Author Topic: Atmocom - A beta tester's review  (Read 293 times)

0 Members and 1 Guest are viewing this topic.

Offline kbellis

  • Forecaster
  • *****
  • Posts: 357
Atmocom - A beta tester's review
« on: May 07, 2019, 11:21:00 AM »
 [ You are not allowed to view attachments ]
(Final product shown above and available for 39.00 / $44.25 U.S. + shipping (to U.S. and Canada, $28.37))

I've been beta testing the ATMOCOM device since mid-January when it arrived from Spain. The package left Arroyo Dela Meil on 12 Jan and amazingly arrived in Ellsworth on 14 Jan!  It's a tiny device, fitting in the palm of your hand, that captures data being sent from the PWS's display console on it's way to local data logging, remote servers, or both types of intake.

 [ You are not allowed to view attachments ]
(Beta device that arrived 14 JAN 2019)

In a limited sense, the ATMOCOM is a WiFi access point, with the only WiFi traffic solely dedicated to your PWS. Configuration of the device is via a minimalistic telnet session, and most recently, using a web configuration page for your browser; however, I've only been using the telnet client.

 [ You are not allowed to view attachments ]

My tests have been exclusively capturing and logging data on a Raspberry Pi 3B+ running Raspbian (NOOBS) off of a 32GB SD card, and in the forwarding of that data to AmbientWeather.net and Weather Underground, both as configured through the display console on the WS-2000, or using the AWNET app for the the WS-2902A. Others may be able to add their reports on how ATMOCOM has performed on their web servers and websites using the php-parsers and php-formaters.

The tests have involved, and in using only one at a time; i.e., one PWS per ATMOCOM device, 1) an Ambient Weather WS-2000; and 2) an Ambient Weather WS-2902A; both shown below.

 [ You are not allowed to view attachments ]
(Atmocom beta device wrapped in an adhoc single business-card-enclosure and resting on top of the RPi)

The data from the PWS is actually being captured and written by two separate programs running concurrently 24/7 on the RPi; a) atomudp36.bin, which automatically creates monthly databases; e.g., /home/pi/atmocom/wxdb/wx201905.db; and b) the weewx-atmocom-udp driver written by Arthur Emerson (with many, many thanks!!!  [tup] ) for weewx. Both of these softwares have been running 24/7 without significant issue and have been updated by Arthur Emerson and Relko Vazic (weewx-atmocom-udp and atomudp36.bin, respectively). Moreover, the ATMOCOM hardware has had absolutely no issues, solidly performing since January 15th.

The most recent test, which is still running since Sunday morning and seen above, has to do with powering the ATMOCOM device using the RPi 3B+ USB port, though officially, I don't think it's technically advised, particularly since the RPi was also powering the keyboard and corded mouse, and in light of maximum total USB peripheral current draw listed for the RPi 3B+. The specific 5v 2.5A power supply for the RPi 3B+ being used in all of these tests (together with the case and tiny fan by Miuzei), appears to have been up to snuff.

 [ You are not allowed to view attachments ]

Without hooking up a multi-meter, it's not exactly clear if at any time there was ever an issue; however, I can report that all of the blinking LEDs looked to be doing their job and have not signaled any issue. Likewise, in browsing the data, both of the incoming streams, and what's been written to the files, it all seems good. I won't be leaving this setup running after this test, but thought that it was still worth mentioning. Once my tests have concluded, the ATMOCOM will be powered from one of the USB ports on the surge suppressor strip wherein the PWS and RPi 3B+ will also be connected.

All communications with the RPi 3B+ have been using my office WiFi, mostly for the convenience of being able to move stuff around un-tethered to a CAT5 cable. Once the tests are completed (?), I plan on connecting the RPi 3B+ to the network using an Ethernet cable.

Because the PWS when it's connected to the ATMOCOM, it's on a subnet of the primary network, and therefore firmware updates of the PWS from your mobile device require that your mobile device be logged into the same subnet. Once the PWS update is finished, the mobile device needs to then log off of the subnet so as to keep traffic dedicated to only your weather station. Also, it's advised to not let your mobile device 'remember' the connection in order to safeguard against an accidental connection.




Offline kbellis

  • Forecaster
  • *****
  • Posts: 357
Re: Atmocom - A beta tester's review
« Reply #1 on: May 07, 2019, 01:50:10 PM »
 [ You are not allowed to view attachments ]

Update - Web page configuration works great!

Quoting Relko Vazic:
Quote
Regarding web page, like Telnet this is only available in config mode. The internal web server which handles Telnet and HTTP client communication is only running when AP config mode is enabled. To access you must...

a) Press+hold button and let LED blink 4 times. Verify that LED now blinks on/off slowly (2-3 seconds interval).

b) Connect computer/smartphone etc. to ATMOCOM SSID. Open web browser and navigate to 192.168.4.1

c) Configure and when done, press Save and Reboot button. Disconnect computer/smartphone etc. from ATMOCOM and select "Forget network" to avoid accidental/roaming connections.

If no config changes were made it is not necessary to press Save and Reboot, just go to d)

d) Only if LED still blinks slowly, press+hold button and allow LED to blink 3 or 4 times to reboot device into capture mode.

Best regards,
Relko
--
Astrogenic Systems
info@astrogenic.com
www.astrogenic.com

 [ You are not allowed to view attachments ]

 

anything