Ecowitt GW3000 user assistance

[gjr80]

Posted in weewx-user to canvass a wider audience.

I'm developing a new Ecowitt driver based on the so called local device HTTP API. This driver will have catchup on startup via either local device history files (for supported devices) or Ecowitt.net (for devices that upload to Ecwoitt.net). Ecowitt's documentation of the local device HTTP API and GW3000 SD card data files is somewhat limited, so I was hoping for some practical help from an Ecowitt GW3000 user (if one exists) who has both tipping (WH40) and haptic (WS85/WS90) rain gauges connected to their GW3000.

What I am after is the header information from the basic sensor CSV file stored on the GW3000 SD card when the GW3000 is connected to both WH40 and WS85/WS90 rain gauges. The basic sensor CSV file name format is YYYYMMX.csv where YYYY is the year, MM is the two digit month number and X is an upper case letter. The basic sensor file can be downloaded via the local device web page or directly in your browser via http://device_ip_address:81/YYYYMMX.csv where device_ip_address is the device IP address and YYYYMMX.csv is the file name as described previously. You can view the availble SD card history files via the device web page or at http://device_ip_address/get_sdmmc_info.

The required data does not contain any sensitive info.

thanks,
Gary

[michael.k...@gmx.at]

Hi Gary, all I am missing is the GW3000, since I was waiting for your updated driver. How about you? If it's the same situation for you, tell me the shipping address, I'll get you one, as a donation for the project, or, if you decline, I'll buy one for myself, but it'll take a while until I'll arrive.

[gjr80]

I have the GW3000 and WS90 so have the haptic sensor covered, and I can surmise the behaviour with a WH40 only, it is just a case of confirming what happens when both WS90 (or WS85) and WH40 are present. Thanks for the offer, but if I don't get any data responses here I will likely just buy a WH40 and wait for the couple of weeks for delivery.

All going well I hope to have the new driver with backfill ready for user testing sometime in 2nd half of February.

Gary

[michael.k...@gmx.at]

I have a spare WH40 for 868MHz, but you are not in Europe, are you? Too bad the GW3000 868MHz is sold out currently, I won't get one before end of February, I assume.

[michael.k...@gmx.at]

I just ordered a GW3000, but it won't ship before March.

[gjr80]

Yes, 433MHz here. I have a WH40 on the way, what's another rain gauge...

Gary

[michael.k...@gmx.at]

My GW3000 is on the way, so it should arrive in a week or so.

[Auchtermuchty Weather]

I have a GW3000 in the UK, Very happy to help if I can.

[Steeple Ian]

Hi Gary,

I have the two types of rain gauge and the GW3000 arrived today.

Attached is a .zip of the 6 CSV files from my SD Card after a few hours running.

Regards,

Ian

[Steeple Ian]

...... It does seem however that it only creates columns for the traditional rain gauge and not a piezeo rain gauge, even if priority is set to piezo. I am not sure if this is a bug, an oversight or deliberate. This happens even when the WH40 is disabled. A somewhat disappointing outcome.

[gjr80]

Thanks for your trouble Ian, I hadn't picked up on the piezo bug. I set my GW3000 and WS90 up and let it run for a couple days in December before looking at the csv files. Unfortunately those two days had no rain so didn't think anything of the lack of rain data. Just looked back over the last 7 weeks and no rainfall was recorded in the csv files.

I saw your wxforum post, so I guess we wait. My WH40 is in the mail so I guess chances are it will be here before any firmware update is released.

Gary

[Ian Millard]

No problem Gary.

I see that Gyvate has submitted the request to Ecowitt however I have no sense on how quickly they can respond. 

Ian

Sent from my iPad.

On 11 Feb 2025, at 10:55, gjr80 <gjrod...@gmail.com> wrote:

Thanks for your trouble Ian, I hadn't picked up on the piezo bug. I set my GW3000 and WS90 up and let it run for a couple days in December before looking at the csv files. Unfortunately those two days had no rain so didn't think anything of the lack of rain data. Just looked back over the last 7 weeks and no rainfall was recorded in the csv files.

--
You received this message because you are subscribed to the Google Groups "weewx-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to weewx-user+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/weewx-user/27c66b98-c8be-4c52-93ec-bc34c60ce637n%40googlegroups.com.

