The LostContact setting

[daj]

Steve

I wonder if there is some further work you could do with the <#sensorcontactlost> webtag?

At the moment this is set to 1 if the FineOffset drops communication with the outdoor sensors. Would it be possible to then stop Cumulus processing? Currently Cumulus keeps uploading data to your website and also produces the realtime.txt file, both of which are now irrelevant.

Over the last few days my Station has lost contact for a few hours at a time and I am still trying to figure it out, but I am also spending time cleaning up data.

just a thought

thanks

[steve]

Would it be possible to then stop Cumulus processing? Currently Cumulus keeps uploading data to your website and also produces the realtime.txt file, both of which are now irrelevant.

Yes - but I'll have to have a look at how much work that is. It may just be a case of stopping the timers, but there is probably other stuff that would need special attention, like wind speed averages. I'd have to spend some time looking at it.

[daj]

Ok -- Please don't put too much effort into it. A nice to have but not if too much work.

I was trying to think of other ways to handle it but by the time Cumulus has issued the lost sensor flag it has also processed and uploaded data. It also uses the date and time of processing as part of its web tags, so I can't do a date/time check as that is changed by Cumulus too

[steve]

It's really just another example of the 'carry on regardless' way that Cumulus works. I didn't have any problems like this with the two stations that I originally wrote Cumulus for (Oregon Scientific and Davis) so I never thought to put in anything to cope with a long-term failure of the data feed, assuming that if the data went missing, it would be back again in a short period of time, so it didn't matter if Cumulus used the last good data repeatedly for a short period.

You say you cleaned up your data - what did you have to do? Delete the log entries?

[daj]

You say you cleaned up your data - what did you have to do? Delete the log entries?

Initially I was editing the monthly log file, and the today.ini to remove spurious data, however I changed my approach....

Once I realised the station lost contact (Got Toolbox to email me) I stopped Cumulus and waited until I had a connection again (first time it was about 8 hours, then 4 hours, yesterday about 40 minutes). I used EW to see if it had a connection yet

I then rolled back to an earlier Cumulus backup (I take one per hour with the Toolbox) and restarted Cumulus and all was fine.

Not sure why it is dropping communication -- can't be the batteries as it does eventually come back

[steve]

Initially I was editing the monthly log file, and the today.ini to remove spurious data,

The way it's intended to work is that it just keeps logging the same (external) values as the last set it managed to read. The internal values (including pressure etc) should be OK. So if I changed it to do something different, the only alternative I see would be to stop logging altogether. You then lose the valid internal data, and if you look at a graph, the external data is a flat line just the same as it would be if there were no entries in the log. Unless I'm misunderstanding what actually happens when sensor contact is lost.

Not sure why it is dropping communication -- can't be the batteries as it does eventually come back

Mine does it frequently, but the way I know is that I see the console values go to dashes, as I don't have Cumulus running with the FO station all the time. Although, now I think about it, I haven't noticed it do it for a while - it seems to prefer the colder weather for some odd reason.

[Charlie]

The only time I see mine lose contact is if I'm tweaking and get my body between the sensor unit and the display unit. I guess when it comes to RF I make a better door than a window! Any chance the warmer weather equals more opportunities for warm boddies to get in the path?

(I'm shooting about 10 meters indoors, and maybe another 30 outside. There are about 4-5 walls to get through. I can see where adding some biomass in the path might make the link tenuous).

[daj]

For me, my sensors are about 20 meters outside of the house and have a window and wall to come through inside. This has ran almost without issue for the last two years but the last few days have been really bad with several hours of nothing. Batteries only a few months old.

Temperature has been around -10C. Oddly on one occasion it was not working at +1 but suddenly started getting data when it got to -8C. It's currently 1C and it has been working great all day. No one can walk in front of the outdoor transmitter as it is up a pole. No one in the house when it fails either

I'm going to try and change all the batteries this weekend, weather permitting. Although we have had 30cm of snow overnight so nothing is easy at the moment

[EvilV]

You probably know this, but the RF output of a circuit is very much dependent on the voltage supplied to it. There is a remarkable drop if you lower the voltage by not very much. It's probably something to do with the Ohms Law power equation. In colder weather the battery voltage will fall somewhat. I'm not sure by how much, but any drop of voltage will have a large effect on an already low level signal struggling to get through the walls of a building.

Yesterday, I was messing about with a low power amateur radio transmitter, and I noticed a huge reduction in power when I changed the power supply voltage from 12 volts to 9 volts. It wasn't a 25% reduction, more like 80%, judging by eye-balling a power meter. Take that as a rough estimate, but you can see what might happen if cold weather reduced your weather station transmitter voltage from 3 volts to 2.5. I'm not sure how far down the voltage will go in cold weather, but there will be tables available from manufacturers, I expect.

EDIT:

I just did some searching on the Energizer website. Ordinary alkaline batteries suffer a very large voltage drop when delivering 250MA continuously at low temperatures. At -20C after doing this for about 50 minutes the voltage available drops to about 1.1. Further periods of that current draw and temperature drops the voltage after one hour to 0.8V. Even at 0C, the voltage drops to 1.25 after an hour. It would have been useful to see how the much lower current draw of a weather station affects the voltage, but the graphs don't cover that. My son just returned from Canada skiing for two months and he said his alkaline battery powered camera just refused to work on the mountains. The temperature was often around -15 to -20C, but the camera wouldn't work at all. I remember my camera packing up while skiing in the Alps at about -10, unless I kept it close to my body before use. Once again, cameras draw more power than a weather station.

We could test the performance by opening the weather station transmitter case if it is accessible and measuring the battery voltage on cold and warm days. That would settle the matter.

[gemini06720]

I'm going to try and change all the batteries this weekend, weather permitting. Although we have had 30cm of snow overnight so nothing is easy at the moment

David, how about a project for the upcoming summer (in 10 thousand years, by the look of it )...

...Installing a solar panel and a battery which would power your weather station...

The Davis Vantage Pro2 ISS (integrated sensor suite) is using a 3-volt lithium cell (C type) and a 0.5 watts solar panel / current drawn: 0.14 mA (average), 30 mA (peak) / battery life: 8 months without sunlight - greater than 2 years depending on solar charging.

[RCE]

Apart from the occasional lost sensor warning, the only time I have had an issue is when I left a bluetooth GPS turned on approx 2m from the main control unit overnight, then I had a log full of lost sensor events.

Strange thing is I have a bluetooth dongle in the pc that the weather station connects to that is always on (used to sense when we are home by detecting the bluetooth on our phones) and this never seems to cause a problem.

[Gina]

The Davis Vantage Pro2 ISS (integrated sensor suite) is using a 3-volt lithium cell (C type) and a 0.5 watts solar panel / current drawn: 0.14 mA (average), 30 mA (peak) / battery life: 8 months without sunlight - greater than 2 years depending on solar charging.

Are you sure it's a Lithium cell? Lithium cells are NOT rechargeable and attempting to recharge them can cause them to explode according to Energizer.

[steve]

The Davis Vantage Pro2 ISS (integrated sensor suite) is using a 3-volt lithium cell (C type) and a 0.5 watts solar panel / current drawn: 0.14 mA (average), 30 mA (peak) / battery life: 8 months without sunlight - greater than 2 years depending on solar charging.

Are you sure it's a Lithium cell? Lithium cells are NOT rechargeable and attempting to recharge them can cause them to explode according to Energizer.

The solar panel charges a super capacitor; the lithium cell takes over if the super capacitor runs out of charge.

[Gina]

Ah. I see. Thank you for the explanation I was afraid someone might try to recharge a lithium cell with disastrous results.

[gemini06720]

The Davis Vantage Pro2 ISS (integrated sensor suite) is using a 3-volt lithium cell (C type) and a 0.5 watts solar panel / current drawn: 0.14 mA (average), 30 mA (peak) / battery life: 8 months without sunlight - greater than 2 years depending on solar charging.

Are you sure it's a Lithium cell? Lithium cells are NOT rechargeable and attempting to recharge them can cause them to explode according to Energizer.

The solar panel charges a super capacitor; the lithium cell takes over if the super capacitor runs out of charge.

Thank you Steve for adding more precision to my suggestion to David.

Gina, I cut and copied the quoted information from the Davis Instruments documentation ... and I do not remember writing anything about charging the lithium battery... I guess Davis Instruments should have been more precise in their statement, such as '...greater than 2 years depending on solar intensity and the charging of the super capacitor...'