2017-04-06 Results from a powerfailure 2 weeks ago
Earlier today there was a local powerfailure. I noticed it while being connected remotely to my server at home, it stopped responding and stayed off-line for a while. I checked the website of the electricity distribution network and it showed a local powerfailure. Initially my reaction was a bit of interest in how low the radio noise levels would be when I could try the radio at home. By the time I got home power was back on so I never found out. Some systems at home had to be started by hand, but eventually everything was up and running again, with new kernels booted. Everything came up fine including PPPoE to the outside world. Later in the evening I noticed the old weatherstation in the shed which measures the temperature and humidity inside the shed seemed inoperational. Checking that weatherstation eventually led to a diagnoses that the 12 volt powersupply for that weatherstation had failed. It was only delivering 7.2 volt. To disconnect that weatherstation for diagnosis I completely shut down the shed computer ritchie. After that it did not come up again. Power came up and the harddisk started spinning but nothing happened after that. So I also took that computer in for diagnosis. Connected to a keyboard and monitor everything came up fine. Disconnected from the monitor nothing came up. Slowly it dawned on me that this might be related to an issue which I had read on the PC Engines Alix.1d page :BIOS update * beta various updates (release MFGPT timers)Which seemed to be the problem which suddenly started acting up for me which it never did before. But the BIOS clock was also reset so I guess the whole setup was erased and I just did not notice because the default settings were good enough to boot anyway and ntpd sets the clock soon after boot. A simple solution would be a BIOS upgrade. But preparing a CF card to boot freedos seemed a long way to flash the bios so I tried it via the Heavy Duty Boot pxeboot environment which booted FreeDOS nicely on the alix.1c but it rebooted as soon as I started SB.COM. I guess sb.com searches for bios images on C:\ and the FreeDOS floppy image is A:\ In the end I searched for a workaround and found mention in PC hangs up when no monitor attached - Ubuntu bug #243257 of a wire between pins 12 and 7 in the VGA connector. A piece of wire was bent and inserted into the vga connector and now the weatherstation computer ritchie boots up again. Next plan is to do the BIOS upgrade via flashrom: In searching for information about the Alix 1.c bios upgrade I'd like to do I came across mention of flashrom.
* beta fix VGA DDC issue (boot hang)
2017-02-23 Ok, not using the available wind meter readings in the neighbourhood 2 months ago
Today is very windy in the Netherlands, with storm with wind gusts of up to 65 km/h and average windspeed of 36 km/h. But the readings shown by the weather station readings shared in the neighbourhood show quite different values:2017-02-23 21:49:14 : HIDEKI Wind sensor Rolling Code: 15 Channel: 4 Battery: OK Temperature: 4.9 C Wind Strength: 15.29 km/h Direction: 337.5 °So I don't think those will help a lot for my Weather station Utrecht Overvecht.
2017-01-27 Nice neighbours sharing their weather readings 2 months ago
A posting about reading 433.920 MHz signals triggered the idea I had ages ago to decode those signals and see what weather stations are available nearby. The original posting 433,92Mhz ontvangen (Dutch) was about receiving remote controls (KlikAanKlikUit) and had a screenshot of some Linux software for receiving those signals but no name of the software (that would be useful information). But a simple google search found me rtl_433 on github which receives and decodes all kinds of signals on 433.922 MHz. I downloaded it on the raspberry pi for radio experiments, and it is working fine receiving weather information from probably nearby weather stations. At least one outside temperature and humidity sensor, one inside temperature and humidity and one wind and temperature sensor. This last one could be nice for my weather station!2017-01-27 21:00:27 : HIDEKI Wind sensor Rolling Code: 15 Channel: 4 Battery: OK Temperature: 3.6 C Wind Strength: 5.31 km/h Direction: 67.5 °and a rain sensor:2017-01-27 21:01:05 : HIDEKI Rain sensor Rolling Code: 0 Channel: 4 Battery: OK Rain: 648.2 mmThanks for sharing, neighbours!
