Topic: WifiLogger Realtime Data Page

[RichG]

I've been using my new WifiLogger for a few days now. Seems good but I have a couple of questions that I could not find answers to in the documentation.

On the real-time data page there is an orange message saying "PC Software Connected". What does this mean? I'm guessing it means that my WeeWx server is receiving the LOOP data but wanted to confirm.

Secondly the real-time data page doesn't update, I'm sure when I first installed it then this was updating every few seconds but not any more. Is this normal?

Thanks in advance.

Richard

[ocala]

I have the original wifi logger so things may be different then your wifi logger2. I do not have that message on my real-time page but mine is connected to my PC via TCP. I use Weatherlink software and Cumulus software which I download data to once a week.
As for updating it should be doing that every 3-5 seconds. If its it's not I would try to reboot the logger. I believe there was a small white button you had to press for a couple seconds to reboot. 

[RichG]

Its the Version 2 logger with WeeWx 5.1.

I have rebooted the logger from the web interface with no change. I presume it has the same impact as the white button reset?

edit: For info WeeWx is using the loop data from the Wifi Logger so should be no different to any other system.

[ocala]

TBH I don't anything about WeeWx. Maybe someone else with some experience could chime in.

[RichG]

So I stopped WeeWx running and the orange "PC Software Connected" message disappears and the real-time  web interface starts up. As soon as the WeeWx server starts reading the LOOP data the orange message "PC Software Connected" pops up and the real-time interface stops? It would be good to confirm whether this is expected behaviour? The documentation for the WiFi Logger badly needs updating to reflect the improvements made in the latest firmware.

[ocala]

Based on that it would seem that maybe you could tweak some settings in WeeWx  to correct the issue?

[davidmc36]

...It would be good to confirm whether this is expected behaviour? ...

That's exactly how all three of mine work. I'm sure it is normal for the device.

[johnd]

A lot of people will use the WFL solely to upload to an online platform such as weatherlink.com, which it will typically do no more than once per minute. This leaves the microcontroller plenty of time to update the real-time data page. But if a local computer is consuming the loop packets immediately they are generated then I'm guessing that there may not be enough time to refresh the real-time data page reliably.

If Wojtek sees this then he will obviously be able to comment further.

[PaulMy]

Hi,

I've been using my new WifiLogger for a few days now. Seems good but I have a couple of questions that I could not find answers to in the documentation.

On the real-time data page there is an orange message saying "PC Software Connected". What does this mean? I'm guessing it means that my WeeWx server is receiving the LOOP data but wanted to confirm.

I don't use WeeWX but use CumulusMX, as well as the several third-party uploads including Weatherlink.com.  The PC Software Connected message comes on when CumulusMX has disconnected connected.  The settings in CMX - Disconnect period:

Code: [Select]

Disconnect period:  10
 Period of time to disconnect from IP loggers to allow them to upload to WeatherLink.com. The disconnect happens once a minute. Time in seconds

Secondly the real-time data page doesn't update, I'm sure when I first installed it then this was updating every few seconds but not any more. Is this normal?

Mine updates but sometimes hard to notice as most data is without decimals.  And it also would only update while the attached PC is disconnected - i.e. 10 seconds in each minute.

Enjoy,
Paul

Edit for typing error

[CamarilloWX]

I don't use WeeWX but use CumulusMX, as well as the several third-party uploads including Weatherlink.com.  The PC Software Connected message comes on when CumulusMX has disconnected through settings in CMX - Disconnect period:

I also use CumulusMX.  What I have found is that you have to have a disconnect period configured in the Davis Connection settings or the Real Time Data for the WifiLogger never updates.  When CumulusMX is connected to the WifiLogger, the Orange Message "PC Software Connected" is displayed.  When CumulusMX disconnects for the period set in the Davis Connection settings, the Orange Message "PC Software Connected" is not displayed.  I have the disconnect period set to five seconds.  This may be in a manual somewhere but I haven't found it.

[RichG]

Thanks everyone for all the replies.

I've checked WeeWx and there doesn't appear to be any way of regularly disconnecting from the LOOP feed to give the logger a chance to update its real-time display. By the sounds of it this is normal behaviour but it was concerning me as I thought the logger had gone off-line as nothing was being updated. In a perfect world in would be nice for the real-time display to have a heartbeat shown every time a LOOP packed is read or something so you can can confirm everything is connected and OK. It would be great if the developer could confirm whether this is to due to a resource limitation on the device.

The other gotcha which I'm now aware of is the device heating issue, I've only ever used the serial logger before but as it wasn't compatible with my newer console and Davis was charging a rediculous amount of money for a new one I bought the WiFi Logger instead. I'll have to have a scour through the posts here to see what people have come up with to mitigate the effect and whether it's linear so a simple offset works? I'm presuming this will also have an impact on the relative humidity as well. Not the end of the world but it would have been nice if ProData had mentioned this on their product page so at least customers are aware.

Thanks again for your responses and also to Storm Bert for giving me something interesting to test on.

[johnd]

I'm presuming this will also have an impact on the relative humidity as well. Not the end of the world but it would have been nice if ProData had mentioned this on their product page so at least customers are aware.

