Tempature Spike Sources?

[ScottVan]

My station has been up and running in its new location for about a month now. I've been getting random outdoor temp spikes for the month. Nothing bad - all taken care of (mostly) by the spike removal function. I'm now more curious to figure out how/where they are coming from. I've put ferrite cores on each end of the Cat 5 TP wire I used to remote-locate the sender unit.

Some RF hazards I can think of in my world:
- Air compressor start up (in garage - big capacitor start motor)
- Electric dryer
- Car start ups
- Military aircraft flyovers (I'm on final for a C-130 airbase)

Since I used only 3 pairs in the cable, would it make sense to ground the other 2? I have an 8 foot grounding rod with a 6 gauge solid core ground wire up to the remote sensors. You may or may note recall I have 2 antennas up on my mast as well but both are well grounded and unused of of today. In fact I did test the 2 meter antenna for 2 days with my 5 watts APRS station - only a few spikes but they continued when I switch off the station.

Otherwise, up and running well. Experienced first snow this weekend - that's another topic!

[AllyCat]

I've put ferrite cores on each end of the Cat 5 TP wire I used to remote-locate the sender unit.

Hi Scott,

Does the cable extend the sensor cables, or have you another remote connection for example to use a distant power supply? I can't see any harm in "earthing" the spare conductors in the cable, but you could connect to a "real" earth, or to the negative battery connection in the transmitter (or both)

AFAIK nobody has definitively identified whether the spikes are in the (received) radio transmission or on the A/D converter inside the transmitter unit. However, probably both can cause spikes in the data:

The transmitter sends a burst of data lasting about 50ms once every 48 seconds but doesn't report (or show) lost data/sensor contact until about 6 - 8 consecutive messages are "lost". So any occasional interference with a duration of tens of milliseconds is likely to kill the whole data block and will not be reported or cause any data corruption. It's more likely to be impulse interference, e.g. car ignition (but all modern cars should be adequately suppressed) which may cause spikes in the data.

Probably interference from electric fences is not relevant in your case, but Gina described her problems in a very long thread where the issue appeared to be caused by interference picked up on the sensor cables. Of particular interest in that thread is where she shows the A/D converter waveforms where it can be seen that the actual A/D converter ramp is applied to the sensor cables. The rain and anemometer sensors are just magnetic reed switches with I believe one side connected directly to the "earth" (battery negative) of the transmitter. So there seems to be ample opportunity for impulse interference to disturb the operation of the A/D converter. I believe Gina's final solution was to use an optocoupler in the anemometer cable, but electric fences can be a particularly aggressive form of interference. With an understanding of the method of operation of the FO external unit it might be possible to devise some passive filtering (but I don't have any specific recommendations).

Cheers, Alan.

[ScottVan]

I did ground the two extra wires in the Cat 5 cable yesterday afternoon. I did not ground those to the negative batt connection. After I did this though my barometer is stuck at 29.91 (other nearby stations all are at 30.48-30.51) and my humidity seems to reading low compared to those other stations as well.

Isn't the barometer in the receiver unit? How could I have affected that? Temp is dead on compared to another thermometer here at the house. Wind reading seem good as well.

Do I need to do some type of hard reset on my base unit?

[philcdav]

Hi Scott.

To eliminate the CAT cable would be a good start, i think.

Just leave the remote unit unplugged from the PC, it doesny need it.

After a couple of days you could dump the data and have a look see.

Cumulus will easily catch up so no data will be lost.

I run an amateur radio station here with 100W RF HF and 50 W @ 2m/70 cm alongside the WX Rx unit and the aerials <5 m from the TX unit. Never had a problem BUT i get the odd spike too

So, my best guess is it will be bad connections caused by wind, rain whu? OR electric fences.

Interestingly, since putting WD40 into the external unit connections i get very few failures after nealy 3.5 years.

[Gina]

I'm back and still getting spikes from electric fences! If not ours our neighbour's. Not looked at the problem since I've been away (from this board) but may get back to it once my observatory is more nearly finished. I shall be moving all the outdoor station sensors and using a new tubular aluminium mast for the wind sensors, with the cable run inside it. The Tx unit will be further away from next doors fence. I'll report any progress.

[ScottVan]

Here's my day's temp graph. Some spiking but much much less than before the grounding.

Any others getting this good/bad readings with 15m of extension?

temp.jpg