2016-04-01 Forcing a dual-band wireless card to 2.4 GHz channels 1 year ago
The wireless card of the weather station computer in the shed is dual-band but with only a 2.4 GHz capable antenna. Since the house access-point is configured to support both 2.4 GHz and 5 GHz channels the system sometimes selects the 5 GHz access and keeps having serious packet loss. I looked at ways to convince the driver to select 2.4 GHz channels only but found none, but then I found out wpa_supplicant can do this. But I configure wpa_supplicant through wpa-* options in /etc/network/interfaces so I had to find out how to configure it using those. The manpages for the interfaces file is very limited on the wpa-* options, but I found an explanation that a lot of wpa_supplicant options are supported, including the one to select frequencies. The sneaky part is that the option in wpa_supplicant.conf is freq_list and the option in /etc/network/interfaces is wpa-freq-list. A rather complete list can be found at Where can I find a full list of wpa-* options for the interfaces file? - superuser.com. So now I have in /etc/network/interfaces:auto wlan0 iface wlan0 inet dhcp wpa-ssid default wpa-psk VerySecret wpa-freq-list 2412 2417 2422 2427 2432 2437 2442 2452 2457 2462 2467 2472The ideal solution is to order a dual-band (2.4 GHz and 5 GHz) antenna. Update: Noticeable absent are channels 12 and 13 which are available for regulatory domain NL but are not listed when I ask the driver for available channels:koos@ritchie:~$ /sbin/iwlist wlan0 chann wlan0 19 channels in total; available frequencies : Channel 01 : 2.412 GHz Channel 02 : 2.417 GHz Channel 03 : 2.422 GHz Channel 04 : 2.427 GHz Channel 05 : 2.432 GHz Channel 06 : 2.437 GHz Channel 07 : 2.442 GHz Channel 08 : 2.447 GHz Channel 09 : 2.452 GHz Channel 10 : 2.457 GHz Channel 11 : 2.462 GHz Channel 36 : 5.18 GHz Channel 40 : 5.2 GHz Channel 44 : 5.22 GHz Channel 48 : 5.24 GHz Channel 52 : 5.26 GHz Channel 56 : 5.28 GHz Channel 60 : 5.3 GHz Channel 64 : 5.32 GHz Current Frequency:2.462 GHz (Channel 11)And now I wonder why those are missing.
2015-07-29 Finally a nearby thunderstorm 1 year ago
Finally a thunderstorm passed our house today, and it got counted on the lightning strike detector! The downside is that switching the light on in the shed a few times causes more 'hits'.
Lightning strikes counted today
2015-01-05 A too cold harddisk 2 years ago
A problem I did not expect to see: a harddisk getting close to being too cold. This disk in the weather station computer in the shed is a laptop harddisk and according to the specs the minimum operating temperature is 5 degrees Celcius, which was reached last week when the shed dropped just below 2 degrees Celcius.
2014-12-16 Adding my weatherstation to the APRS maps 2 years ago
I still had the wish to submit the Weather station Utrecht Overvecht to the APRS network to get the data on the APRS maps and maybe even added to the Citizen Weather Observer Program. APRS stands for Automatic Packet Reporting System which allows lots of information to be automatically exchanged. The most well-known one is position reporting, but there is a lot more, including weather stations. It wasn't very hard to get this done, I just had to find the time to do it. The trigger was the mention of a new APRS server in the Netherlands at Amsterdamse APRS server - PA4TW. I played a bit with the perl modules Ham::APRS::IS and Ham::APRS::FAP. Ham::APRS::IS is for communicating with the APRS-IS network. And Ham::APRS::FAP is for parsing APRS packets. I used this module to make very sure my outgoing packets parse correctly before actually submitting them to the network (and still making a few mistakes in the process). I used the explanation and example at APRSWXNET information by Philip Gladstone to find out what to submit and in which format. The interesting point is that wind measurements can not be omitted, but since I don't have a working wind measuring setup I have to enter the '...' values to show that the measurement is not available. And temperatures are measured in Fahrenheit so I have to convert first from my measurements. The current outgoing packets look like:PD4KH-1>APRS,TCPIP*:/161010z5206.6 N\00507.1 E_.../...g...t045h91b10088http://weatherstation.idefix.net/And now the weatherstation shows fine: PD4KH-1 weatherstation on aprs.fi and Weather stations near PD4KH-1 on findu.com. Note how aprs.fi converts to degrees Celcius and findu.com stays in Fahrenheit.