[Rainer Lang]

there are a few more things to be considered from the bigger perspective with the GW3000
- some referring to the SD card archiving of also other sensor values (e.g. LDS01, WH57 ...), the file name change etc - all needs to be streamlined - already a few things were sorted with the last firmware upgrade
- some referring to the customized server posting

The WiKi is quite up to date regarding the expected-to-be situation and partly also regarding the "missing feature" situation.
https://meshka.eu/Ecowitt/dokuwiki/doku.php?id=start#gw3000
there is a table with the so-far existing observations in the CSV files corresponding to the header items - and the known missing ones

all this also needs to be synchronized with the WS6210 console, which is the other one having the API accessible SD card option

Some field testing may also be needed

an educated guess would be the end of the month - but they might be faster

To view this discussion visit https://groups.google.com/d/msgid/weewx-user/B854BC65-02F5-47D9-BA6A-6BF480C1C89B%40btinternet.com.

[Rainer Lang]

Ecowitt has confirmed that the missing piezo data, the WH7 time stamp of the last lightning event and the LDS01 depth will be added to the SD card data with the next firmware for the GW3000 and WS6210.

@Gary: the header text now contains for temperatures a one-character representation of °C and F - this will become two letters/characters "°" and "C" resp. "F"

All the LDS values have already been added to the WiFi firmware of the other consoles (those which have two firmwares) for the Customized Server and for those people using the Interceptor driver it can now also be retrieved
(of course you have to use an extended version of the driver and not the "barebone" version at GitHub - that could get an update with all now existing sensors and values ...)
The world has significantly changed in the past years ... there are user created (tested and working) complete versions available though
Gary's local API driver looks up to date (probably LDS01 to be added)

[steepleian]

Good news indeed. Thank you for your update.

Ian

https://claydonsweather.org.uk

On 17 Feb 2025, at 10:50, 'Rainer Lang' via weewx-user <weewx...@googlegroups.com> wrote:



To view this discussion visit https://groups.google.com/d/msgid/weewx-user/f25dbea6-eb22-48d0-8f01-929628903c85%40gmail.com.

[michael.k...@gmx.at]

In the mean time my GW3000 is installed already in my test environment, so if there is anything I can help with, let me know. I've got no problem even with a still-not-ready, potentially buggy, bleeding edge version of the driver in my test environment.

[Steeple Ian]

I see that the new firmware has arrived. GW3000A_V1.0.1

[Steeple Ian]

Having said that, it is strange that version number has gone backwards and no sign of piezo rain data in the card .csv files after the old files had been deleted.

[michael.k...@gmx.at]

That was a false rollout they did. 
See https://www.wxforum.net/index.php?topic=48216.msg486394#msg486394

[Rainer Lang]

roll-out mistake (almost) in time detected, reported and corrected - the future now hopefully final SD card header layout coming with firmware 1.0.4 is in the WiKi
for the other network readable SD card device, the WS6210 WLAN/4G gateway, additional observations apply - see also WiKi, table of possible headers
https://meshka.eu/Ecowitt/dokuwiki/doku.php?id=start#gw3000-csv
VPD (vapour pressure deficit), unit kPa, an agriculture related observation, will be added to the YYYYXXZ.csv file behind the "Feels Like" observation

To view this discussion visit https://groups.google.com/d/msgid/weewx-user/94a32c0f-ec10-4a02-8ba9-75b7720f1afbn%40googlegroups.com.

[Steeple Ian]

Firmware 1.04 has finally been rolled out for GW3000

[Steeple Ian]

Sorry, that should be 1.0.4
Ian

[Rainer Lang]

the just released firmware final 1.0.4 upgrade for the GW000 should cover all missing entries in the csv files

To view this discussion visit https://groups.google.com/d/msgid/weewx-user/bfc20d8a-5be2-47bf-ab8d-0fa578363cb4n%40googlegroups.com.

