Welcome to the Cumulus Support forum.

Latest Cumulus MX V3 release 3.28.6 (build 3283) - 21 March 2024

Cumulus MX V4 beta test release 4.0.0 (build 4018) - 28 March 2024

Legacy Cumulus 1 release v1.9.4 (build 1099) - 28 November 2014 (a patch is available for 1.9.4 build 1099 that extends the date range of drop-down menus to 2030)

Download the Software (Cumulus MX / Cumulus 1 and other related items) from the Wiki

Download data missing with WS3085

Discussion specific to Fine Offset and similar rebadged weather stations
davepaterson
Posts: 145
Joined: Tue 01 Jan 2019 8:12 pm
Weather Station: HP2552 , WH40/WH68+WH57/GW1100C
Operating System: Windows 11/11
Location: Hamilton, Scotland UK

Re: Download data missing with WS3085

Post by davepaterson »

Both stations working as they should using Cumulus1. Did find that the supplied USB cable has sideways movement which if moved could cause Cumulus to lock up, (still added to the logfile but with same data only time changed).
The only time I get any loss of data is a couple of minutes when restarting after a overnight shutdown.

Did find this though in the Cumulus1 help Files.
Fine Offset USB problem Synchronise Fine Offset reads
This is an attempt at a work around for the 'USB lockup' problem with more recent Fine Offset stations. It works by avoiding communicating with the station at 'critical' times, i.e. when the console is updating memory. The station does this every 48 seconds, and every 60 seconds. Cumulus performs a synchronisation once a day, where it determines when the console is doing these updates, so it can avoid them.

MX and supplied cable out, gone back 7 years and Cumulus1 in.
davepaterson
Posts: 145
Joined: Tue 01 Jan 2019 8:12 pm
Weather Station: HP2552 , WH40/WH68+WH57/GW1100C
Operating System: Windows 11/11
Location: Hamilton, Scotland UK

Re: Download data missing with WS3085

Post by davepaterson »

C1 ran ok up to the 30th March when there was another lockup. After a lot of searching through the various forms and readme files I took all my programs and files out of the Cumulus folders and changed the way I read the logfiles and ticked the synchronise data box. No problems through April but on the 7th May we got good sunshine (eventually) and both stations started showing zero then 15UV and over 2kW/m2 for solar. More searches showed that the reading I had been taking for true were almost twice what they should be. I could only find calibration for the UV and not the solar on C1, so back onto the forms and saw the latest MX had several Fine Offset improvements so gave it another try with UV and Solar calibration set at .5 Had a missing data\lock up on the 9th after another new build went in but everything after that had been OK. Most of the problem came after an overnight shut down so now leave it on and give a few hours rest during the day. On the Greenfrog scientific web site the WS2085/3085 now has the info THAT IT IS NOT RECOMENDED TO RUN CONNECTED 24/7 , both use the new all in one sensor suite. Asking questions on the Greenfrog support form is going nowhere.
sfws
Posts: 1183
Joined: Fri 27 Jul 2012 11:29 am
Weather Station: Chas O, Maplin N96FY, N25FR
Operating System: rPi 3B+ with Buster (full)

Re: Download data missing with WS3085

Post by sfws »

mcrossley wrote: Thu 04 Mar 2021 3:18 pm I do not know how Cumulus 1 did it, but the MX code uses the value of first byte of each data block to determine the number of minutes to skip FORWARD to timestamp the next logger record.
Steve Loft did once document how Cumulus 1 did it, as I remember reading it.
As stated here, all forum content prior to June 2008 was lost.
Unless somebody can find this description on current forum, I guess Steve's description of how he read Fine Offset has been lost.

The Wiki did not start until August 2009, and was therefore unable to take any information from older forum.

Dave's Fine Offset with solar console will (AllyCat to confirm) interrogate transmitter for solar data every 60 seconds and for other data every 48 seconds.

The interval between Fine Offset logging of data will increase by one minute, every time there is a clash between a multiple of 48 seconds and a logging due time. So with a default logging interval of 30 minutes, approximately every 96 hours the interval will be 31 minutes, (although maths is a weak point of mine), and the extra minute is added on every few hours for most frequent logging intervals.

For the synchronization feature that Cumulus did on start-up, 1.9.4 managed to identify when the Fine Offset did latest update to its circular buffer.

What Cumulus 1 did, was two scans of the Fine offset memory.

On the first scan, done working backwards in time from that latest memory update time, Steve used the number of minutes (stored at byte zero of that memory) to construct a "table of logging times", as there is no date-time record on logging entries in that circular buffer, and byte one tells you how many minutes have passed since previous logging.

This is reverse of what Mark says for MX, if he means what he said. For 1.9.4, byte zero is determining the number of minutes to skip BACKWARDS, and in default case, as I say varies between 30 and 31.

On the second scan, working forwards in time, all those logs where the calculated log time exceeded the last update time in today.ini were processed.


One final point, if Dave is swapping between 1.9.4 and MX using the same Cumulus.ini file; my own view is that is not wise.
I stopped using synchronization (I had used that with 1.9.4), when I moved to MX, and while (with 1.9.4) I used a "EWpressureoffset=x.y" as it seemed to reduce lockups, I removed that for MX.
I have not tried it with a recent release of MX, but with releases more that a year ago, IIRC synchronization meant MX skipped every reading after first of each hour, and the pressure is more accurately read from a Fine Offset by MX (since an early change by Mark) than by 1.9.4 (which read wrong memory bytes)
davepaterson
Posts: 145
Joined: Tue 01 Jan 2019 8:12 pm
Weather Station: HP2552 , WH40/WH68+WH57/GW1100C
Operating System: Windows 11/11
Location: Hamilton, Scotland UK

Re: Download data missing with WS3085

Post by davepaterson »

I used C1 from the 30th March up to the 7th May with no lockups and used the cumulus.ini from MX with no problem. Only the today.ini needed a change. After moving to MX again and adjusting the solar and UV to .5 with MX I retried C1 with the MX data and cumulus.ini, it showed up 3 errors but didn't adjust the solar reading. I am keeping MX and the synchronization ticked for now as long as the lockups don't happen, so any other help in reducing the possibility would be more than welcome. If I have to go back to Cumulus1 I will be doing a clean start and rewrite my programs to adjust the C1 UV and Solar readings.
davepaterson
Posts: 145
Joined: Tue 01 Jan 2019 8:12 pm
Weather Station: HP2552 , WH40/WH68+WH57/GW1100C
Operating System: Windows 11/11
Location: Hamilton, Scotland UK

Re: Download data missing with WS3085

Post by davepaterson »

UV and Solar Question.
I know from reading that the UV of 15 will not happen in the UK and have estimated that the reading the stations give me is twice what it should be. I am totally new to the sun readings and what I can't find is what are the usual solar readings from the UK especially Scotland Central Belt. I have the Met Office UV website to compare UV but not solar. I have checked both station, one showing Lux the other W/m2 and both keep the same readings (.0079 calc). Accurate (the cost of the WS3085 probably compares with a sensor for the Davis) as of now readings are not needed (unless your next door), I am only looking for what is normal in W/M2 or Lux for cloud, bright, and sun in May.
Post Reply