2014-11-28 Moving the lightning strike detector to the shed 2 years ago
I have noticed the lightning strike detection in Weather station Utrecht Overvecht goes completely mad when I transmit on the 20 meter amateur band. With the detector being quite close to the antenna I can understand this. The solution will be to find a place to mount the detector in the shed. It will be lower (less reception of the radio energy of the strikes) but it will also be further away from my interference. That also means the reading of the detector will have to be done using w1retap since that is what I use on the shed weatherstation computer. I was a bit confused whether w1retap supports this counter but I found out it's based on the DS2423 counter chip which is supported in w1retap, as part of a wind speed meter in a TAI8515 weather station, but w1retap will give the count on readout and the conversion is up to the user.
2014-11-07 (#) 2 years ago
Interesting new problem: transmitting on 20 meter PSK31 causes the lightning strike detector of the Weather station Utrecht Overvecht to go slightly mad. If you look at the graphs it seems like there was a really heavy thunderstorm last Thursday but that's just the result of trying 20 meter PSK31 all day and making 11 QSOs (radio connections). Time to move the lightning detector to a better location: in the shed with the rest of the weather station. Update 2014-11-08: And another side-effect: I used a really cheap gimmick USB hub to hook up the radio to the workstation because I needed an extra USB port for something else. The next thing that happened was the USB chip in the serial cable to the radio (CAT interface) completely locking up on transmitting. I also hear buzzing on speakers when I transmit on 20 meters so this is a big hint I need to change some things. At least a balun for the attic dipole is on the way.
2014-05-25 (#) 2 years ago
After testing the gps sky view it's now time to test with ntpd. First step was to recompile ntpd because the debian default package had no pps support. Recompiling on a 500 MHz AMD Geode takes a bit of time. Results look ok for a first test:root@ritchie:~# ntpq -p remote refid st t when poll reach delay offset jitter ============================================================================== +greenblatt.idef 18.104.22.168 2 u 17 64 377 1.177 101.732 63.403 *metronoom.dmz.c .PPS. 1 u 17 64 377 19.499 100.512 6.722 +auth1.xs4all.nl 22.214.171.124 2 u 11 64 377 18.403 104.008 3.669 oGPS_NMEA(0) .GPS. 0 l 6 8 377 0.000 114.073 7.364 root@ritchie:~# ntpdc -c loopi offset: 0.001986 s frequency: 94.434 ppm poll adjust: -30 watchdog timer: 342 s root@ritchie:~# ntpdc -c kerni pll offset: 3.3e-08 s pll frequency: 94.434 ppm maximum error: 0.175258 s estimated error: 2e-06 s status: 2007 pll ppsfreq ppstime nano pll time constant: 3 precision: 1e-09 s frequency tolerance: 500 ppm root@ritchie:~# ntpdc -c sysi system peer: GPS_NMEA(0) system peer mode: client leap indicator: 00 stratum: 1 precision: -19 root distance: 0.00000 s root dispersion: 0.00749 s reference ID: [GPS] reference time: d72cb010.dc91595e Sun, May 25 2014 20:08:16.861 system flags: auth monitor ntp kernel stats pps jitter: 0.006714 s stability: 0.000 ppm broadcastdelay: 0.000000 s authdelay: 0.000000 sIt will need some more calibration probably. Update: It keeps looking nice after some calibration. Stats gathered at NTP server ritchie.idefix.net stats. This does mean one of the old project sundial goals has been met: the weather station computer in the shed is now also a time server.
2014-05-07 (#) 2 years ago
Free unscheduled UPS test this morning courtesy of the local electricity company this morning. As logged by the UPS:Wed May 07 08:47:09 CEST 2014 Power failure. Wed May 07 08:47:15 CEST 2014 Running on UPS batteries. Wed May 07 09:27:39 CEST 2014 Battery power exhausted. Wed May 07 09:27:39 CEST 2014 Initiating system shutdown!The weatherstation computer for Weather station Utrecht Overvecht doesn't power up automatically, so it's not available at the moment. It's annoying that my websites are unreachable and Internet access is down. But no money is lost, so there is no reason to invest in backup links and resilient hosting.
