Cumulus Support

looks like when the wind speed starts rising and at a certain speed during the rise (and occasional the fall) they appear

some soort of build up resonation

Don't know...

I admire your tenacity on this!

Have you tried tuning a radio to the station's frequency to see if there's any RF interference with your transmission? I'm fairly sure this is what causes my occasional spikes.

While my station is on 868MHz and my interference is fairly infrequent, 433MHz is a far more crowded band. (I think I read in another thread your station is a 433MHz?)

I used a scanner in AM mode and recorded the audio into Audacity. By comparing times, I found the spikes often correlated quite closely with interference on the radio.

ben-s wrote:I admire your tenacity on this!

Thank you

Have you tried tuning a radio to the station's frequency to see if there's any RF interference with your transmission? I'm fairly sure this is what causes my occasional spikes.
---

I haven't got a receiver that covers the 433MHz frequency. I did try making up a wide band AM receiver and feeding the detected output into Audacity but found nothing.

I used a scanner in AM mode and recorded the audio into Audacity. By comparing times, I found the spikes often correlated quite closely with interference on the radio.

That's interesting. At one stage I did get interference from an electric fence into the rain gauge but have had no trouble since.

The frequency of spikes (and amplitude) has been enormously reduced by the change in cable and addition of RC termination. Even running at a 48 second logging rate I only get a spike every few hours rather than several an hour and sometimes several degrees worth. I guess it's possible that I may have totally cured one source of spikes (anemometer) but another, lesser one, remains. If it is interference pickup, putting the cable inside the hollow mast may cure it. I expect to do that when the weather is more suitable! I would like to improve the stability of the mast too but without increasing the weight - I can just about manage it as it is, as long as there's not much wind.

Fantastic work Gina and thanks for sharing the results!
At one point you had a capacitor across the anemometer - is it still there?
It's probably bad luck to mention this, but my setup with 3 @ 10nF caps (one across the anemometer, and one from each or it's leads to the negative battery terminal) has pretty much eliminated spikes for me for about a month now. Of course, my extended cable uses CAT-3, which is unshielded, solid, untwisted, and is only about 8-9 meters long! And your measurements show at least one of those caps is probably useless...

Charlie wrote:Fantastic work Gina and thanks for sharing the results!

Thank you

At one point you had a capacitor across the anemometer - is it still there?
It's probably bad luck to mention this, but my setup with 3 @ 10nF caps (one across the anemometer, and one from each or it's leads to the negative battery terminal) has pretty much eliminated spikes for me for about a month now. Of course, my extended cable uses CAT-3, which is unshielded, solid, untwisted, and is only about 8-9 meters long! And your measurements show at least one of those caps is probably useless...

One side of the anemometer goes to battery negative, so one of your capacitors connects earth to earth and the other two are in parallel across the anemometer. So you have 20nF across the anemometer. I have 47nF and 100 ohms in series across the anemometer connections at each end. 22nF was sufficient to stop the spikes, 100nF would limit the maximum wind speed to something around 100mph or maybe less. Actually, one capacitor (at the bottom end) is still 22nF - I got too cold to change it!

This plot shows the temperature data since I changed cables etc. A couple of bursts of spikes and a couple of large singles. There are also several small spikes of a couple of tenths of a degree. This is with 48 second logging.And this is the last 24 hours at 5 minute logging - real and apparent temperatures.

OK so what's happening now?!?!? Here's the temperature plot for the last 24 hours. Reasonable until about 0300 this morning, then....More detail here in the plot for the last 6 hours...Those were with 48 second logging. Here's the standard 5 minute logging for the last 24 hours.It doesn't seem related to wind - here's the 24 hours wind plot.Yet more detail - 2 hours plot.

sorry to see it's still there.

first thing in my mind was "that's interference".
maybe a long shot,but still a possibility:
http://www.telepermit.co.nz/fence.html
more results:
http://www.google.com/search?client=ope ... 8&oe=utf-8

maybe from one of your neighbours(doesn't have to be next to you)

Thanks Hans - a good point I've just gone out to the shed and turned off the electric fencer. We have no animals out at night this time of year and our neighbour who often taps off ours is not using our fence supply ATM.

Time mark - electric fence switched off at 20:00.

maybe a hare or cat or whatever you have running in the wild,
in that case you could see by your temp readings if a rabbit is "well done"
for christmas diner

Fencer off for 20 minutes now and no change.

Switching the electric fencer off last night made no difference - nor did switching back on this morning.

As of 14:30 I have unplugged the wind sensors from the transmitter, to confirm that it still is something to do with them. That only takes a minute or two of being outside in the Arctic conditions. I now have a couple of opto couplers so can try one later, assuming a good temperature plot without wind sensors.

Just out of interest, how are you going to use optos with a circuit that reads a resistor array for wind direction? Are you building an external ADC of some sort?

It appears to be the anemometer (wind speed) that causes the problem. That is simply an on/off switch (reed switch) and it connects a digital input to ground (0v) or leaves it open-circuit. The opto-isolator features an IR diode and a photo-transistor - the diode can be turned on by connecting the anemometer reed switch to a resistor and battery. The photo-transistor can be connected directly in place of the anemometer connection to the transmitter. BTW - I don't claim originality for this idea, it was suggested by Hans earlier in this thread.