Date   

Re: Wsprdaemon location error (mine) & noise reporting

Gwyn Griffiths
 

Hello Bruce
Your noise data are being reported to the database. There's lots of documentation on wsprdaemon.org on how to query the database, but a quick query brings up the following as a few recent lines on 40 m:

tutorial=# select * from wsprdaemon_noise_s where site like 'KX4AZ%' and band='40' order by time desc limit 15;

        time         |  site   | receiver | rx_grid | band | rms_level | c2_level | ov |  seqnum  

---------------------+---------+----------+---------+------+-----------+----------+----+----------

 2021-10-16 14:56:00 | KX4AZ/T | KX4AZ_1  | EN74gc  | 40   |   -132.02 |  -143.22 |  0 | 25616876

 2021-10-16 14:54:00 | KX4AZ/T | KX4AZ_1  | EN74gc  | 40   |   -134.97 |  -142.96 |  0 | 25616570

 2021-10-16 14:52:00 | KX4AZ/T | KX4AZ_1  | EN74gc  | 40   |   -129.19 |  -142.62 |  0 | 25615900

 2021-10-16 14:50:00 | KX4AZ/T | KX4AZ_1  | EN74gc  | 40   |   -127.36 |  -144.05 |  0 | 25615125

You will see that the site name, as set in your wsprdaemon.conf for the upload id, is still as /T and not -T in this listing.
However, it is as -T in the webpage with your noise graphs at http://wsprdaemon.org/graphs/KX4AZ/

Any rigorous calibration of antenna system and local noise is so fraught with problems and pitfalls, there is no real need to be concerned about "polluting" the noise database. If someone is really serious about using your noise data I am sure they would get in touch to learn more about your setup etc.

best wishes
Gwyn G3ZIL

 


Wsprdaemon location error (mine) & noise reporting

Bruce KX4AZ
 

After a bit of a struggle with naming conventions for my remote receiver site 'KX4AZ/T' at EN74gc, I was able to get wsprdaemon going yesterday on the Pi4, so that spots are now being reported from my new KiwiSDR. A huge thank you to Rob for creating/maintaining this software!

Alas, I just discovered that I entered the wrong location (EN73hx versus true location of EN74gc) into the conf file for wsprdaemon.  It's only about a 9 mile location error - and is the location where I am currently running a 'KX4AZ' transmitter at about 10 microwatts RF output (a separate "story").  I have just corrected the location error and rebooted the Pi.

It bothers me a bit that I have "polluted" the WSPR database with a location error like that, but at least it is not a gross location error, as in another US state(!).  But I also enabled noise reporting yesterday, first using 'KX4AZ/T', eventually changing it to 'KX4AZ-T' after realizing that the '/' character might be problematic, as it was for the receiver name.  Before I go any further, I wanted to ask if the noise data are being reported properly to the database (albeit with the location error that I have just fixed).  And secondarily, since I have not yet done any calibration for the antenna setup, I want to make sure I am not "polluting" the noise database by enabling this feature while I am still studying/learning about it.  I am happy to shut it off if that would be beneficial.


Re: HB9VQQ testing Kiwi SDR

Glenn Elmore
 

Rolf
Using "Quiet receiving site", curve C and "galactic noise" curve D from ITU-R P.372.8 as a guideline, one can probably stand a system noise more than 20 dB above KTB before it impacts SNR  from a losslessly matched antenna. This presumes that not only receiver side noise and NF but also IMD, ADC non-linearity and common-mode and similar QRM ingress is no higher.  It also recognizes that ground losses and ground noise currents can be a significant part of the feedpoint impedance and noise floor even  up at 10 MHz and particularly for horizontally polarized antennas.

If the SNR present on the antenna being used is substantially due to far-field sources, propagated signal and propagated noise, then except for IMD and ADC issues the receiver used shouldn't impact things much. In real environments of course that's the rub.  Both near-field sources and strong signal overload from out-of-band signals can make this not the case.

