Posted on May 17, 2011 at 12:45Ok Russ1
I will give you some directions since the subject of RF and transmission
requires a lot of expertise and some consider it ''black art''.
Separate first of all two worlds and one is the radio part
and the second is the digital part .
It's funny to evaluate radios in terms of standby current and data rate
since for radio guys the real parameters judging the reliability of a radio
are usually hidden or specified very shortly.
Keys parameters in a radio design regarding the receiver are it's noise figure , compression point, image rejection, adjacent channel rejection and a long list of other parameters that mainly describe the range you will achieve and the behaviour your link will have if it's surrounded by other closely operating rf links .
Regarding the transmitter you have parameters such as output power , adjacent channel leakage , modulation efficiency , harmonics suppresion etc. Finally the antenna is another serious factor not to be taken lightly as is the path loss calculation, fading etc.
These chipsets you are referring on the market are not designed for anything serious regarding radio performance.
For example it's nice to see those +10 dbm chipsets operating at 2.4 GHZ but it's doudtfull the results you will achieve operating in frequencies occupied with other wlan's . Don't expect a long range link . On the other side if your demands are short range then you can give it a try.
Follow these rules
Use a formula to calculate open space path loss so you evaluate ''link budget'' offered by the chipset.
If you see high sensitivity in a chipset check immediately compression of the receiver.
If you need extra db's power check antenna alternatives .
Adding an amplifier will add up to a mess that usually leads to trouble since these chipsets are not easy to deal (extra isolation etc.) Finally legal problems if you transmit higher than power dictated by regulations and if you don't have a spectrum analyzer handy usually spurious and harmonics are the rule.
IN the past i have designed for a customer short range in house wireless link based on the nRF2401 .Speed i have achieved is 57600 bps and it was working with an AVR . No protocol is required since it was just a virtual serial port. What nRF2401 lacks is RSSI , easy amplifier connection , and the receiver is not of serious performance (as most of these chipsets anyway) . The specs though regarding distance and cosumption were ok so i implemented it .You don't need a protocol to implement what you want .Code size was about 2K written with a c compiler . What the nRF2401 is missing is an RSSI measurement and easy external amplifier connection (the rx-tx switch is inside) .Although the link budget was marginal i implemented prototype based feed closed dipole antenna printed on FR4 that gave omnidirectional coverage and added another 2.1 dBi (If you don't have a vector network analyzer then you need black art !!! to tune it) that increased the link performance to acceptable values.
From my research recently, i like the CC1101 with STM32Fxx which seems more adequate for long range communications (with external modifications) and low power consumption but if you don't test these things by yourself you can't easily express an opinion .
Sorry for my long talk but it's a big subject.!!!
[ This message was edited by: mkefalas on 29-01-2009 20:23 ]
