WS3083 reset and consequent relative pressure issue

[6thNapoleon]

I recently performed a factory reset and memory wipe on the console of my WS3083, calibrating relative pressure according to an online calculator (http://www.srh.noaa.gov/epz/?n=wxcalc_altimetersetting). It informed me to add about 2.2 hPa to the absolute pressure reading.

However, when initialising Cumulus (version 1.9.4), I am informed that it is:

Ignoring dubious relative pressure reading, using offset of zero instead.

Consequently, historical data points recorded by the console and downloaded by Cumulus are rendered nonsensical. Moreover, looking at Jun14log.txt, there appear to be an insufficient number of data points recorded when the console is disconnected from my computer. Overnight, I typically obtain three (entered at random intervals?) over an eight hour period when entries should be made every half hour. Toggling through the data points on the console show that it is making them every half hour. They are also ‘sensible’. So I am not sure there is an issue with its memory. Nor am I certain it’s anything to do with its USB connection. Cumulus successfully displays information from the console ‘live’ when it is connected to my computer.

I have consulted the FAQ. At different points not only has there been a discrepancy in pressure values between the console and Cumulus, but an alarm sounds on initialisation to inform me pressure is dropping very rapidly (always by about 340 hPa/hr). As a result, I added

Code: Select all

EWpressureoffset=x.y

to cumulus.ini, altering x.y to 2.2. The warning about dubious relative pressure and unlikely pressure drops then disappears on initialisation, but the same issue of downloading insensible and insufficient data points from the console remains.

I would be immensely grateful for any assistance. I feel I have reached a bit of a cul-de-sac.

[steve]

Resetting the console while Cumulus has data needed to catch up will cause problems, as the data that it reads when it starts up will not be from the correct area of memory.

But it now seems to be getting 'garbage' data from the first block of memory, which is where the pressure offset is stored. It will also cause invalid data to be read from the logger for the 'catch up' process. All I can suggest is that you try resetting the console again. Do this immediately after closing Cumulus, and make sure you have the EWpressureoffset setting in cumulus.ini - in the latest run you seem to have taken it out again, so it's again getting a bad offset from the console.

You're not running some other software with the station (e.g. Easyweather) at the same time as Cumulus, are you?

As an aside, using a pressure offset purely based on your altitude is often not the best way to go about it, as it assumes that your pressure reading is accurate to begin with. It is better to set the relative pressure to a known good source, e.g. a nearby Met Office station, as this also compensates to a certain extent for any inaccuracy in the station. Now is a good time to do this, with the high pressure system in control.

[6thNapoleon]

Thank you for the swift response.

Have reset the console again (disconnected from my computer) and reintroduced the offset in cumulus.ini. After looking up a local station, I have tweaked the relative pressure and offset. Thank you for illuminating why this is more appropriate.

Will have to wait a while to see if the data catchup behaves itself.

To answer your question, I'm just running Cumulus.

[6thNapoleon]

No luck I'm afraid. Just connected up my console again and run Cumulus only to be confronted by an alarm that pressure's falling more than 300 hPa/hr. No flashing error message this time, but past data points downloaded by Cumulus on catchup are still either missing or nonsense.

For reference, my console currently reads relative pressure at 1020.9 hPa, Cumulus relative pressure at 1020.8 hPa and the offset in cumulus.ini is set at 2.2 hPa.

[steve]

If you zip up the diags folder again and attach it, I'll take a look.

I'm assuming it was all working fine before the reset that you mentioned at the start of the thread?

[6thNapoleon]

Thanks.

No, there was an error on 20 June. As far as I recall (and sorry for being nonspecific), it was something to do with the station not being initialised. I shrugged it off, resetting the console and the station. That appeared to solve it. I didn't get that error message again and the console was plugged in for the remainder of the day. I only became aware of this pressure issue when I connected it the next morning, only to have the pressure alarm sound and to discover that my overnight data was scrambled.

[steve]

The garbage data when you started it up at 14:35 was a result of the reset; Cumulus hadn't been running for an hour, so it tried to download the logger data to fill in the gap. As I mentioned above, if Cumulus needs to download past data and the console has been reset, it will get incorrect data.

Following that initial download, it all seems to be working correctly, and the first block of memory containing the pressure offset etc looks OK.

[6thNapoleon]

Right, I have a few questions. Going forward, can I expect all future data to be scrambled on catchup? If I wish to obtain weather data overnight do I need to keep the console connected, and Cumulus running, all night? Or manually edit Jun14log.txt with historical data points recorded by the console?

[steve]

Right, I have a few questions. Going forward, can I expect all future data to be scrambled on catchup? If I wish to obtain weather data overnight do I need to keep the console connected, and Cumulus running, all night?

If your weather station continues to work correctly, then Cumulus will be able to download the data correctly. If your weather station (or the USB connection) starts scrambling the data, then of course the data received by Cumulus will be scrambled.

You don't need to keep Cumulus running all the time; lots of people using Cumulus close it down overnight and don't have issues with data download when it starts up in the morning.

You would of course get better results by leaving Cumulus running all the time, even when your weather station is behaving correctly; particularly as you have the station's logger set to 30 minutes, and thus much of the data is being discarded while Cumulus is not running.

[AllyCat]

Hi,

... as you have the station's logger set to 30 minutes, and thus much of the data is being discarded while Cumulus is not running.

Most people set their FO logger interval to 5 minutes using Easyweather, or Steve's excellent utility. Even one minute is possible for overnight storage, but 5 minutes will permit logging for up to a couple of weeks.

For "holiday" periods it may be necessary to extend the logging interval, particularly with the 308x versions which (AFAIK) have a somewhat smaller capacity than the 108x versions (~3,000 vs ~4,000 records).

Cheers, Alan.