It does say that for a well matched antenna (well matched to radiation resistance, not necessarily the presented feedpoint resistance),  NF of the preamp and rx probably don't matter much. When one starts using poorly matched antennas, particularly a problem for electrically small antennas, then the requirements on the rx system hardware and ingress mitigation get greater.

If you are seeing appreciable difference between receivers/detectors then I suspect you have a problem with ingress,ADC/IMD and that these need to be eliminated.

Just my opinion from HF active antenna, preamp and receiver investigations...

Glenn n6gn


Re: HB9VQQ testing Kiwi SDR

KD2OM
 

Rolf,
I am using three different amplifiers built by N4CY Everett Sharp. They all are 11 dB gain. He has changed the version several times, one has a relay to disconnect the antenna when no power is supplied. I am feeding power to them using a bias T, also built by Everett. 
I do need to increase the gain for my beverage antenna, candidates are a 20 dB gain amplifier from Advanced Receiver Research and a 23 dB amplifier designed by K8ZOA from DX Engineering.

73,
Steve KD2OM 

On Oct 15, 2021, at 13:44, Rolf Ekstrand <rekstrand@...> wrote:

Hi again y'all

I have run my Kiwi  here with AM filter for more than a year now almost 24/7 and I am very happy with it.  I am only using an old ground mounted  Cushcraft R7.  I am still planning to expand with additional kiwis  for the low bands and a vertical log periodic towards middle EU.  In the meantime I like to try a LNA ( with a 50 ohm HP on let say 10 Mhz) on the vertical.     

I guess I could cobble together a Norton amp or similar and the HP filter, but before doing so,  what are the "big gun" WD kiwi stations out there using if they need a boost??

73  Rolf


Re: HB9VQQ testing Kiwi SDR

WA2TP - Tom
 

Not sure I’d consider myself a big gun, but I have had very good results with high spot counts. 

I am not using any and pre-amplification on the resonant antennas. 

The only antenna with a pre-amp here, is the wellbrook active loop. 

I have been plagued with CM paths and local noise sources, which is what I focused on.

I built my own filters for the AM BB problem stations. 

Tom
WA2TP 

On Oct 15, 2021, at 7:44 AM, Rolf Ekstrand <rekstrand@...> wrote:

Hi again y'all

I have run my Kiwi  here with AM filter for more than a year now almost 24/7 and I am very happy with it.  I am only using an old ground mounted  Cushcraft R7.  I am still planning to expand with additional kiwis  for the low bands and a vertical log periodic towards middle EU.  In the meantime I like to try a LNA ( with a 50 ohm HP on let say 10 Mhz) on the vertical.     

I guess I could cobble together a Norton amp or similar and the HP filter, but before doing so,  what are the "big gun" WD kiwi stations out there using if they need a boost??

73  Rolf


Re: HB9VQQ testing Kiwi SDR

Rolf Ekstrand
 

Hi again y'all

I have run my Kiwi  here with AM filter for more than a year now almost 24/7 and I am very happy with it.  I am only using an old ground mounted  Cushcraft R7.  I am still planning to expand with additional kiwis  for the low bands and a vertical log periodic towards middle EU.  In the meantime I like to try a LNA ( with a 50 ohm HP on let say 10 Mhz) on the vertical.     

I guess I could cobble together a Norton amp or similar and the HP filter, but before doing so,  what are the "big gun" WD kiwi stations out there using if they need a boost??

73  Rolf


Re: New install on Pi4 question

Bruce KX4AZ
 

On Fri, Oct 15, 2021 at 08:46 AM, KD2OM wrote:
I think the problem may be the names. There is no need to use a call as the receiver name.
You are correct!  Right after my last post I went back to the original example conf file example (from the web site) and changed the receiver name to remove the '/T' suffix from the call sign, while leaving it present for the uploaded spots - important for me to distinguish different locations I use.  After a restart the spots  began uploading correctly.  I also did some more editing of the example conf file to change all of the "no" values to "yes" for the noise related reporting.  Hopefully that is working correctly too - I have a lot ore learning to do to figure out how to verify that.


