interference and fading. In practice, you'll need to allow 6dB or so for fading in line of sight. For interference, you either do freq. hopping (which some Digi 900MHz products do), or if the interference is brief, just delay and retransmit n times - presuming you checksum/CRC the message, or rely on such in the link layer.
Of course, the required signal strength depends on the channel bandwidth and modulation method. Narrower channel = lower modulation rate = better sensitivity, and so on.
The receiver noise figure has to be considered to, and down at -100dBm and weaker, and with channels hundreds of KHz wide, or more, it gets to be a consideration.
so a 100mW transmitter and essentially no antenna gain is marginal at 1Km line of sight. Beware the null in the antenna pattern of dipoles - parallel to the rod, off each end. "Pointing" th end of the dipole at the far end is not what you want to do!