Well, since I wrote our original WFL web page then you can blame me for that.  But I no longer maintain the page - someone else does that now, so cannot easily change the text. But I would make a couple of points:

First, the small degree of console warming that occurs with WFL also happens with any enhanced logger like WeatherlinkIP, Meteobridge Nano etc. The extra controller onboard such loggers and WiFi or network chips just unavoidably generates more heat than the very simple circuit of the serial/USB loggers. And heating also occurs of course if you use the backlight more than momentarily. The point being that it's not an issue that is unique to WFL.

Second, I've always regarded the inside temp/hum sensor as a slightly weak point of the console design. The sensor is buried inside the console and so is inevitably somewhat slow to respond to changes in room T/H and vulnerable to any heating effect from additional circuit components. Would have been much better in my view at least to have a socket on the console for an optional external T/H sensor. (Actually, from memory, the wireless Envoy does have such a socket.) But water long since under the bridge in respect of the 6312 console..

[RichG]

First, the small degree of console warming that occurs with WFL also happens with any enhanced logger like WeatherlinkIP, Meteobridge Nano etc. The extra controller onboard such loggers and WiFi or network chips just unavoidably generates more heat than the very simple circuit of the serial/USB loggers. And heating also occurs of course if you use the backlight more than momentarily. The point being that it's not an issue that is unique to WFL.

Second, I've always regarded the inside temp/hum sensor as a slightly weak point of the console design. The sensor is buried inside the console and so is inevitably somewhat slow to respond to changes in room T/H and vulnerable to any heating effect from additional circuit components. Would have been much better in my view at least to have a socket on the console for an optional external T/H sensor. (Actually, from memory, the wireless Envoy does have such a socket.) But water long since under the bridge in respect of the 6312 console..

When I first used the WiFi Logger I set maximum Tx Power due to a distant location and also set the Powersaving to Maximum Power as there was zero guidance or mention of the heating issue, this caused a 4C temperature increase which I would not call small? Reducing the transmit power to 11dBm and turning the power saving back on reduced this to about 2C and removing the battery cover down by another 0.5C. I just feel this known issue should be published by the seller and be in the documentation so at least the buyer is informed.

I couldn't agree more with the design of the console, there should have been a flying lead with the sensor as a minimum or the ability to plug in a module as you say. If the new console from Davis had a local data feed and a decent battery life then I would have bought that instead but whether that is an improvement? At some point I will build a RFM69 / ESP32 logger but until that time the WifI logger is fine.

[ocala]

I'm presuming this will also have an impact on the relative humidity as well. Not the end of the world but it would have been nice if ProData had mentioned this on their product page so at least customers are aware.

Well, since I wrote our original WFL web page then you can blame me for that.  But I no longer maintain the page - someone else does that now, so cannot easily change the text. But I would make a couple of points:

First, the small degree of console warming that occurs with WFL also happens with any enhanced logger like WeatherlinkIP, Meteobridge Nano etc. The extra controller onboard such loggers and WiFi or network chips just unavoidably generates more heat than the very simple circuit of the serial/USB loggers. And heating also occurs of course if you use the backlight more than momentarily. The point being that it's not an issue that is unique to WFL.

Second, I've always regarded the inside temp/hum sensor as a slightly weak point of the console design. The sensor is buried inside the console and so is inevitably somewhat slow to respond to changes in room T/H and vulnerable to any heating effect from additional circuit components. Would have been much better in my view at least to have a socket on the console for an optional external T/H sensor. (Actually, from memory, the wireless Envoy does have such a socket.) But water long since under the bridge in respect of the 6312 console..

Just curious if this heating issue with Wifi logger is just with the Wifilogger2?
I have a VP2 Console running the original Wifilogger. While your right it is slow to respond it matches the thermostat in my house. They are mounted at the same height about 5ft apart.   

[johnd]

I'm not aware of any difference in heat generation between WFL1 and 2. There are many factors that determine how much heating you see. For example, a logger that is passing LOOP data to a local computer every 2 seconds via WiFi is obviously doing more work than a configuration where the WFL is just there to upload to a remote web server say once every 5 or 10 minutes. And then eg what is the ambient temperature of the room where the console is located. How well ventilated is the area around the console. etc etc.

[RichG]

The WFL 2 uses a much faster ESP32 module, whether it gives out more heat or not I couldn't say? The console is hung on the wall so there is very little gap between it and the console back, I am going try and move it away by 5mm as that will help further.

[RichG]

I've also discovered that my data has stopped being uploaded by the WiFiLogger to WOW and Weather Underground, no idea whether this is related to having WeeWx connected? WeeWx is also filling its logs up with unable to wake console messages even though the WiFi signal is excellent and nothing else on the network is having issues.... I'm going to start a new thread over in the WeeWx section to see if anything can be done to debug this.

EDIT: Update, having looked through the WiFi Logger options it seems there is a choice on how the WFL prioritises the TCP access, default is set to "Pause with disconnect". The tooltip states this will pause the port every 60 seconds to allow the logger to upload and update, so from this setting I would have expected the Real-time data to update every 60 seconds and for the data uploads to work. I've moved the data uploads to WeeWx and disabled uploads on the WFL and set the TCP Access to Pause which gives the TCP port priority so we'll see if this helps matters.