2014-02-21 (#) 3 years ago
I almost started to digres in my last post getting an APRS report out using an android device, aprsdroid and a wouxun radio about the privacy implications but decided to separate that issue. It works, and I am not going to invest in it. APRS is nice, but for me it has privacy implications: it reports my position in real-time which is more than I want to share with the world, even as an amateur radio experiment. I know there are people who will post their bicycling tours or runs on-line as they happen but I don't like publishing my wereabouts, especially not in (near) real-time. Enough people can browse the mobile telecom location registers as it is. So I think I'll leave the APRS location experiments at this and I'll go look at other amateur radio stuff. There is enough to play with! I know APRS is bigger than just reporting location. Getting my weatherstation in Utrecht Overvecht to report weather to the APRS citizen weather observer program is on my wishlist.
2014-02-03 (#) 3 years ago
In searching for information about the Alix 1.c bios upgrade I'd like to do I came across mention of flashrom, a command line utility which can do firmware flashing for a lot of systems right from the commandline. No more long downtimes due to being busy with bootable floppies, cd's, usb sticks or slow update programs. Initially the program gave a warning on ritchie, but with one "I am sure what I am doing" option:root@ritchie:~# flashrom --programmer=internal:laptop=this_is_not_a_laptop -r oldbios.bin flashrom v0.9.6.1-r1563 on Linux 3.12-1-486 (i586) flashrom is free software, get the source code at http://www.flashrom.org Calibrating delay loop... OK. ======================================================================== WARNING! You may be running flashrom on an unsupported laptop. We could not detect this for sure because your vendor has not setup the SMBIOS tables correctly. You can enforce execution by adding '-p internal:laptop=this_is_not_a_laptop' to the command line, but please read the following warning if you are not sure. Laptops, notebooks and netbooks are difficult to support and we recommend to use the vendor flashing utility. The embedded controller (EC) in these machines often interacts badly with flashing. See http://www.flashrom.org/Laptops for details. If flash is shared with the EC, erase is guaranteed to brick your laptop and write may brick your laptop. Read and probe may irritate your EC and cause fan failure, backlight failure and sudden poweroff. You have been warned. ======================================================================== Proceeding anyway because user forced us to. Found chipset "AMD CS5536". Enabling flash write... OK. WARNING: unexpected second chipset match: "AMD CS5536" ignoring, please report lspci and board URL to firstname.lastname@example.org with 'CHIPSET: your board name' in the subject line. Found SST flash chip "SST49LF040B" (512 kB, LPC) at physical address 0xfff80000. Reading flash... done.Checksums for the saved bios file match with the bios image I have from the last upgrade.
2014-02-03 (#) 3 years ago
I had a closer look at the available PC Engines alix 1.c bios updates and noticed alix0.bin and alix2.bin images in the .zip file. A bit of searching found 3D3 BIOS-Upgrade: Which one is the right file? - PC Engines support forum which lists alix0.bin for alix.1 systems and alix2.bin for alix.3 systems. Should work with the provided sb.com.
2014-02-02 (#) 3 years ago
The change to the new weatherstation computer also means the powerline network is now gone. And I mean gone, not just disconnected "just in case". The adapters are back in a crate. And this should mean the HF spectrum should look nicer than in my earlier measurements. Future measurements may confirm this. The weatherstation computer now uses Wi-Fi to talk to the rest of the network. Since the access-point is on the same ground level the signal quality is good and speed is high enough.
2014-02-02 (#) 3 years ago
Today there was time to change the computer in the shed to ritchie, the alix 1.c based system. The system decided this was a great moment to run into the "system won't boot without monitor present" issue. I found a funny workaround: I disabled the video memory. This does mean the display doesn't work at all, but at least it boots. I'll do a bios reset and upgrade sometime in the future. The PC Engines alix1c product page lists possible problems with a vga boot hang as fixed in a newer bios. After that the system started working nice, talks to the weather station hardware and Weatherstation Utrecht Overvecht is back up. This system with a standard power usage of about 5W will be 'greener' than the replaced older PC. There is a free serial port for a gps, but since the gpskit gps was showing strange problems in the ntp server test I decided to take that gps back inside to have a look whether that can be used or I want to search for some other gps unit to have my own home time server. Maybe the wardriving gps unit. Some sort of plastic container to keep all the electronics together and shielded is also a good idea.
