[werk_ag]

Definitely not supported in Raspberry as well:

"Problem USB to Serial Adapters
HL-340

CH340 Chipset - Currently not supported by RPi but there is a patch of kernel code here, but it is for a 2.X kernel. If you find you have bought one of these, then it may work under Windows, but as of writing there is no support for RPi. Otherwise you can have a go at getting the patch to work."

I can't agree. I have an RX test unit, built with an Arduino with the CH340 chipset running CumulusMX for weeks in a PI B+ without any major problem.

Question: If I am to replace the CH340 Nano with an FTDI Nano will it work without modification (I mean with a C3 instead of an R3)?

I tested it recently, and can confirm that the anti auto-reset circuit with the C3 capacitor, also work with Arduinos with the FTDI chipset.

[hvalentim]

I can't agree. I have an RX test unit, built with an Arduino with the CH340 chipset running CumulusMX for weeks in a PI B+ without any major problem.

Yes. With Cumulus (Mono) it runs in Linux in general flawlessly. I think they may be thinking about other stuff.

I must say than it communicates via minicom pretty well too. So the specific reason for it not to work with Weewx (Python) is enigmatic. But it should lie in the lack of consistency of the CH341.KO driver. For instance here you will find someone reporting a problem related with pyserial not initiating the serial port speed like minicom.

I tested it recently, and can confirm that the anti auto-reset circuit with the C3 capacitor, also work with Arduinos with the FTDI chipset.

Great. That seems to be the way for me to go.

[werk_ag]

Great. That seems to be the way for me to go.

Maybe not only for you...

[hvalentim]

Just installed my new Nano featuring an FTDI chip:

Code:

Bus 001 Device 003: ID 0403:6001 Future Technology Devices International, Ltd FT232 USB-Serial (UART) IC

[  101.916498] usb 1-1: new full-speed USB device number 3 using xhci-hcd
[  101.941989] usb 1-1: New USB device found, idVendor=0403, idProduct=6001
[  101.948780] usb 1-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[  101.955967] usb 1-1: Product: FT232R USB UART
[  101.960371] usb 1-1: Manufacturer: FTDI
[  101.964229] usb 1-1: SerialNumber: XXXXXXXX
[  102.133952] calling  usb_serial_init+0x0/0x167 [usbserial] @ 2966
[  102.140355] usbcore: registered new interface driver usbserial
[  102.146221] usbserial: USB Serial Driver core
[  102.150678] initcall usb_serial_init+0x0/0x167 [usbserial] returned 0 after 1                                                                0233 usecs
[  102.190443] calling  ftdi_init+0x0/0x97 [ftdi_sio] @ 2966
[  102.195972] USB Serial support registered for FTDI USB Serial Device
[  102.202644] ftdi_sio 1-1:1.0: FTDI USB Serial Device converter detected
[  102.209484] usb 1-1: Detected FT232RL
[  102.213170] usb 1-1: Number of endpoints 2
[  102.217315] usb 1-1: Endpoint 1 MaxPacketSize 64
[  102.221977] usb 1-1: Endpoint 2 MaxPacketSize 64
[  102.226640] usb 1-1: Setting MaxPacketSize 64
[  102.232640] usb 1-1: FTDI USB Serial Device converter now attached to ttyUSB0
[  102.240004] usbcore: registered new interface driver ftdi_sio
[  102.245785] ftdi_sio: v1.6.0:USB FTDI Serial Converters Driver
[  102.252482] initcall ftdi_init+0x0/0x97 [ftdi_sio] returned 0 after 55124 use

While trying to start Weewx 3.1 (not 3.0.1) I end up with exactly the same error messages described in my post #14 above.

[werk_ag]

Don't know what to say.
Is it working with Cumulus MX in your Linux version?

[hvalentim]

In case it helps I am attaching logs I got of serial communication between weewx and weatherduino using interceptty,

I tried downgrading Weewx from 3.1 to 3.0.1 so as to match the version you are using with no luck.

As far as Cumulus is concerned, it seems to be working as before.

Note: the attached log files include the results of "wee_config_device" and "service weewx start" threefold:
1) the command line issued and results seen;
2) what showed up in syslog;
3) what data was effectively exchanged through serial comm.

[werk_ag]

You haven't replied to my question, so I do it again.
Is it working with Cumulus MX in your Linux version?

If yes, please attach the last .txt file from MXDiags folder

[hvalentim]

As far as Cumulus is concerned, it seems to be working as before.

I would consider this to be a conclusive positive answer, bearing in mind that elsewhere I wrote a number of times already that it does work fine with Cumulus.

I am attaching the requested log. Better still, the all folder (I suspect the last log may have a lot of garbage from connection issues).

[werk_ag]

I would consider this to be a conclusive positive answer, bearing in mind that elsewhere I wrote a number of times already that it does work fine with Cumulus.

Yeah... But the CumulusMX logs revelead some important thing.

It seems that even with Cumulus MX, sometimes your system have problems reading data from the RTC eeprom, which is what is happening with WeewX.

2015-05-16 08:56:22.512 Serial device = /dev/ttyUSB0
2015-05-16 08:56:22.580 Connected OK
2015-05-16 08:56:22.609 Reading firmware version
2015-05-16 08:56:22.864 0A-0D-4F-4B-0A-0D-31-2E-39-30-0A-0D-00-00-00
2015-05-16 08:56:22.865 FW version = 1.90
2015-05-16 08:56:22.869 Reading reception stats
2015-05-16 08:56:23.171 0A-0D-4F-4B-0A-0D-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00
2015-05-16 08:56:23.172 0 0 0 0 0

2015-05-16 16:18:00.479 Connected OK
2015-05-16 16:18:00.492 Reading firmware version
2015-05-16 16:18:00.776 Array index is out of range.
2015-05-16 16:18:00.781 4F-4B-0A-0D-0A-0D-4F-4B-0A-0D-31-2E-39-30-0A
2015-05-16 16:18:00.782 FW version = OK
2015-05-16 16:18:00.787 Reading reception stats
2015-05-16 16:18:01.089 0A-0D-4F-4B-0A-0D-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00
2015-05-16 16:18:01.090 0 0 0 0 0

Readings and writings to the RTC eeprom are done via the I2C bus, if I'm not wrong, you installed the R5 and R6 resistors that are stated in the parts list, as components not used or reserved for future use.
Remove them (just cut the legs), maybe they are causing problems on the I2C bus.

Another suggestion is that you try to completely wipe both eeproms (Arduino and RTC), and restart the system again.
To do this, just uncomment this line

Code:

EEPROM.write(FirstRunCheck_Eeprom_adr, 0);

program it, let it start, comment the line again, and before 2 minutos reprogram it again.

[hvalentim]

R5 and R6 are empty. I never installed them.

Tried the EEPROM.write procedure. What happens is: after the second reprogram, after around two minutes, the RX hangs for a few seconds and then reboots alone..

When it comes back, after a while, it hangs one more time (without self rebooting, I have to unplug the power) and then apparently starts to work normally.