More mapping…

I’ve spent a while this morning manually updating some log entries in Ham Radio Deluxe – all of the entries imported from last week’s RTTY Makrothen Contest which uses a 4-character Maidenhead locator code as the exchange didn’t show up correctly on my HRDLog.net map that is generated from the log data uploaded to HRDLog.net.  Seems that it doesn’t like 4-character locators so I’ve appended them all with “ll” which is roughly the centre of the square to make the 6-character locators (e.g. “KN67” becomes “KN67ll”), re-uploaded the log entries and now they appear in the right places on the map.

I’m obsessing about the map a little bit at the moment – something to do with looking for patterns within real data and the childish pleasure of putting pins in the map on far off places.

Let’s have a look at an embedded Google Maps link in WordPress:


View Larger Map

GIS for radio maps…

For a while now I’ve been enjoying the ability to export ham radio logging data from Ham Radio Deluxe Logbook direct to Google Earth for visualisation of my contacts.  However now that I have some far reaching contacts, Google Earth no longer presents a full view of radio coverage:

QGIS early days - Google EarthWhat I’d like to do now is be able to plot the same data onto a Great Circle projection map, centered on my home location to present a view of radio contacts’ direction and distance.  A friend (with a specialism in Geographical Information Systems, GIS) has pointed me at the open source QGIS software.  Some light reading around the subject suggests that I basically have two tasks to perform:

  1. Import the data as a Vector Layer into QGIS;
  2. Change the project’s Coordinate Reference System (CRS) to a great circle type projection, centred on my location.

Importing Data to QGIS

Free world map data is available here: http://www.naturalearthdata.com/  …download suitable files and import appropriate shapefiles (.shp) as new vector layers.

Export selected contacts from HRD Logbook to Google Earth (right-click, “Lookup…”) opens an exported KML file in Google Earth and the contact data appears in Google Earth’s “Temporary Places” folder.  The unmodified KML file doesn’t appear to import correctly into QGIS (the folder structure that is generated by HRD Logbook’s export seems to mask the contents) but the individual points and lines can be moved to a single new folder in Google Earth (without their individual folders) and this will save as a KML file that does import nicely into QGIS:

QGIS early days - 10 contacts imported

A slightly hacky way of making the HRD0Logbook output KML file readable by QGIS is to use a text-based ‘find and delete lines’ routine to remove the folder structure which seems to be the problem.

Notepad++ is a good tool for this kind of work.  Open up the KML file and go Search, Find… open the Mark tab and check the Bookmark line box.  Now do a Mark All for the terms “folder” and “All”.  This will bookmark the lines that are to be removed.  Now back the the menu bar, Search, Bookmark -> Remove Bookmarked Lines.  Now save the modified KML file and this should import nicely into QGIS:

QGIS early days - all contacts importedSo now I can get all of my contacts data input to QGIS from HRD Logbook without too much fuss…

Creating a Great Circle (Azimuthal Equidistant) Map

Importing a basemap shapefile or two from Natural Earth is pretty easy and gets you a world map in the standard WGS84 projection (here I have a coastline layer and a 10° graticule layer):

QGIS world map - WGS84 CRS

The next thing to do is to change the Coordinate Reference System (CRS) from WGS84 (the “normal” but distorted view we’ve come to expect from world maps) to a great circle projection, centered on my location (approximately 51°N 1°W).  My initial mistake here was to search with the terms “great circle QGIS” and this brought me lots of information about plotting great circle route lines but not maps.  I realised “great circle” is a fairly ambiguous phrase which is used in varying ways in the GIS community.  Changing the search terms to a more precise “azimuthal equidistant QGIS” quickly yielded guidance on the projection I was looking for.

In QGIS go Settings, Custom CRS… and create a new user defined CRS using the following parameters:

+proj=aeqd +R=6371000 +lat_0=51 +lon_0=-1

…where my location is 51°N 1°W, enter your lat/long figures accordingly.  This gives me the result I’m after:

QGIS world map - Azimuthal Equidistant CRSRight, next thing I need is radials and range circles…

Range Circles

I’m sure there’s an easier way to do this directly in QGIS with buffers but here’s one simple method (and probably easier to understand for a GIS n00b like me).  This online tool creates KML files of circles of a given radius and centre.  They don’t seem to import straight into QGIS as vector layers but they do open in Google Earth and that allows me to group them into a Places folders and export as a KML file that imports as vector layers (a bit laborious but only needs to be done once for any given location):

QGIS world map - Azimuthal Equidistant 1000km range circlesI’m currently looking for a way to draw some radial lines with their origin at my location…

Results

So, all this allows me to plot radio contacts on a great circle type map and helps me relate coverage to antenna configuration, operating frequency, etc…

Contacts 06-10-13 great circle plotThe use of a GIS for this task gives a great deal of flexibility on the look and feel of the output which is quite exciting and something I shall be experimenting with.  🙂

CQ WW RTTY 2013 Contest…

Stumbled on wall-to-wall RTTY on the bands this weekend and discovered the 2013 CQ WW RTTY Contest.  I’d read about the tail-end of contests being good times for QRP operators to pick up easy contacts so I quickly read up on contesting QSO format, watched for a while and set up some macros to do a bit of “search and pounce”…

By the nature of the contest the contacts were all extremely brief (and all the “599” signal reports totally meaningless) but over the few hours I found to play over the weekend I managed over QSOs with many transatlantic contacts, mostly east coast but a few further in including Alaska (more trans-arctic than transatlantic) and a very exciting contact with the far end of Texas… if that gets confirmed and the location is correct, at 5180 miles it qualifies for a 1000 miles-per-Watt award!

