Page 1 of 1

All data flat-lined

Posted: Sat 13 Mar 2010 8:56 am
by christer
Hi Steve.
This morning I found flat figures in the graphs. At around 21:50 yesterday something happend. Since also the inside temp is flat it cannot be the RF from the outdoor sensor.
So I started debug.log and saved it to my PC.
After that I stopped and restarted Cumulus and after that the graphs showed new data. I don´t have access to the console nor the USB cable since the station and PC is located in a remote place. I use VNC to connect the PC via Internet.

The attached zip containing debug.log, today.ini, bugreport.txt, cumulus.008 and 2 pictures.
I´m using Cumulus V1.8.8.855.

Regards
Christer

Re: New Weirdness - frozen data

Posted: Sat 13 Mar 2010 11:59 am
by steve
christer wrote:This morning I found flat figures in the graphs. At around 21:50 yesterday something happend. Since also the inside temp is flat it cannot be the RF from the outdoor sensor.
Your rain counter is doing strange things, at one point it jumped up by over 1000 mm. Cumulus thinks that the data is therefore corrupt and is ignoring it all.

9116.8133 : Ignoring bad data: raintot = 1031
9116.8133 : prevraintotal = 7

Restarting Cumulus has caused it to accept the new value and start again from there.

I notice you have a negative pressure offset, is this intentional?

2010-03-12 07:00:24 : EWUSB: Read pressure values: rel=1003,1 abs=1006,4 offset =-3,3

Normally you would need to be below sea level to need a negative offset, but it may be that you are close to sea level and that's the setting required to make the reading correct?

Re: All data flat-lined

Posted: Sat 13 Mar 2010 12:52 pm
by steve
I've had a look at the code, and I've changed the way this works in 1.8.9, to try to avoid it getting stuck still thinking the rain total is still bad and needing a restart.

Re: All data flat-lined

Posted: Sat 13 Mar 2010 1:51 pm
by Gina
I've found the rain sensor (or its cable) is rather sensitive to radio interference, giving a reading many times greater than if you had poured a bucketfull of water into the gauge! Correcting the resulting chaos is somewhat time-consuming. FO units seem to store the rain total in the transmitter unit and the console plus Cumulus (like other software) stores the data is various files.

Re: All data flat-lined

Posted: Sat 13 Mar 2010 2:01 pm
by steve
They seem to be able to go backwards temporarily as well! But I think that's not an interference problem. But whatever causes it, it's a pain trying to cater for all the little foibles. 1.8.9 tries a bit harder, but there's still work to be done.

Re: New Weirdness - frozen data

Posted: Sat 13 Mar 2010 2:32 pm
by christer
steve wrote:
christer wrote:This morning I found flat figures in the graphs. At around 21:50 yesterday something happend. Since also the inside temp is flat it cannot be the RF from the outdoor sensor.
Your rain counter is doing strange things, at one point it jumped up by over 1000 mm. Cumulus thinks that the data is therefore corrupt and is ignoring it all.

9116.8133 : Ignoring bad data: raintot = 1031
9116.8133 : prevraintotal = 7

Restarting Cumulus has caused it to accept the new value and start again from there.

I notice you have a negative pressure offset, is this intentional?

2010-03-12 07:00:24 : EWUSB: Read pressure values: rel=1003,1 abs=1006,4 offset =-3,3

Normally you would need to be below sea level to need a negative offset, but it may be that you are close to sea level and that's the setting required to make the reading correct?
Steve, thank´s for your answer!
I have not entered any offset to the pressure. All settings in the calibration dialog are default from the installation. Maybee some setting in in the console is wrong?
However, I has followed the wiki and korrected the faulty rain figures.
I think I have to visit the place where the PC and the station is and upgrade to latest version of Cumulus soon.

Thank you wery much for your explanation!
//Christer

Re: New Weirdness - frozen data

Posted: Sat 13 Mar 2010 2:37 pm
by steve
christer wrote:I have not entered any offset to the pressure. All settings in the calibration dialog are default from the installation. Maybee some setting in in the console is wrong?
If your station is above sea level, you need to set the relative pressure so that it matches a nearby station. Otherwise you will be logging incorrect pressure values.

Re: New Weirdness - frozen data

Posted: Sat 13 Mar 2010 3:08 pm
by dev0
christer wrote: I think I have to visit the place where the PC and the station is and upgrade to latest version of Cumulus soon.
Im using log me in... With that you can do everything. Including install new programs and restart, it work like a normal screen, just over the internet...
im not familiar with that program you are using, but im sure you can do the same, otherwise you can install log me in on the remote computer...

Re: New Weirdness - frozen data

Posted: Sat 13 Mar 2010 3:32 pm
by christer
dev0 wrote:
christer wrote: I think I have to visit the place where the PC and the station is and upgrade to latest version of Cumulus soon.
Im using log me in... With that you can do everything. Including install new programs and restart, it work like a normal screen, just over the internet...
im not familiar with that program you are using, but im sure you can do the same, otherwise you can install log me in on the remote computer...
Thank´s for your tip.
But RealVNC kan also do that. But since the PC is connected to a wireless modem (Ice.net) is the speed not so fast.

//Christer