Re: New install on Pi4 question

KD2OM
 

I think the problem may be the names. There is no need to use a call as the receiver name. It gets the call from the config file. Your receiver names should be like Kiwi_1, Kiwi_2, etc. if is is an audio channel it must be Audio_0 or Audio_1? 
In the schedule you need to use MERGED_x for the list of receivers even if you are scheduling only only one receiver.

My names are BE, DI, VE for the receivers and the schedule list looks like this:
“MERGED_RX_1  BE                                 KD2OM FN12GX password”
“MERGED_RX_2 BE,DI,VE,AUDIO_0      KD2OM FN12GX password”


Then the schedule:
(“01:00
             MERGED_RX_1,2200
             MERGED_RX_2,20”
)


Steve KD2OM 

On Oct 15, 2021, at 06:15, Bruce KX4AZ <bruce@...> wrote:

Also, these are the current contents in the wsprdaemon.log file:

Fri 15 Oct 2021 09:40:49 AM UTC: watchdog_daemon() starting as pid 2904
Fri 15 Oct 2021 09:40:49 AM UTC: WARNING: spawn_upload_to_wsprnet_daemon() found a stale uploading.pid file with pid 1438. Deleting file /home/pi/wsprdaemon/uploads.d/wsprnet.d/spots.d/uploads.pid
Fri 15 Oct 2021 09:40:49 AM UTC: WARNING: spawn_ftp_upload_to_wsprdaemon_daemon() found a stale file '/home/pi/wsprdaemon/uploads.d/wsprdaemon.d/uploads.pid' with pid 1452, so deleting it
sed: -e expression #1, char 16: unknown option to `s'
sed: -e expression #1, char 16: unknown option to `s'
sed: -e expression #1, char 16: unknown option to `s'
sed: -e expression #1, char 16: unknown option to `s'
sed: -e expression #1, char 16: unknown option to `s'
sed: -e expression #1, char 17: unknown option to `s'
./wsprdaemon.sh: line 5375: /tmp/wsprdaemon/kiwi_gps_status/KX4AZ/T_last_gps_fixes.log: No such file or directory
cat: /tmp/wsprdaemon/kiwi_gps_status/KX4AZ/T_last_gps_fixes.log: No such file or directory
date: /tmp/wsprdaemon/kiwi_gps_status/KX4AZ/T_last_gps_fixes.log: No such file or directory
./wsprdaemon.sh: line 5381: 1634290885 -  : syntax error: operand expected (error token is "-  ")
 


Re: New install on Pi4 question

Bruce KX4AZ
 

Also, these are the current contents in the wsprdaemon.log file:

Fri 15 Oct 2021 09:40:49 AM UTC: watchdog_daemon() starting as pid 2904
Fri 15 Oct 2021 09:40:49 AM UTC: WARNING: spawn_upload_to_wsprnet_daemon() found a stale uploading.pid file with pid 1438. Deleting file /home/pi/wsprdaemon/uploads.d/wsprnet.d/spots.d/uploads.pid
Fri 15 Oct 2021 09:40:49 AM UTC: WARNING: spawn_ftp_upload_to_wsprdaemon_daemon() found a stale file '/home/pi/wsprdaemon/uploads.d/wsprdaemon.d/uploads.pid' with pid 1452, so deleting it
sed: -e expression #1, char 16: unknown option to `s'
sed: -e expression #1, char 16: unknown option to `s'
sed: -e expression #1, char 16: unknown option to `s'
sed: -e expression #1, char 16: unknown option to `s'
sed: -e expression #1, char 16: unknown option to `s'
sed: -e expression #1, char 17: unknown option to `s'
./wsprdaemon.sh: line 5375: /tmp/wsprdaemon/kiwi_gps_status/KX4AZ/T_last_gps_fixes.log: No such file or directory
cat: /tmp/wsprdaemon/kiwi_gps_status/KX4AZ/T_last_gps_fixes.log: No such file or directory
date: /tmp/wsprdaemon/kiwi_gps_status/KX4AZ/T_last_gps_fixes.log: No such file or directory
./wsprdaemon.sh: line 5381: 1634290885 -  : syntax error: operand expected (error token is "-  ")
 