50 kHz of solid RTTY in the 2013 CQ WW RTTY Contest
50 kHz of solid RTTY in the 2013 CQ WW RTTY Contest

Interesting points:

  • In just a few days after the contest the rate of return of Logbook Of The World confirmations is much higher for these
    TTY contest contacts than for all my previous casual PSK31 contacts.  I guess serious contest participants are much more concerned with electronic logging and log submissions.  Useful to know if I’m going to build up the DXCC count on LoTW.
  • It was extremely useful to have a spectrum scope of each HF band from my SoftRock SDR setup as I was tuning up and down the bands with my Elecraft K2.  This gave a good awareness of the spread of the RTTY traffic across the bands and which bands had traffic on.  When a band seemed very quiet it was easy to see individual RTTY bursts on 10m appearing on the 192kHz-wide SDR waterfall and then quickly tune to them with the K2 I would have definitely missed these contacts if I didn’t have the SDR spectrum scope capability.
  • There were, obviously, a few co-site interference problems with having the wideband SDR receiver on while transmitting from the other radio and both antennas in the loft so I am now planning on building a remote HF SDR that I can locate somewhere in the local area, perhaps at a non-ham friend’s house.  Ideally somewhere close so the receiver is vaguely representative of my location and within surface-wave distance so I can monitor my own transmissions.  Current plan is to have a look at an HF upconverter for the ultra cheap USB-RTL dongle receiver and a Raspberry Pi running the show…

Loft antenna developments…

Some good developments here over the weekend.  On Friday afternoon I added new elements to a 20m dipole in the loft to give me a multi-band parallel-fed dipole covering 20m, 17m, 15m and 10m bands.  This is in addition to the ~25m horizontal loop I have around the perimeter of the loft coupled with my SGC SG-211 battery ATU at the feed point.  The current setup looks something like this:

Horizontal Loop and Multi-Dipole v2

 

The dipole elements have been spaced away from each other to minimise the interaction between them but with no particular attention to orientation – experience from the 20m dipole over the last month or so doesn’t show much of a dipole pattern so I expect the shallow-V profile and the unavoidable loft clutter (wiring, light, water tank, etc…) to make it broadly omni-directional.

Loft multi-band dipoleDipole element lengths have been adjusted (with the help of a borrowed MFJ-259B analyser) to suit the low end of the bands for data modes and it’s fed by coax to a 1:1 balun.  Performance looks reasonable so far and some QSOs have been made on all four bands including my furthest yet, 5500km deep into Russia on 17m.

Interestingly I changed some connections around to have a look at the level of co-site interference between my transceiver and the loft SDR receivers now I have two multiband antennas and noticed SWR on 20m and 17m rise above 2:1 on the dipole.  Reconnecting both antennas back to the transceiver (through a 2-way coax switch) restored SWR to below 1.5:1 on all bands, as before.  This had left one of the coax feeds unconnected and running close to the 20m and 17m elements so the dipoles appear to be very sensitive to their surroundings!

First contacts …and some pins on the map!

After months of building and tinkering and receiving, I’ve finally started to transmit at my new QTH.  Using 5 Watts from my Elecraft K2 I’ve made my first few PSK31 contacts on the 20m band using the horizontal loop I have strung around the perimeter of my loft (approximately 28m in circumference, matched by a battery operated SGC 211 auto-ATU).  First contact was a very pleasing 3000km into Russia and seeing my callsign come back to me from that far away really brought back the magic of radio!

A quick snapshot of these contacts and I shall have to figure out a good way to embed this kind of map into the website…

Contact map 1

…more contacts, another map!

Contact map 2Some more contacts made and a pattern may be emerging of single and double hop coverage…?

Contact map 3

 

Rx Antenna Switch – Interference

While I’m in the process of setting up receivers and antennas in the loft, I’m becoming a dab hand at locating sources of interference.  Unfortunately the latest gremlin is my first attempt at a USB-controlled receive antenna switch…

I’ve acquired a USB-controlled Denkovi 8 channel relay board to implement some control of things in the loft (primarily receive antennas but also thinking of other things that may want remote on/off functionality).  For the first stab at using this as an antenna switch I’ve simply wired receive antenna signal lines straight through the switch relays with no attention paid to maintaining impedance.  I’m also concious that with no real RF ground arranged in the loft space yet, ganging all the grounds (DC supply return and coax shields) together may not be the best idea, however I’ve put this configuration together in the spirit of try-it-and-see-what-happens.

