If you are not familiar with Instromet is rather difficult to describe the
wiring set up, but here goes:

The system runs from a inside junction box which receives signals via
cabling from the various external sensors - around 5. The output from the
junction box goes to both a large analogue display panel and separately to a
PC routed via a datalogger. The link to the PC uses an RS232 cable. However,
the system is linked to a second PC (used as a back-up) via a data transfer
switch - this splits the signal from the datalogger. All the sensors are
elevated some 8m off the ground and the wind and sunshine sensors an
additional 5m approx. on top of a mast. So I guess yes...it is acting rather
like an aerial.

However, the spikes are narrow - perhaps just one or two a day. I am
therefore assuming the problem lies with EMI. I've attached a number of
ferrite cores on the various cables as well as the serial leads and now just
have to wait and see if this solves the problem.

Mike

"Geo" wrote in message
...
On Fri, 28 Mar 2008 19:58:59 -0000, "Mike Lewis"
wrote:

I've discovered that the problems we've been experiencing on our Instromet
weather station with regards to spikes in rainfall and windspeed data is
due
to RF interference. The source of this is going to be difficult to track
down, but one solution I've read is to fit a ferrite? to the serial cable
connected to the PC.
You have not described the unit wiring/cabling setup. Is it a single
remote unit
with a single RS232 cable back to the PC? Or is there more wiring
involved? How
is/are the remote unit(s) powered?. Is the signal cable screened? Length
of
cable? Ground level or elevated?
Basically does it /look/ like an aerial?
RF signal are usually present for enough time to pass a message of some
sort
(speech/data) and a local transmitter could result in your aerial picking
up
enough to upset the sensors (probably not the PC). If you have virtually
continuous recording data and only get narrow spikes then I would be more
inclined to suspect the power source.

Geo