OK, I'll answer my own post after further investigation...
I discovered that if the "see-saw" was slid close to the reed switch,
the reed switch was not opening in one of the rest positions. It did
open if the see-saw was slid to the other extent of its travel/slack
(only 1mm or less of movement).
Quite why this caused the multiple registrations, I'm not sure.
Perhaps there was some horizontal movement as it hit the rest
I tried moving the replacement reed switch 5mm further up to ensure
the switch opened at rest, but found that the switch opened briefly
at the mid point of travel (not good!).
I eventually used a 10mm reed switch, and even that switch only just
opens at one rest position with the see-saw close to the reed switch.
I used a digital oscilloscope to check for switch bounce. With this
reed switch, there is some bounce as the switch closes, but it is all
over within 40uSecs. Switch opening is clean. With this reed
switch, the switch is closed for between 110 and 150mSecs during the
I have tested it by dripping water in, and also by dribbling in at a
high rate, and have observed no missed or multiple registrations.
My recommendation to people replacing their reed switch is to test it
very carefully with a multi-meter before putting it back in service.
- test it with the see-saw up close to the reed switch, and far from
the reed switch,
- test it carefully to make sure the switch opens at rest with the
see-saw close to the reed switch.
--- In Lacrosse_weather_stations@yahoogroups.com
, "J.L. Blom"
> Do you hear the swicht click multiple times?
> It looks like you have a bad switch which bounces.
> On Wed, 2007-12-05 at 11:54 +0000, Ian Steer wrote:
> > I noticed while testing my rain gauge after replacing the reed
> > that dribbling water in caused the WS3600 to register mutiple
> > registrations for a single tip of the buckets. With my
> > connected, I could see the reed switch closing multiple times
> > should have just closed once. It seemed OK if I just dripped
> > one drop at a time.
> > Now I have become aware of the problem, I notice it all the
> > Whether it used to happen with the original reed switch - I don't
> > Has anyone found a solution to this problem ?
> > I'm thinking the best solution is to make a small electronic
> > between the rain gauge and the WS3600 that only
> > pulses to the WS3600 (ie filters-out the short pulses that could
> > possibly be correct). However, this seems like a lot of work to
> > Anyone got a simpler solution ? (because this is really giving
> > sh%$#@s now that I am aware of the problem!)
> > Yahoo! Groups Links