Page 1 of 2

Cumulus 1.9.4 beta build 1071 - UV data with a decimal point

Posted: Fri 03 May 2013 3:58 am
by Werk_AG

Code: Select all

Cumulus 1.9.4 beta build 1071 is available.

•For 'easyweather.dat' input method: Read UV and Lux values from easyweather.dat file if 'extra sensors' selected.
Hi Steve.

Many thanks for having extended suport for this two new data fields in 'easyweather.dat' input method.
Any chance of a future version, the UV data field accept values ​​with a decimal point? Like 1.9, 3.4 etc etc

Regards

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Fri 03 May 2013 6:36 am
by steve
I made it an integer because that's what Fine Offset stations supply. I suppose I could look at making it accept a floating point value, for those people using it for other purposes.

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Fri 03 May 2013 11:21 pm
by Werk_AG
steve wrote:I made it an integer because that's what Fine Offset stations supply..
Yes I know. We talked about it ...
It would be good that the field could also accept data with a decimal point. The resulting graphs would look much better.

Thanks for your attention

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Sat 04 May 2013 7:12 am
by Werk_AG
Also I noticed what may be a small bug.
Using easyweather.dat as input method, the names of the data fields (Light, Solar Rad, UV) in the Solar Section, are only shown correctly if in the station setup menu I activate the option "use B/L". If this option is not active, the fileds are show with the names Temp, Dew p and Hum, but the values are displayed correctly. Only the names change!

Regards

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Sat 04 May 2013 7:51 am
by steve
Ah, yes, the list of station types known to have optional solar sensors doesn't currently contain the easyweather.dat type; I can easily fix that. There may be other similar issues scattered about the code. As the easyweather.dat input method is deprecated, I have tended not to consider it when doing any changes.

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Sun 05 May 2013 3:41 am
by Werk_AG
I understand that easyweather.dat input method is deprecated, almost all recent weather stations have PC connection. Anyway, I hope at least this option remains as it is the only way possible for all who use homemade stations, based on microcontrollers like Arduino.
Currently from my original station (Auriol H13726), I'm just using the rain gauge, and anemometer (data collected also by an Arduino with a RF receiver), all other data come from sensors connected to Arduinos.

I am currently trying to build a solar radiation sensor based on studies by Chuck Wright: Measuring Solar Radiation
Preliminary tests are promising, but none of this would be any interested if I did not have the support of Cumulus to use these data.
I think now you understand why I was so pleased with the fact that you have recently extended the possibility of easyweather.dat file format.

Thank you, thank you so much

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Sun 05 May 2013 7:56 am
by steve
Werk_AG wrote:Anyway, I hope at least this option remains as it is the only way possible for all who use homemade stations, based on microcontrollers like Arduino.
I won't be taking it away; I was going to some time ago, when it was only there as the 'old' way of using Fine Offset stations, but I know that quite a few people now use it as a way of using Cumulus with unsupported stations. There are alternative approaches, of course; I know of at least two people who have made their station emulate one of the supported stations.

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Sun 05 May 2013 8:34 pm
by AllyCat
steve wrote:I know of at least two people who have made their station emulate one of the supported stations.
Hi,

Yes, that's what I hope to do (eventually) since I now have a number of additional or duplicate sensors (Solar, Hydreon, Greenhouse temperature, etc.). I don't run my PC(s) 24/7 so find the downloading facility of Cumulus very useful. Ideally, I'd like to use a serial interface for hardware/software simplicity and to permit a second instance of Cumulus to run, but which is the "best" station to emulate?

The "obvious" protocol appears to be Davis' due to the availability of their "open" documentation and the range of sensor types supported. However, there have been some comments that their protocol is complex to implement. What else is available? The instromet looks interesting, but is a specification available or was the data format provided in confidence?

Thanks, Alan.

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Mon 06 May 2013 2:28 am
by Werk_AG
steve wrote:There are alternative approaches, of course; I know of at least two people who have made their station emulate one of the supported stations.
No doubt that emulate one of the existing stations would be the best approache. Possibly one of the Fine Offset models. I've tried searched the internet about it, but found nothing ...
I think I still do not have knowledge to do it without help ... I started just a few months ago trying to learn C + + to play with Arduinos.

If someone could post some part of the code, it would be a great help ...

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Mon 06 May 2013 8:48 am
by steve
Charlie has built his own station which emulates the Fine Offset interface, see this thread: https://cumulus.hosiene.co.uk/viewtopic.php?f=16&t=7390

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Tue 07 May 2013 12:01 am
by Werk_AG
I will try to talk with Charlie. Thanks for the info...

Thank you also for the adjustments in last Beta Version. Already instaled... seems fine.

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Tue 07 May 2013 10:42 am
by AllyCat
steve wrote:Charlie has built his own station which emulates the Fine Offset interface, see this thread: https://cumulus.hosiene.co.uk/viewtopic.php?f=16&t=7390
Hi,

In the first post of that thread Charlie says:

"USB is way harder than it looks. So a solid 3 months later ......."

which persuades me (even more) that I want to try to use a ("RS232") serial interface. Any ideas which stations can use such an interface (and ideally which might be easiest to emulate)?

Thanks, Alan.

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Tue 07 May 2013 11:39 am
by steve
AllyCat wrote:which persuades me (even more) that I want to try to use a ("RS232") serial interface. Any ideas which stations can use such an interface (and ideally which might be easiest to emulate)?
Someone else did an emulation of the older Oregon Scientific stations, which have a serial interface, but I think if I were to do it I'd be looking to emulate a Davis, for which the protocol is officially documented - http://davisnet.com/support/weather/dow ... wareID=235

Instromet stations also have a serial interface, so that is now a possibility.

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Tue 07 May 2013 2:16 pm
by AllyCat
Hi,

Thanks Steve. I suspect that you missed my post above (because another was added before you were able to reply).
AllyCat wrote:..........

The "obvious" protocol appears to be Davis' due to the availability of their "open" documentation and the range of sensor types supported. However, there have been some comments that their protocol is complex to implement. What else is available? The instromet looks interesting, but is a specification available or was the data format provided in confidence?
Does anyone know which parts of the Davis protocol are "essential" (or is it all done by the Davis .DLL)? Is there any protocol documentation available for the Instromet?

Cheers, Alan.

Re: Cumulus 1.9.4 beta build 1071 - UV data with a decimal p

Posted: Tue 07 May 2013 2:28 pm
by steve
AllyCat wrote:Thanks Steve. I suspect that you missed my post above (because another was added before you were able to reply).
Yes, I did.
The instromet looks interesting, but is a specification available or was the data format provided in confidence?
There are no copyright notices on the document I've been given, but I don't think it would be correct for me to pass it on. It's quite a simple text-based protocol.
Does anyone know which parts of the Davis protocol are "essential" (or is it all done by the Davis .DLL)?
That would be easier for me to answer if it weren't for the fact that I'm using the Davis DLL in Cumulus, and I don't know for certain which commands are invoked for the various functions that I call. In my 'other' versions of Cumulus which don't use the DLL, all I'm currently using are the LOOP and DMPAFT commands, for live data and archive data respectively, but I'll almost certainly end up using more than just those.