Cumulus Support

I am researching the practicalities of installing a weather station at a local RC model aircraft flying field. Mainly to provide current wind and temperature status on the clubs website for pilots to know what the weather is doing at the site before setting off from home.
The site doesn't have electricity or access to a phone line, so I know that power needs to come from a battery/solar/wind power combination and to upload data to the net, a mobile dongle will need to be used, although 3G access is patchy but there is solid GPRS data connectivity.
Additionally, the site is remote with no neighbours, so security could be an issue.
Given all of the above, does anyone have any knowledge/experience of similar setups that they could share to enable me to make the right decisions of which routes to explore?

I will be using my redundant OS WMR 200 weather station, and most likely a netbook with a 3G dongle plugged in for internet access initially. So will Cumulus manage it's uploads given the relatively slow internet access? If/when a Linux Cumulus version is released, I am sure a Raspberry Pi version or similar will quickly follow which would vastly reduce power requirements compared with a netbook PC, but what power options could people suggest based on a netbook?
And with regard to security, has anyone "camouflaged" their external instruments to make them less obvious to prying eyes but still retaining their effectiveness?

Have a word with Stuart2007.

He installed remote kit on a windmill https://cumulus.hosiene.co.uk/viewtopic.php?f=4&t=8966 although he did have power.

Hi,

Sorry, I have no experience of remote/off-grid stations, but perhaps I can do some of the "maths" for you.

Firstly, I would hope that the weather sensors/station itself wouldn't be a significant power or security issue. Presumably they can be fitted at the top of a secure (inaccessible) mast with little apparent (or even actual) value? I don't know about the WMR 200, but power consumption should be negligible, with a wireless link to a hidden receiver in a "secure" location. A Fine Offset (Maplin) offers that for £50 (currently), with the most expensive "spare" sensor (the temperature-transmitter) £15. I don't know how you'd camouflage the sensors, but you could put a "fake" yagi antenna on the mast, to imply that the "receiver" is miles away.

However, running without mains power looks to be a major issue. A good netbook battery is rated at about 50 watt-hours and may run for 10 hours, so let's assume you need 5 watts (with 24/7 operation). A Raspberry Pi should be better, but I guess you may still need a watt or two.

So the netbook may require 120 watt-hours per day, or say 10 Amp-hours from a 12 volt car battery. I believe a typical modern car battery is about 40 Amp-hours, so even a large, new one is likely to struggle to run the netbook for a week.

As for a solar panel, Cumulus will show you that peak sunlight is about 800 watts/square metre in Summer, down to around 400 watts in Winter. But solar (PV) panels are only about 15% efficient, so even on a reasonably "sunny" day in winter you may get an average of 40 watts for 6 hours or 240 watt-hours. So a half-square-metre PV panel appears to be the minimum requirement. That is likely to have significant cost/value and may not be easy to "camouflage". Then you will probably still need an alternative weekly "recharging" strategy for long periods when light levels are low.

IMHO a wind generator may be even worse. Check your own wind records, bearing in mind that windmills generate no useful energy below about 6 or even 8 mph (that's a wind run of around 200 miles/day). Furthermore, "useful" amounts of energy may need much higher windspeeds, but then a point is reached where automatic "safety" shutdown systems are required to prevent mechanical damage in high windpeeds. And I don't know how you camouflage a windmill !

Cheers, Alan.

Thanks Beteljuice, will have a look at that topic.

Thanks for your input Alan. I'd been contemplating a couple of 110mah leisure batteries to try and achieve at least a weeks worth of power. Theoretically there is someone at the site every week, so the batteries could either be recharged by a generator whilst someone is on site or they get swapped out each week by fresh batteries and recharged at home. But the "forgot to" or "can't be bothered" factor may well creep in hence the reason for thinking solar recharge.

I've been trying to understand solar panels and recharging battery's from the UK's usual pitiful sunlight power and was beginning to draw the conclusion it couldn't be done cheaply or without a big bank of solar panels that shouted "steal me"!
To add to the issues, there is a hut (think prefab portable office) with a flat roof, but it is overshadowed by trees on the southerly side, basically blocking any useful sunlight.

Hmmm, camouflaging a windmill? Now there's a challenge. Make it look like a tree?

I've had solar-powered stations with GPRS comms operating in 2 or 3 different configurations, but unfortunately it's not easy to do simply and cheaply, especially if it's a site that might be prone to vandalism/theft. One of the problems is the power that the 3G modem uses, which is nontrivial if it's powered up continuously.

To go down your present line of thinking, one solution - if feasible - might simply be a new hut in a more sunny position. A small new wooden shed isn't hugely expensive and its pitched roof (obviously oriented towards the south if possible) can provide a good and less obtrusive position for a large solar panel.

Another consideration is how often you want to see data updates. If you're content with eg 5 or 15-minute updates then the best solution is to power down the 3G modem other than when you're actively transmitting an update. But that's easier said than done unless you have some experience in designing the system.

But if this sort of system would be acceptable then you can base its energy budget on a Davis Connect unit - this uses a 5W solar panel (relatively small for this sort of device) and a 12Ah battery and can transmit year-round in England with updates down to every 5 minutes. It wouldn't be running Cumulus because there isn't the computing power available to do so (because of energy considerations) - it's more Arduino than even an RPi, but it is able to upload data to the weatherlink.com network. I know a Connect is going to be too costly for your application (it would be about the £1K mark), but just an illustration of how the energy budget works out. I don't know what Chris at http://www.weatherelement.com is planning to sell his units for - they might well be cheaper than the Davis option, but would need importing from the US.

Thanks for your input John.
Those Davis units aren't in the budget unfortunately. The site is bordered by trees on the southern edge and this would be the only place where another hut could go, so unfortunately that doesn't achieve anything either.

With regards to internet updates, for the purposes of RC flying we only need updates during daylight hours and even then more like 9.00am and 9.00pm (when flying is allowed).
Obviously I can't schedule updates as such with Cumulus, but I can probably frig a schedule for when the internet is available, and just let the Cumulus errors build up in between.

My main sticking point is definitely how I power a netbook for running Cumulus. I suspect my cheapest option is going to be a bank of leisure/car batteries that get charged up weekly. Or, has anyone got the instructables for a hydrogen power station in a matchbox?