[cjthomson]

Hello there

I'm looking for a little bit of advice. I've been having a few problems with my WeatherDuino2 RX V6 in terms of stability when connected to a Raspberry Pi 3. I've found that if I leave the RX unit disconnected from the RPi it works quite happily but if I hook it up to my RPi and start Cumulus MX it might continue to work for up to 12 hours or more often than not a few minutes - it just seems to stop communicating over the USB interface. I've tried different cables and I have switched to an older version of Raspbian (Jessie) with Mono 3.2.8.

I'm getting to the point where I feel that I need to rule out the Nano as being the problem and recall reading that a diode would need to be removed from it to stop it from being powered over USB. However the photos I've seen of other Nano's here have a very slightly different component layout.

I've attached a photo of the underside of my spare Nano (the side which has the CH340G controller) and I'd be grateful if someone might be able to confirm which component it is I should be removing - I think it's the component that sits between the USB port and the end of the CH340 controller.

   

Kind regards
Chris

[werk_ag]

... I think it's the component that sits between the USB port and the end of the CH340 controller.

Yes, its that component (black with white trace in one of the sides).


Before the radical step of removing the Nano from a Pro2 v6 board, may I sugest that you try to running the system with exactly the same version of CumulusMX but running on a PC, just to check if the USB connection will break.

Is the PI3 executing any other tasks than running CumulusMX?

You may also try to add this parameter to your Cumulus.ini file in the Station section:

Code:

VP2SleepInterval=1500

[cjthomson]

... I think it's the component that sits between the USB port and the end of the CH340 controller.

Yes, its that component (black with white trace in one of the sides).


Before the radical step of removing the Nano from a Pro2 v6 board, may I sugest that you try to running the system with exactly the same version of CumulusMX but running on a PC, just to check if the USB connection will break.

Is the PI3 executing any other tasks than running CumulusMX?

You may also try to add this parameter to your Cumulus.ini file in the Station section:

Code:

VP2SleepInterval=1500

Hi Werk

I'll dig out my old Windows laptop and setup CumulusMX on it and see what happens. The PI3 is running as a headless system purely for connectivity to the RX board so I'm not running any other tasks on it. 

Before I do that I will try adding that parameter to the Cumulus.ini file and see what happens.

One thing I've noticed is that when Cumulus appears to stop getting data from the RX unit the LCD display connected to the RX board doesn't update either. However when disconnected from the Pi3 the LCD display starts updating again.

Kind regards

Chris

[cjthomson]

Hi Werk

I've tried connecting the RX to my Windows laptop and ran CumuluxMX - (the latest release). CumulusMX was able to read from the RX fine for half an hour before it appeared to stop pulling data from it. This is similar to what I was experiencing with the Pi3 - although sometimes I can get up to 12 hours from it before it stops. Unfortunately I didn't think to switch on the additional logging features at the time   

I've switched logging on for subsequent sessions - these sessions seem to read the initial current values from the station and then nothing else apart from the indoor temperature and pressure. I've attached the Cumulus.ini file and a log file which I hope may give you more information.

Thanks 
Chris

Cumulus.txt Size: 16,6 KB  Downloads: 314

20180101-141520.txt Size: 56,02 KB  Downloads: 264

[werk_ag]

Hi Rich,

The CumulusMX log can be usefull (there is no need to enable the Cumulus debug log option), however you should keep the system running until it loses connection.
From your CumulusMX log, I only can see a CumulusMX complain about a file with a invalid value:

Code:

2018-01-01 14:15:22.321 Error at line 1 of data\Dec17log.txt : Value cannot be null.
Parameter name: input
2018-01-01 14:15:22.321 Please edit the file to correct the error

Don't know if this error can cause any problem.
For testing you could try starting CumulusMX (on the test PC) by firts cleanning all files in the data and backup folders.

Maybe I'm wrong but I have a feeling that the problem is not with the Nano, but something else. I saw, that you have the realtime upload set to 30 seconds, have you checked that the ftp upload to your server occours without any problem?

Thank you

[cjthomson]

Hi Rich,

The CumulusMX log can be usefull (there is no need to enable the Cumulus debug log option), however you should keep the system running until it loses connection.
From your CumulusMX log, I only can see a CumulusMX complain about a file with a invalid value:

Code:

2018-01-01 14:15:22.321 Error at line 1 of data\Dec17log.txt : Value cannot be null.
Parameter name: input
2018-01-01 14:15:22.321 Please edit the file to correct the error

Don't know if this error can cause any problem.
For testing you could try starting CumulusMX (on the test PC) by firts cleanning all files in the data and backup folders.

Maybe I'm wrong but I have a feeling that the problem is not with the Nano, but something else. I saw, that you have the realtime upload set to 30 seconds, have you checked that the ftp upload to your server occours without any problem?

Thank you

Hi Werk

I was playing about with the Rpi3 and my RX unit this evening and one thing I've noticed is that the display stops updating the data that it should be receiving from the TX unit outside when I have got the USB cable connected between the Rpi3 and the RX. As soon as I disconnect the USB cable the display updates the windspeed and outside temperature values on the display. Please note that at this point the Cumulus MX software was not running.

I have now tried about four different cables this evening and it has made no difference. I'm not really sure what I should be checking next but I have a feeling that this may have something to do with it.

Kind regards
Chris

[werk_ag]

Hi Chris,

Well, that is new info, which indeed configures an issue that seems related with the diode on the Nano. Some post above you tell me that your Pro2 RX board is a v6, unless some error occured during the assembling (as you know Pro2 RX v6.x boards are suplied fully assembled) the Arduino Nano installed on those boards already have the diode removed. That is the reason I started by discarding a problem with the diode.
Its easy to check if the diode was been removed or not. If the Nano becomes powered by just inserting the USB cable, then the diode is there.

Regards