KiwiSDRs reactivated

After a holiday, and little amateur radio activity, I have reactivated the 2 KiwiSDRs from the home location.

The performance is not as good as at the previous remote and RF quiet location, but is adequate enough for general listening and will provide further opportunities to pursue noise reduction strategies.

KiwiSDR1 is currently devoted to WSPR, KiwiSDR2 is available for public access from the top menu.

I have suspended 70 MHz reception for the moment to free up an antenna port for a KiwiSDR – activity levels have been very low on this band so far this season.

9K2GS received on 50 MHz FT8

The consensus among the VHF community that this week would see the real start of elevated conditions (particularly for those of thus in more northerly latitudes) has been confirmed by regular openings to the Mediterranean countries. This morning, Friday 4th May 2018, the first medium distance DX, 9K2GS, was received on FT8 mode.

Remote site issues and 2m FT8

The recent inclement weather in the UK caused a number of issues at the remote site which hosts the KiwiSDRs and other receivers.

I had originally listed here a number of issues which had to be overcome, such as the breaking of the 6m/4m mast in the wind, but recent events have made these points redundant.

I have taken down and dismantled the remote site antenna and all receivers due to the termination of the operating agreement by the site owner. The search for a new site will begin shortly.

The 2m antenna has been relocated to the home location (IO83LS).  The breaking of the mast was fortuitous as I was not happy with the performance of the antenna at the remote site and having retreived it safely I now had an opportunity to test it in a different environment. It was mounted on a pole about 3m above the ground, just below the 6/4m duoband Yagi.

Temporary VHF setup, 2m 4 Element Quad, 4/4 Element 6/4m Yagi

After reading the good results that G3XBM was getting on his blog, I tuned to 144.174 MHz,  FT8 mode, using WSJT-X 1.9 rc2.

F1DRN received QRB 1253 Km on 2m FT8

Given the modest antenna – a 4 Element Quad at 3m AGL, no receive preamp or power amp, and an IC7100 transceiver there is a surprising amount to be heard. I’m pleased I gave it a chance.

Aircraft scatter is common, and sometimes helpful:

GM4JJJ on 2m FT8, with Aircraft Scatter

K3RWR received on 2200m using WSPR-15 mode

K3RWR (Hollywood, Maryland, FM18QI) was decoded last night on 2200m (137.612 kHz) by G0LUJ (Bowland Forest, Lancashire, IO83QV) using WSPR-15 mode. It is hoped that there will be further tests with slow data modes – it would be ideal if more slow modes were implemented in the current release of WSJT-X as this would encourage their use.

K3RWR trace visible at 04:00-04:15

Adding 6m to the KiwiSDR using a transverter

The second Kiwisdr at the remote location has been under-utilised, so a re-purposing was considered. The plan was to purchase a 50 – 28 MHz transverter with minimal complications (as transmission was not required) and interface it with the KiwiSDR.

A search on uncovered a circuit board from a source in Ukraine. At a total price with postage of less than £20 it was good value, and arrived from Ukraine in just over one week.

Connecting up was simple – Pins 1 and 2 to the KiwiSDR antenna input, 7 and 8 to the 6m antenna and 5 and 6 to to the 12V supply.
To use the transverter, the KiwiSDR ‘Frequency Scale Offset (kHz)’ box in the config page of the administrator settings must be completed with the correct offset, in this case 50000-28000 kHz = 22000 kHz. Also, the ‘Max receiver frequency’ dialog was changed from 30 to 32 MHz to allow tuning from 50 to 54 MHz.

Initial testing began with WSPR transmissions from the home location to determine frequency accuracy and stability, and the sensitivity of the new combination.
The test transmission was conducted at a power setting of 0% on the transmitter. An estimation of the power would be 1W or less, over a 50 Km path. The WSPR signal was decoded and the drift was well within limits. The stability and sensitivity are up to standard.
The frequency accuracy is almost there, but not quite. The receiver is tuning about 500 Hz low. So, in the above example, after the WSPR decoder is started the frequency in the bottom right dialog needed to be changed from 50293.00 to 20292.50 to correctly position the signal. That being said, for most purposes being 500 Hz out in frequency will not make much difference. As the transverter is stable – and turning trimmers on the board may affect that – the first approach is to see if the KiwiSDR software will allow greater granularity on the frequency offset to correct it that way, which is work in progress.
With the caveat of that small amount of unfinished business, the implementation of the transverter/KiwiSDR project was a low cost and quick operation which has produced something which will be useful to observe 6m band activity as conditions improve.

The KiwiSDR as source image for a 10m grabber

Using spare capacity on the KiwiSDR observations were made on the beacon area of the 10m band. Over a couple of days any CW beacons heard were marked on the KiwiSDR using the instructions below:

It soon became apparent that the KiwiSDR, even fed with a low band antenna such as a Wellbrook Loop, was highly effective at receiving European beacons on 10m during small Sporadic-E openings.

A KiwiSDR session was initiated on a spare computer and a snapshot is taken every three minutes. It is available here:

Testing the Red Pitaya on 70 MHz (4m)

The Red Pitaya should not be able to receive signals on the 70 MHz band. The user manual states a range of DC-62.5 MHz. That being said, I decided to ignore the manual and perform some tests using WSPR to see if any results could be obtained.

Firstly, the write-c2-files.cfg file in Pavel Demin’s WSPR transceiver software was modified to add a line which included a centre frequency of 70.0925 MHz (70.091 + 1500 Hz).

Modified write-c2-files.cfg file

WSPR transmissions were started from the home location (IO83LS) which is 29 km from the receive site (IO83QV). In the absence of a 4m receive antenna, the 6m antenna was used. Despite this, multiple decodes were achieved.

70 MHz WSPR transmission successfully decoded

To get additional confirmation, further tests were conducted with Bri, G0MJI, near Liverpool.

G0LUJ received by the Red Pitaya of G0MJI

The ability of the Red Pitaya to receive on 70 MHz will be useful in the forthcoming tropo and sporadic-E season as CW Skimmer Server will be able to continually scan the band for beacon and other CW activity and provide early indication of elevated conditions.

Searching historical WSPR data using a Python script

The historical records held by WSPRnet are a valuable resource for analysis and data-mining, however the file sizes and number of records make data manipulation difficult and time-consuming.

Last year, I noticed a post by I2GPG – WSPR Log file creator – which used a small python script to search for and extract records from the gzipped monthly files that matched a particular callsign, producing more manageable file sizes ready for importation into spreadsheet applications. KI7MT has been developing something similar: KI7MT/wspr-ana, and there are probably many other approaches.
If you wish to try a search using I2GPG’s script (Windows):
1) Download and install Python 2.7 if required – link.
2) Create a ‘C:wspr’ directory.
3) Download as many .gz monthly files as required from WSPRnet into ‘C:wspr’.
4) Download I2GPG’s python script to ‘C:wspr’ – (save as) link.
5) Run the python script according to the instructions from I2GPG.
Alternatively, I have modified the script to allow separate receive and transmit callsigns to be entered, allowing more targeted search – (save as) link.
Search in progress
Completed .csv file viewed in LibreOffice