[Steeple Ian]

Gary,

I was just wondering how the development of the new driver is progressing? If you need any help with testing, I am very happy to help.
Ian

[gjr80]

Thank you Ian,

I have a working version running on a system at home. It needs some tidying up plus the backfill via device history files needs a touchup after the v1.0.4 firmware release. The absence of any complete and reliable local device HTTP API documentation is not helping. I am away on holiday in months time and am aiming to have a version suitable for public use by then.

Gary

[Steeple Ian]

Thank you for the update Gary.

Ian

[michael.k...@gmx.at]

Hi Gary, are there any news regarding a new version?

And by the way: will backfill work when the driver is configured as a service?

[gjr80]

It's close, I actually packaged up a beta last week but then found another issue. The lack of API documentation and dealing with the large number of unit combinations have been challenging (eg the CSV files used during backfill may all have different units and some of those units don't exist in WeeWX). I'm not putting a timeframe on it as I have done so before and then had to eat my words. But it's close.

Backfill is a function of a WeeWX driver, so whilst the Ecowitt local HTTP driver can function as a driver and a service, backfill will only be supported when operated as a driver.

Gary

[f4n...@gmail.com]

How does the GW3000 store wind gust data when the interval is set to 5 or 10 minutes? Will it store the max gust that occurred in between the last 5 or 10 minutes the right way, or just the current gust at minute x? Will wind gust backfill work correctly even when the weewx reporting interval and the GW3000 sd store interval differ? Is there a limit of backfill timeframe or could the GW3000 be left in a remote location standalone and the readout of e.g. 6 months or 1 year worth of data would work flawlessly? Or even a few years worth of data in case of a corrupted weewx-database? How fast will the backfill process work e.g. on a raspberry?

A lot of questions but I just want to get an idea what'll be possible with the gw3000. Keep up the good work, with this amazing driver/gw combination there will be no need anymore to go for expensive weather stations/logging solutions (e.g. the outdated, limited davis loggers) as long as power is provided to the gw3000 at least.

fankyy

[Rainer Lang]

I suggest you look up your questions in the Fine Offset / Ecowitt WiKi first (link see below)
it's all there - read the GW3000 section, read the chapter on Ecowitt outdoor sensor arrays
and there are many more topics in the context.

You sort of ask your questions in the wrong forum - www.wxforum.net with the Ambient Weather and Ecowitt and other Fine Offset clones subforum,
from where the above mentioned WiKi originates, would be the better place 
Or just read the WiKi:

https://meshka.eu/Ecowitt/dokuwiki/doku.php?id=start#dokuwiki__top

To view this discussion visit https://groups.google.com/d/msgid/weewx-user/a480cc33-a640-4656-81c5-9933e7aba61bn%40googlegroups.com.

[scottjl]

Was this ever released? I just came across this thread and would be very much interested in trying it out. I happen to have a GW3000, a WS90 and a WH40H and a few other sensors and would be happy to share my data (no secrets here, it's just weather data) if you need more samples. I'm running firmware GW3000B_V1.1.0.

--scott

[michael.k...@gmx.at]

Yes and no and maybe. You may want to check this thread: https://groups.google.com/u/0/g/weewx-user/c/6Cc9qcWsa9A and this one: https://groups.google.com/u/0/g/weewx-user/c/-rbfcQsNo_s 

[Ian Millard]

This was never released as finished version by the original author as he appears to be no longer active and his Github account is closed.

However, there are a couple of forks from the development branch that Gary was working on (which was nearing completion).

You can find my fork at https://github.com/Millardiang/weewx-ecowitt_local_http/tree/development

Also Werner Krenn has a further modified version at https://github.com/WernerKr/Ecowitt-or-DAVIS-stations-and-Season-skin/blob/main/ecowitt_http/ecowitt_http.py

Ian

--
You received this message because you are subscribed to the Google Groups "weewx-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to weewx-user+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/weewx-user/59587d73-34b0-4795-8572-5e8d8ed87ad5n%40googlegroups.com.

[Ian Millard]

This was never released as finished version by the original author as he appears to be no longer active and his Github account is closed.

However, there are a couple of forks from the development branch that Gary was working on (which was nearing completion).

You can find my fork at https://github.com/Millardiang/weewx-ecowitt_local_http/tree/development

On 5 Aug 2025, at 02:51, scottjl <sco...@gmail.com> wrote: