Sandaysoft

I've just found a bug in Cumulus in the downloading of the history data from a WS23xx which is causing this problem

I need to check that I haven't made the same stupid error elsewhere in the WS23xx code, then I'll have a fixed version available which should sort out your problem, and will help anyone else who is seeing incorrect data being downloaded from the logger.

I'm really sorry about this.

Steve

New build available with bug fixed, hopefully:

http://www.nybbles.co.uk/downloads/beta185/CumulusSetup.exe

I've worked out that I introduced the bug in build 688.

Steve

WSH

I don't know how you keep track of all the tweaks you must try out !

Have a drink on everyone !!!

_________________

......................Imagine, what you will KNOW tomorrow !

Hi Steve

Brilliant, many thanks. Just one small point though, the Dew Point graph is still behaving as before.

See http://www.sedgleymanor.com/wx/trends.htm

_________________
Website : http://www.sedgleymanor.com/wx
on the Indian Ocean coast of Western Australia (down the bottom left-hand corner of Oz)

Those changes in the dew point are all from data downloaded at start up from the logger, I assume? Cumulus calculates it from the temp and humidity, and you have no odd changes in those, so that's strange.

Can you let me have the diags file for that start up, and your Feb09log.txt, please?

Steve

See attached...

Thanks

Ian

_________________
Website : http://www.sedgleymanor.com/wx
on the Indian Ocean coast of Western Australia (down the bottom left-hand corner of Oz)

Have you got Cumulus set to calculate dew point, or are you using the station values?

Steve

I'm using the Station

_________________
Website : http://www.sedgleymanor.com/wx
on the Indian Ocean coast of Western Australia (down the bottom left-hand corner of Oz)

The higher figures are the ones which came from the history download, e.g. 0750-081 and 0910-1110, and those dewpoints are clearly wrong, eg.

09:00 (in real time) temp=19.2, hum=53, dp=9.4 - which is correct
09:10 (from logger) temp 19.5, hum=49, dp=14.6 - which is wrong, it should be about 8.5

Cumulus is clearly reading the correct history entries and timestamping them correctly, because all of the other data follows on quite nicely, as you can see from the graphs. So I don't know what to make of it. The dew point from the history entries is downloaded in the same block as all the other data. It's unlikely to be a corruption in the data transfer, because the values are all consistent. So I don't know what to suggest, other than seeing what Heavyweather displays for those periods, and/or setting the option to get Cumulus to calculate the dew points.

Steve

Thanks again for all your work Steve.

I switched to Cumulus's calculation and exited it at about 2300hrs and restarted it about 30 mins later. It now appears to be working correctly. I'll see what happens when I boot up in the morning.

Ian

_________________
Website : http://www.sedgleymanor.com/wx
on the Indian Ocean coast of Western Australia (down the bottom left-hand corner of Oz)

Ok, thanks.

Steve

Good Morning Steve

Cumulus' Dew Point calc seems to be working perfectly - brilliant

Many thanks once again

Ian

_________________
Website : http://www.sedgleymanor.com/wx
on the Indian Ocean coast of Western Australia (down the bottom left-hand corner of Oz)

Experimented with switching back to Station DewPoint calculations and it goes haywire again. Cumulus calc is the way to go!

_________________
Website : http://www.sedgleymanor.com/wx
on the Indian Ocean coast of Western Australia (down the bottom left-hand corner of Oz)

I don't really understand why the station figures go wrong like that. But it's just as well there's the Cumulus-calculated alternative. I imagine the numbers come out the same as the station anyway. I suppose you could compare the Cumulus figure with the console, just out of interest.

Steve