Cumulus Support

Well thank you very much for the warm welcome Gina!

Lol I already need to work out how to stop it reading around 1.5°C over the top when in direct sunlight. Given it is mounted 10ft high it may be tricky I feel

I had exactly the same problem! The suggested arrangement of mounting all the outdoor sensors next to each other is far from ideal. What many of us have done is to separate them and mount the temperature/transmitter unit separately at about 5ft and in a decent Stevenson screen. The supplied screen is next to useless. Plenty of examples in the Homebuilt forum (including mine). There are problems with the rain gauge where they suggest too. Again the Homebuilt forum has examples of mods. I think there may be some in this forum too. My weather website (see sig) and others also show what people have done to improve things.

nking wrote:I wouldn't be too quick to discount the 1081 as some owners have had issues with the 1080 clock facility and I have personally found the 1081 clock to keep very good time

I've had the 1080 from Maplins for 6 months or so now (£76 inc delivery at the time?). I can confirm a) they only had the 1080 not the 1081 when I ordered on the web b) it keeps excellent time c) I've had no real problems until this month when I've had 2 cases of lost connection, maybe due to batteries (see other thread).

Just purchased a WH1081 from Maplins this afternoon. All set up but just need to download cumulus and get to grips with it all. There will probably be a number of basic/silly questions posted in the not to far distant future asking for help!

Yesterday I brought my second station from Maplins Console for running on another PC with C2 software so that I can keep C1 working on my other Pc and keep me website up to date - rest can be for spares

Welcome to the Forum Kent Weather, I hope you find it very interesting as I do.

Just set up my new (second) console on my XP box, installed and ran Easy Weather and now running Cumulus 1. Just need to sort out why it doesn't seem to be uploading to my Cumulus web site. Once I've got Cumulus 1 working properly I'll give C2 a try Probably tomorrow.

Hi, my first post so a quick Hi

I bought the Maplins offer weather station and Im pretty pleased with it, my main hobby is astronomy so I was looking at a way of logging temperature and humidity which has a direct bearing on how well you can see & image things.

I managed yesterday to get my weather station online and the website is in my sig.

Welcome to the forum upgrader, and another astronomer joins the fold It seems astronomers are also obsessed with the weather

mcrossley wrote:Welcome to the forum upgrader, and another astronomer joins the fold It seems astronomers are also obsessed with the weather

It goes hand in hand
I also wanted a accurate record of my local temperature for using with my dark imaging library, if your into astrophotography you will know what I'm on about.

I now have my second console wired up to my XP box and running Cumulus 1. I found the reason it wasn't uploading - XP's firewall was blocking it. Once I told it to unblock Cumulus, it worked Just need to set the webcam address in the menu link and fix a few other little things.

http://ginad.org.uk/weather/ for Cumulus version of Blackdown Hills Weather

hi folks another newbie here,ive also just bought this unit from Maplins.must say i'm very happy with my purchase,i have it sited on a 10 ft pole in the garden.i want to iron out any problems before it goes up on it's permanent position.
great site this and very helpful thanks bob.

Hi,

First post here! I have been running one of these units for over a month now, and am generally pleased with it. I could tell you tales of the fun I've had siting the sensors and calbrating so that I can maintain continuity with my old temperature records going back many years, bu that would be a long story.

One thing it did notice though is that I have occasionally had a wildly inaccurate rading, no doubt corrupt data. A couple of days ago I got a temperature reading of -24C, which Cumulus duly saw as a new all time low temperature record. This was just one reading, and the reading either side where over 30 degrees warmer.

It seems to me that Cumulus could be made to eliminate such readings. I would suggest that any reading which is wildly divergent from the proceding reading, or plainly ridiculous should be ignored. This could be done on two vases

1. If the reading is well outside the expected range.
2. If the rate of change is excessive (30C in 10 minutes for example).

As it is, I was able to edit the appropriate files manually to remove the errant reading. But it should be possible to do this automatically.

BTW you can find my station at http://homepage.ntlworld.com/m.dixon4/Cumulus/index.htm

Martin

Martin

MartinDn wrote:It seems to me that Cumulus could be made to eliminate such readings. I would suggest that any reading which is wildly divergent from the proceding reading, or plainly ridiculous should be ignored. This could be done on two vases

1. If the reading is well outside the expected range.

See 'EWTempdiff' in the cumulus.ini file - http://wiki.sandaysoft.com/a/Cumulus.ini

2. If the rate of change is excessive (30C in 10 minutes for example).

Cumulus does do a lot of filtering of bad data already, but I do intend to try to add further checks at some point.

steve wrote:Cumulus does do a lot of filtering of bad data already, but I do intend to try to add further checks at some point.

Steve, something I've found in my experiments is that the station will sometimes return nonsense data, presumably when a block of memory is being updated by the station at the same time as being read. I now read all blocks twice (or more) and only use data that is consistent between two readings. I don't put in a time delay between readings - the slowness of USB seems adequate. (Although as my software is written in Python and running on a very low power computer it might be a bit slow anyway.)

Another less common source of wrong data is that the station updates its 'current memory address' pointer before writing data to the new address. Hence it is possible to read old data (two weeks or more old) from the new address. This data is usually credible though - no -24C temperatures.

jim-easterbrook wrote:Steve, something I've found in my experiments is that the station will sometimes return nonsense data, presumably when a block of memory is being updated by the station at the same time as being read. I now read all blocks twice (or more) and only use data that is consistent between two readings. I don't put in a time delay between readings - the slowness of USB seems adequate. (Although as my software is written in Python and running on a very low power computer it might be a bit slow anyway.)

It's interesting that you say that, because I actually saw some official (?) documentation from Fine Offset somewhere, and it did seem to be suggesting that the station might not be ready so a retry may be necessary. I couldn't work out how you would do that (other than by the method you suggest of comparing the data). I'll see if I can find the documentation. The existing filtering in Cumulus will effectively already be doing this to a certain extent though.

Another less common source of wrong data is that the station updates its 'current memory address' pointer before writing data to the new address. Hence it is possible to read old data (two weeks or more old) from the new address.

Yes, I spotted that one a while ago - I now skip a read when the address changes.