2014-01-06 (#) 3 years ago
Searching for the print header - COM port cable in the PC parts junk box for the new weatherstation on the alix.1c board gave 5 of them, 2 with db9 connectors and 3 with db25 connectors. I tested them all with a real telephone modem. In total 3 worked, 2 didn't but that may be because there are two different standards for it.
2014-01-01 (#) 3 years ago
Current todo list new weatherstation installation on ritchie:
That should get things moving to a level comparable to the current setup. And the wishlist for improvements:
- Set up udev rules for libusb access by non-root user so w1retap can run. Done, via a w1retap group. ✓
- Get w1retap to run on the system. Done ✓
- Get backups running using the amanda system. Done ✓
- Add serial cable from alix1c COM2 header to outside. I found out that there are two types of cables for this and the "Intel" version won't work. So if the PC parts junkbox has these cables they still need to be tested. Done ✓
- Get the conrad serial port weather station running on that port. Done ✓
- Something to put the system in so it is in a good place in the shed and somewhat shielded from temperature/humidity influences.
- System statistics gathered. Wishlist: cpu load and states, processes, temperatures and voltages mainboard, temperature disk, entropy, ntp stats. ✓
- Logging done in a way data will be on the home server and interruptions of the network link will be handled. I'm starting to use the term telemetry to describe this process of gathering timestamped data, preparing it for transmission over maybe flaky network links and processing it on the server. Partly done, some data is still gathered in the 'old' style.
- A working gps receiver so it can work as a standalone time server.
- Get the rain meter of the conrad weather station set up.
- Get the wind speed meter (anemometer) of the conrad weather station set up.
- Get the wind direction meter set up.
- Upload the available weather information to the CWOP (citizen weather observatory program) using APRS-IS.
2013-12-30 (#) 3 years ago
First attempt to reinstall the alix.1c system named ritchie as the 'shed' computer. Since I had a lousy image on a lcd monitor I chose 'flat panel' in the setup which gave me no image at all. So I had to look in the PCEngines Alix.1c manual to see which jumper resets the bios settings back to failsafe defaults. The debian installer boots fine, detects the hardware, uses a good framebuffer driver which gives a crisp image on the lcd monitor and even finds the wireless card (atheros chipset, ath5k drivers). I decided to select 'wlan0' as network device for the rest of the install and it is working fine after typing the complete wpa2 key for our home wireless network. The box contains a 40 Gb ex-laptop disk. Guided partitioning suggested a 500 Mb swap partition which should work fine. At the moment there isn't a specific Geode kernel loaded, I'll see after reboot whether one is available or has to be built. If I need to built stuff I will do the compiling on a different system with more cpu. Update: The installation did not save the wireless config so I had to set it up again. This was easy with the example in Wifi/HowToUse debian wiki. IPv6 connectivity is no problem at all so I can now log in via ssh and do the rest remote.
2013-12-27 (#) 3 years agoOlder news items for tag weatherstation ⇒
When I get around to wardriving lately I use the wigle app for android, so I decided to repurpose the wardrive box as shednet computer. This will also mean I can stop using PLC network to link the shed computer to the rest of the network, replacing it with wifi. Using PLC offends the radio amateur in me and people at the radio club now ask me whether I dumped it already. So, time for different hardware which as a plus will also use less power. I already built in a laptop harddisk in the wardrive box and removed the compact flash holding the wardriving box image (I'll save it so I can restart using the wardriving box if needed). The alix 1c can work nicely with an IDE laptop harddisk.The installation will probably be done with a recent debian jessie distribution. Plans for the new shednet computer include an ntp server, but since the current gps unit (gpskit) seems to fail after running for a few days in the test setup that may have to wait.
The laptop harddisk built into the ritchie mini-itx case