A diary of the work carried out installing a new Davis Vantage Pro 2 to replace the previous Davis Weather Monitor II Station (Circa 1992) at the Ottery St Mary location.


Friday 5th November 2010

The Davis Vantage Pro 2 Plus has arrived today from Prodata Weather Systems with a Wireless Envoy, Weather Link Data logger and Daytime FARS Kit. I was pretty organised for once and get everything out of the box and spread out so I could see what I was looking at. I then opened all the individual boxes and collected the manuals for a little light reading later in the day.

In the evening, it was time to fit the Daytime FARS unit, and although there are fitting instructions, I could not really see from them what the plate order or why I had to remove the insulating ring (Which was pictured, but without the Temp/hum sensor on it) I have put it together how I think it should be, logically and subsequent testing has confirmed the fan operates when in sunlight.


Sat 6th November 2010

Unit testing today, with the ISS placed out in the garden on a chair, with the console in the Lounge and the Envoy in the office. First the console was configured which was fairly simple to do. Disappointed that the backlight does not have much intelligence. I always liked the Irox backlight that had a sensor for ambient lighting and switched it on when it went dark, although to be fair that's the only think I miss on the Irox !! Everything else on the console was straight forward to configure and I have the information I want displayed. I do find the console a little cluttered, especially between the Hum% and the Baro which from where I am sitting, looks like a long string of numbers. Operationally however it's working fine.

I then cabled up the Envoy to my EeePC testing machine. I use this machine for testing anything on Windows as it's my last remaining Windows machine and it's very portable too. I've loaded WeatherLink, Weather-Display and Cumulus for comparison and testing. So far, all work with the VP2, however I don't think that Cumulus can display the Solar and UV so probably won't go with this software, although I think the web part of Cumulus is great and looks very modern and easy to use for guests to a website. WL is ok, but still feels like a 1980's app although it does process the data and graph it well. I think I'll stick with WD though as it's been my software of choice for the last 5 years and the support is excellent from Brian Hamilton in NZ.

I also purchased a new 6ft x 4inch Fence post and Metpost for the garden for mounting the ISS and some 1-Wire waterproof boxes to. I'll see about knocking in that 2 ft spike into the lawn tomorrow so I can locate the ISS in free air whilst no being too far away for the anemometer cable to reach. I also want to monitor the rain gauge as I can't decide if to relocate the ISS Rain gauge to my original ground location that was chosen by the Envronment Agency as being the best location or use the elevated one on the ISS. One for science I think !!


Sun 7th November 2010

Been testing a raw install of WD today. After 5 years, I've bound to have some really odd configurations in the config files of WD so have been testing setting up from scratch and loading in old datafiles from the last 5 years. So far so good and I have loaded the whole of 2010 into WD. Also changed the colours and configured Solar and UV for display.

Got the MetPost into the ground, all but 2 inches of it, but it's going no further !! Put the post in and level'd it, drilled holes and bolted the ISS into the post, Raingauge Level, Slight tweek on the Solar and UK for centring and it's up !! I've put the Anemometer onto the mast, albeit a bit lower than it will be in final situation, however i have some re-wiring and tidying to do on the mast and whilst I'm testing, it'll be fine there. I also need to decide on the Anemometer Transmitter or the Long wire at the moment too.

Slight panic when I lost the Hum and Temp readings, however I'd knocked out the RJ-11 in the ISS SIM when passing the Wind Sensor RJ-11 through the grommit. All fixed now.

I've just noticed that the WMII is showing 9.4c and the DFARS on the VP2 is showing 8.2c. Amazing what the FARS can do to stabilise the actual air temp.


Tuesday 9th November 2010

Running a new station against your old one is always quite interesting. I've been comparing the various parameters against each other to see what the new set-up changes in the way of DFARS and the elevated Rain Bucket. Surprising news is that the rain tips have increased over the ground located TBR. I think this is because being in an elevated position as it is now above the fence line on both sides and thus the open space around the rain gauge has opened up to an extent that I am gathering more rain due to increased exposure. The DFARS evens out the temperature in situations where the screen is exposed to direct sunlight, so am pleased with the performance on that one. Solar and UV are doing well considering they are in an enclosed garden.


Sunday 12th December

Finally installed the metpost into the ground, trimmed the post to the correct length and checked the levels. All looking good !! Whilst the mast was down for the new anemometer, I decided to remove the rotator and 4m beam antenna and some of the cables that are not used. I've also removed the Satellite dish which was used with MythTV for experimental purposes a while ago and which I don't have the time to continue with at the moment. The Dish is finding a new home at a friends house for FreeSat.

Final work left to do before Jan 1st :

  • Build 1-wire sensors for 3 x Soil Temps, Lightning and Leaf Wetness and install.
  • Backup Datafiles from Live Server.
  • Completely remove WD.
  • Install latest, new compiler version WD.
  • Install Network USB Server Software and connect Startech USB IP server.
  • Configure and upload archive datafiles to WD.


Sunday 19th December

Spent all day sorting out the supplementary sensors for the new station, which are the 5cm and 30cm Soil, Grass temp and Leaf Wetness based on 1-wire. I went back to basics on the bus and re-cabled everything using CAT5E and RJ45 where possible. All boxes are now IP65 and use waterproof cable glands and the leaf wetness sensor has been re-mounted onto the front of a IP65 box. The ROM ID's were then noted for configuration later, and then soak tested.

All this was then tested using the Startech USB IP Server, meaning that USB devices are now available to all servers on the network. This is part of an experiment for High Availability projects, and latterly to use as a remote weather station sender over IP. This seems very stable and quick to set-up.

Remaining tasks are to finish the 1-wire Lightning Sensor refurbishment, and then mounting in the garden. This will be the first time we've had a Lightning sensor since October 2008 when the last one died as a result of the Ottery Storm !! It was planned to build a Blitzorung Directional Sensor for Lightning, however time has not permitted. As much as I would like a Boltek LD-250, we do not get enough lightning here to use it properly.


Monday 27th December

After a full backup and cleardown of the weather server, Weather Display has been re-installed and configured for the Davis Vantage Pro 2. The weather station is now live.

Further work required for :

  • Datalogger testing
  • 1-wire configuration
  • Configuration of Custom Tags, and website datafiles

Further testing revealed that the new compiler version of WD seems to be having issues within my Virtual Machine. In light of this, a new version of WD is now running on my Test Asus EeePC and integrated with the original backend webserver. We will be re-starting the Forecasting routines and Twitter as and when the integration is completed.

Monday 3rd January 2011

After a whole load of fault finding, it looks like the VM cluster cannot cope with the increased CPU load from the increased datafeed and the new compiler version of WD which seems to use more CPU than the previous versions.

We've started building a replacement physical server for all weather operations, which should be completed by the end of the week. Until then there will be a limited service from the site such as Twitter, Email Forecasts and Site Forecasts.


Monday 10th January 2011

New Intel Core 2 Duo server is now live handling Weather Display, WXSIM and all weather related operations. WeatherLink has also been installed to manage the Davis Envoy when needed although we will be using it as a secondary weather recording repository on an ad-hoc basis.


Saturday 15th January 2011

Weather Station now completed, with 1-wire and Davis sensors reporting to the station. Wiring completed, with some small items to complete when weather permits.


January 1st 2012

Davis re-located at The Old School, Combe Raleigh.