Re: New install on Pi4 question

Bruce KX4AZ
 

Thanks for the suggestion, I removed the _1 in the receiver name used in the conf file, rebooted and restarted wsprdaemon.  Still not uploading any spots, though the status shows a longer list, with  _1 in the receiver name and a "never ran" error (snapshot attached).  Will continue studying the documentation from the web site, has to be something real basic that I am doing wrong.


Re: New install on Pi4 question

Jim Lill
 

I'm not sure it likes _ in the RX name or if the various people who track spots do

On 10/14/21 9:11 PM, Bruce KX4AZ wrote:
When I went back and re-read the original installation message I realized it was simply asking me for permission to create a RAM drive, so I answered YES and proceeded from there.  After editing the example wsprdaemon.conf file for my specific setup, I now have it running and can see that it is connected to the KiwiSDR.  It's not uploading any spots so far, and when I check the status, each of the five receivers says "Got PID xxx from file, but it is not running".  I am assuming I made a simple mistake somewhere along the line.


Re: New install on Pi4 question

Bruce KX4AZ
 

When I went back and re-read the original installation message I realized it was simply asking me for permission to create a RAM drive, so I answered YES and proceeded from there.  After editing the example wsprdaemon.conf file for my specific setup, I now have it running and can see that it is connected to the KiwiSDR.  It's not uploading any spots so far, and when I check the status, each of the five receivers says "Got PID xxx from file, but it is not running".  I am assuming I made a simple mistake somewhere along the line.


Re: New install on Pi4 question

Jim Lill
 

see man fstab

On 10/14/21 5:35 PM, Bruce KX4AZ wrote:
Thank you, I located the fstab file in the etc directory and will halt the wsprdaemon and add that line.....I am curious about the "0 0" after the size=300m term - what function does it serve?


Re: New install on Pi4 question

Bruce KX4AZ
 

Thank you, I located the fstab file in the etc directory and will halt the wsprdaemon and add that line.....I am curious about the "0 0" after the size=300m term - what function does it serve?


Re: New install on Pi4 question

Jim Lill
 

There is also an access speed issue that the ram disk helps.

this line needs to be added to the end of your /etc/fstab file

tmpfs /tmp/wsprdaemon tmpfs defaults,noatime,nosuid,size=300m    0 0

you may not need 300m, I use that as I have 26 channels running from 2 kiwi.  After adding that reboot and see that you that ram disk   and then I'd simply reinstall wd

On 10/14/21 4:10 PM, Bruce KX4AZ wrote:
Very new to Linux & Pi, so being very cautious...but also not afraid to ask the very dumb questions that follow below.  When I initiated a new installation of Wsprdaemon, it displayed a warning message about the /tmp/wsprdaemon directory not being configured as a RAM drive, so that every 2 minutes the WSPR write cycle will eventually wear out the microSD card in the Pi.  I decided to pause the install until I get some input.  So my questions are...
1) If I proceed as is, what might be a typical lifetime for a 32GB Samsung microSD card? In other words, is the warning a real concern for a new user, or more of a standard disclaimer ("I TOLD you this might wear out a solid state drive!").
2) Is there a VERY simple set of commands (e.g. a short narrative entitled "300 MB RAM Drive Set Up for Dummies") that I can use to create the RAM drive for wsprdaemon to use?
3) If I proceed without setting up a RAM drive, is that something I can do after the wsprdaemon installation, or do I need to I halt things and do that first?

