WS 2300 rainfall counter

[maylandmanta]

Hi
Having downloaded the latest version of Cumulus I have noticed that the rainfall counter keeps wanting to load the yesterday,month and yearly totals up as 174mm plus the actual rainfall levels for that day.I have even deleted all the data from the WS 2300`s memory and any history files saved to ensure clean data is received by Cumulus but the next time I opened the programme the same thing happened.I also noticed that briefly the figures were double this before reverting back to the 170`s.
Is this a bug in the programme or could it be something my end.

Cheers
Steve

[steve]

It's not clear to me how long you have been running Cumulus. The monthly and yearly totals are calculated at start up from the daily totals in dayfile.txt, but if you only started using Cumulus today you won't have any entries in that - in which case the total for today is probably wrong.

Cumulus calculates today's total by subtracting the start of day total from the current total. It stores the start of day total in the today.ini file at the start of day. If it got an incorrect value for that from the station it will therefore get today's total wrong. Have a look in today.ini and see what the Start= value is and how that compares with the current total in the Jan09log.txt file (the format is in the help).

What are the 'history files' that you say you deleted? When you say 'latest version of Cumulus' do you mean 1.8.4?

Steve

[maylandmanta]

Hi Steve
With reference to my post reguarding the rainfall counter, I am now running version 1.8.5 build 674.The monthly and yearly rainfall figures still keep showing in the 170`s at start up, but only briefly before reverting to the correct figures.However if I uncheck the data logger box this problem no longer occours.Is this normal with the WS2300 and Cumblus ?
Other than this minor hickup I am finding this a great piece of software so many thanks for the time and effort you are putting into this.

Reguards
Steve C.

[steve]

Hmmm... sounds like a bug to me, possibly similar to the one I fixed in 674. Are you using a rainfall multiplier?

I'll have a look at the code.

Thanks,
Steve

[maylandmanta]

Steve
Thanks for the quick reply.In answer to your question about the multiplier ,it is set to 1.00 as it set itself up on installation which as I understand it should have no effect.

Thanks
Steve C.

[steve]

Yes, that's right. I'm wondering if your startup values are just hanging around from before you installed the latest beta. I do understand you correctly - it's effectively now not a serious problem because it goes away a few seconds after startup (on the first read of the rain counter, I guess)?

It may sort itself out tomorrow.

Steve

[maylandmanta]

Steve
Will check again tommorow and keep you posted.The only problem it causes is that I have to keep ammending the all time records as it throws the false figures into there aswell

Steve C.

[steve]

Ah, I see. Could you send me a file from the Diags folder for a run where the problem occurs, please? And your today.ini and dayfile.txt files.

Thanks,
Steve

[maylandmanta]

Hi Steve
Thanks for the reply via e-mail.What I have decided to do is wipe Cumuls off computer and reinstall.I have set up 2 instances of the program - one with the data logger checked and one without.Will run for a few days and let you know the outcome.
Cheers
Steve C.

[maylandmanta]

Steve
Since my last post earlier today I have been randomly switching both instances of Cumulus that I set up on and off and can confirm that these false figures only occour at start up when the data logger box is checked.
I dont know if i`m being totally thick here but Ihave played with the rainfall multiplier figure as below :-

Changed multiplier to 0.01(smallest number will allow to input) and my false figure at start up changes from 173.5 to 1.7 mm (Understandably as moving the decimal point).If it were possible to set the multiplier to 0.00 would it not mean that these figures would be cancelled out .
As I say may be missing the plot somewhere here as I am getting confused with all these figures but thought i would mention it


Steve C.

[steve]

Is the figure that's being added in (without the multiplier) the 'total count' from the station? It's the number after the pressure in jan09log.txt. Actually, I think yours was around 189, so that's not it.

If you set the multiplier to zero, you will never have any rain at all. And Cumulus will crash.

I think I will have to try to pretend to be a WS2300 and step through the code in the debugger with some sample data.

Steve

[steve]

I see what the problem is now, from the log file you sent me. I don't know why I didn't notice it before. The total rain counter being read from the station is different depending on whether it's being read from the history data or during normal running. The history data value is in the region of 190 and the normal running one about 15. So the rain calculation comes out differently in the two cases.

The difference appears to be that one is the counter in the station, and the other in the rain unit itself. It looks like one has been reset at some point but not the other, so they are out of step. I'll have to look at which memory locations I'm reading, and which ones I should be reading, to fix this. The WS2300 has a number of slightly different 'total' counters!

Steve

[w5afw]

FYI---> My ws-2315 is acting the same way. Any fix yet?

[steve]

Yes, sorry, I normally try to post to the relevant thread when I fix a problem, but this one I must have overlooked. It was fixed about a month ago, in build 688.

Steve

[w5afw]

Steve, I am using ver 726 and the symptoms are the same. It could be a different cause but acting the same. I am trying to correct with the .001 multiplier trick but calibration is impossible. The guage and wiring seem to be fine, but the receiver is out of reach so I am not sure what it is doing. Suggestions?