RF transceiver chipset with low BER

[john smith]

I currently think of building a low power RF modem (433Mhz) + datalogger to
collect some data. I have look around at different RF transceiver chipset
manufacturers (chipcon, atmel, analogue device,microchip,mircel ect) and
they are all pretty much have a BER of 10^-3. Is possible to reduce the BER
by adding external components e.g. preamplifier on the RX pin?



Radiometrix made ones with BER of 1 ppm! anyone know what chipset they use?



thanks



John

 

[Tim Shoppa]

e.g. preamplifier on the RX pin?

That will just amplify the noise too (and probably make you more
susceptible to intermod etc.)

The 10^-3 BER comes out in certain conditions. Increase power,
decrease loss (e.g. move receiver closer to transmitter), lower noise,
or lower your data rate and you get much better BER's. Shannon's law.

Note that your real BER will be very different than the advertised one
because you're using it in a non-lab environment.

Tim.

 

[Guy Macon]

john smith wrote:

I currently think of building a low power RF modem (433Mhz)
+ datalogger to collect some data. I have look around at
different RF transceiver chipset manufacturers (chipcon,
atmel, analogue device,microchip,mircel ect) and they are
all pretty much have a BER of 10^-3. Is possible to reduce
the BER by adding external components e.g. preamplifier on
the RX pin?

Radiometrix made ones with BER of 1 ppm! anyone know what chipset
they use?

(Note: I am not an RF expert)

The app note at [ http://www.radiometrix.co.uk/apps/apnt100.htm ]
seems to indicate that a good SNR is the key to getting the BER
down. Do you have reason to believe that the transceiver chipset
adding noise at low input levels is a problem? If so, it would
seem that a good preamp would help. If, on the other hand, the
problem is too much RF noise in your band, it would seem that a
preamp would simply amplify the signal and the noise.

Have you considered using error correction? Look at the CRC at
[ http://www.radiometrix.co.uk/products/bimsheet.htm#pac-data ]...

I am sure that the real RF experts will have better suggestions.

--
Guy Macon <http://www.guymacon.com/>

 

[PeteS]

The BER is going to be directly proportional to the S+N/N ratio. That
will be determined by the received signal strength, although different
modulation techniques will give a different curve of BER vs. Rx signal
absolute strength.

You should note in the data sheets that the BER listed (for a typical
device from Microchip,
http://ww1.microchip.com/downloads/en/DeviceDoc/70090a.pdf ) is at a
specific sensitivity. For instance, it states that the BER shall be
<=10^-3 (for a narrownband FSK signal) when the received signal
strength is -109dBm. (Put that in perspective - it's a received signal
power, at the device, of about 14fW).

At higher receive signal strengths, the BER will be much better. 10^-6
is known as 'carrier class', incidentally. for any installation with
transmitter and receiver fairly close, you should easily achieve 10^-6
BER.

Cheers

PeteS

 

[Lasse Langwadt Christense]

john smith wrote:

I currently think of building a low power RF modem (433Mhz) + datalogger to
collect some data. I have look around at different RF transceiver chipset
manufacturers (chipcon, atmel, analogue device,microchip,mircel ect) and
they are all pretty much have a BER of 10^-3. Is possible to reduce the BER
by adding external components e.g. preamplifier on the RX pin?

Think you misunderstand the spec. the usual specification for sensitivity
is the input level to the reciever that gives you a BER of 10^-3

you can get pretty much any BER you like, it is just a matter of how high the
input levet is ...

So you are comparing apples and oranges ...

Radiometrix made ones with BER of 1 ppm! anyone know what chipset they use?

it may have a BER of 1 ppm at and input level of -80dBm, and a BER of 100ppm at -100dBm
it is different measurements and can directly be compared


-Lasse

 

[Lasse Langwadt Christense]

Lasse Langwadt Christensen wrote:

john smith wrote:

I currently think of building a low power RF modem (433Mhz) +
datalogger to collect some data. I have look around at different RF
transceiver chipset manufacturers (chipcon, atmel, analogue
device,microchip,mircel ect) and they are all pretty much have a BER
of 10^-3. Is possible to reduce the BER by adding external components
e.g. preamplifier on the RX pin?


Think you misunderstand the spec. the usual specification for sensitivity
is the input level to the reciever that gives you a BER of 10^-3

you can get pretty much any BER you like, it is just a matter of how
high the
input levet is ...

So you are comparing apples and oranges ...



Radiometrix made ones with BER of 1 ppm! anyone know what chipset they
use?


it may have a BER of 1 ppm at and input level of -80dBm, and a BER of
100ppm at -100dBm
it is different measurements and can directly be compared

cannot be be directly compared ...

-Lasse