Thanks in advance for any advice.  I WAS able to get the new Pi4 set up with a solid wifi connection, running headless with SSH & VNC access enabled, so looking forward to giving wspr daemon a try.


New install on Pi4 question

Bruce KX4AZ
 

Very new to Linux & Pi, so being very cautious...but also not afraid to ask the very dumb questions that follow below.  When I initiated a new installation of Wsprdaemon, it displayed a warning message about the /tmp/wsprdaemon directory not being configured as a RAM drive, so that every 2 minutes the WSPR write cycle will eventually wear out the microSD card in the Pi.  I decided to pause the install until I get some input.  So my questions are...
1) If I proceed as is, what might be a typical lifetime for a 32GB Samsung microSD card? In other words, is the warning a real concern for a new user, or more of a standard disclaimer ("I TOLD you this might wear out a solid state drive!").
2) Is there a VERY simple set of commands (e.g. a short narrative entitled "300 MB RAM Drive Set Up for Dummies") that I can use to create the RAM drive for wsprdaemon to use?
3) If I proceed without setting up a RAM drive, is that something I can do after the wsprdaemon installation, or do I need to I halt things and do that first?

Thanks in advance for any advice.  I WAS able to get the new Pi4 set up with a solid wifi connection, running headless with SSH & VNC access enabled, so looking forward to giving wspr daemon a try.


Re: HB9VQQ testing Kiwi SDR

Rolf Ekstrand
 

Thank you Rob

No doubt there could be a difference between a narrow band Air Spy  and the Kiwi in a band hopping application.  Another possible issue that could influence the outcome will be the splitter design on a single antenna.  

73 Rolf


Re: Web Noise Plots

admin
 

Hello,
I am not sure what you mean with changing colors but I did a small change to the Station Noise Stats page to make the colors there more "stable". Hope that resolves your issue.

Regards
Arne

On 13.10.21 21:18, KD2OM wrote:
Why do the colors of the lines change? The lines from my three receivers change, it seems to vary based on the noise level.

73
Steve KD2OM


Re: Web Noise Plots

Gwyn Griffiths
 
Edited

Dean
The link you gave is to wspr.live, which is written and maintained by Arne.
He takes the wsprdaemon noise data and provides these plots in Grafana.
Your queries are best answered by him.

If I have the right Dean Shutt, AL7CR, your noise plots on the wsprdaemon.org site are all showing at:
http://wsprdaemon.org/graphs/AL7CR/ 

Steve - the color changes are a feature of Arne's code in Grafana. Rob's note on blue and red being fixed only refers to the plots
at http://wsprdaemon.org/graphs/index.html and  http://wsprdaemon.org/graphs/yourcallsign

regards
Gwyn G3ZIL


Re: Web Noise Plots

Rob Robinett
 

There are two noise level measurement techniques reported by wsprdaemon:

In Red is the RMS noise level for the whole WSPR band taken from the 0-1 seconds and 110-120 seconds of each 2 minute WSPR cycle when all transmitters are supposed to be quiet.  However this measurement can be disturbed by the presence of an RF carrier (e.g. CW station) which persists through those measurement windows.

In Blus is a FFT noise level which is derived from the quietest 4 Hz channels of the WSPR band.  This measurement looks at the whole 2 minutes of the WSPR cycle, so it averages the noise over a much longer period but may be perturbed by transient noise events like thunderstorms.

In general the FFT measurements seem more useful and frequently show slightly lower noise levels, but we chart them both.

Gywn and Clint have written about this noise subject and you can find links to their documents at http://wsprdaemon.org/technical.html

The noise level measurement database we are building seems to be unique in the radio research world and we have started some efforts to exploit it for new insights into the ionosphere.

On Wed, Oct 13, 2021 at 12:18 PM KD2OM <steve@...> wrote:
Why do the colors of the lines change? The lines from my three receivers change, it seems to vary based on the noise level.

73
Steve KD2OM



--
Rob Robinett
AI6VN
mobile: +1 650 218 8896

361 - 380 of 856