IMAG0219This first setup has shown that the Denkovi relay board is a good source of interference to the signal lines running through the box.  The two obvious problems are the introduction of 1 kHz-spaced lines over a large portion of the HF spectrum when the USB lead is plugged in and an increase in the noise floor by about 10dB when the software establishes the USB connection.

Thoughts now are to acquire a pair of coaxial relays (one for HF, one for V/UHF) and use the relay board to provide USB control of those while keeping its USB lead as short as possible and paying attention to placement and shielding of the control box.  Also, now that the main components of the loft receiver setup have come together, it’s time to tidy things up a bit and have a think about an effective RF ground as the tinkering to date has shown that there is much improvement to be made in this area!

Loft PC upgrade

After some testing of my loft receiver system on my main desktop PC showed that the Loft PC was probably hitting the limits of its performance I’ve decided to do a limited hardware upgrade.  The Loft PC is made from components dated around 2005 which were mid-range back then.  Without spending too much money and keeping the Socket 939 motherboard as the core of the system, the components to take the motherboard to its maximum (what would have been high spec back in 2006) are now available 2nd-hand on eBay.  I have therefore acquired a dual-core Athlon X2 4600+ CPU and an additional 2GB of DDR400 memory as a ‘final push’ on this PC…

Passmark PerformanceTest Summary Results (performed locally, not via VNC) :

LOFT-PC (pre-upgrade, Athlon 64 3000+, 1 GB DDR400): 

LOFT PC (post-upgrade, Athlon dual-core X2 4600+, 3 GB DDR400): 

Desktop PC (Max. power saving mode, Intel i5  750 quad-core, 8 GB RAM): 

The bars show Passmark summary scores relative to my desktop PC running in “max. power saving mode” which was set when the successful SDR/VNC tests had been achieved.  The detail of the Passmark tests shows various gains from the upgrade (and interestingly an increase in memory latency on the Loft PC – due to non-matched memory module pairs?*) but this gives a feel for the relative size of the upgrade.

Most importantly, the upgrade appears to have smoothed out data mode reception on the system so I’m currently running SDR#, VAC, IPSound, Digital Master 780 and RealVNC Server with good performance – success!

Loft PC - running Softrock Ensemble with SDR# and DM780

Quick, time to make an image of the system before something breaks…

* (Later Edit) Well, my instincts proved correct!  It wasn’t long before I had a couple of reports of registry failures and some boot errors (missing/corrupt system folders and hal.dll).  After a couple of unsuccessful re-installs of Windows today I think I’ve traced it to the new RAM (2x1GB low-density DDR400) and the old RAM (2x512MB high-density DDR400) not playing nicely together.  Old RAM now removed and the system seems stable and working well at 2GB RAM.  That increase in memory latency was probably a clue to the problem.  Will re-run the Passmark tests when this fresh system has completed all of its automatic updates…

Loft Receivers v2…?

At the start of my Loft Receivers project there were two obvious approaches to the system architecture:

  1. A loft-based PC running SDR hardware and applications, access this PC via remote desktop method over network (VNC/IPSound)
  2. Loft-based SDR hardware, I/Q outputs streamed over network, SDR applications running on remote PC

Option 1 has been the focus so far due to the relative ease at which it could be assembled using spare PC hardware.  In a number of ways, Option 2 presents a more elegant system (cheaper total system cost, lower power consumption, lighter network footprint…),  but I expect it will demand much more configuration time.  In that respect, Option 1 is conceptually simple – set up a basic SDR PC then connect to it via VNC/IPSound.

In reality there has been a lot of trial and error in getting the Option 1 station up and running and I am still trying to optimise the performance for data modes reception.  With the extra overheads of running VNC and IPSound I may be hitting the limits of the Loft PC’s modest specification (Athlon64 3000+, 1GB RAM).  I had thought this would be plenty enough for the job but perhaps I was being optimistic.  If this is the case, I’m loath to spend a couple of hundred pounds on upgrading the PC’s motherboard/CPU and RAM for what was meant to be a “low cost” project…  Additionally, Spring has arrived and the temperature up in the loft has suddenly risen giving me concerns that it may be difficult to run a full powered PC up there during the summer months.  These two factors may combine to see Option 2 be taken up sooner than expected.

This could be a nice project to put my Raspberry Pi to work.  I’ve read somewhere about using a Raspberry Pi for sending the RTL-USB dongle’s I/Q output via TCP and SDR# can receive this.  The SoftRock Ensemble II receiver would need a USB soundcard plugged into the Raspberry Pi and then the Pi will need to be configured to send the line-in out over the network using Pulse Audio or something…  A quick search shows a promising guide:

http://www.crazy-audio.com/projects/raspberry-pi-for-audio-distribution/

My next steps on this project are to temporarily set the SoftRock receiver up on my more powerful desktop PC (Intel i5, quad-core, 8GB RAM…), access this via VNC and see if the performance difficulties disappear with the extra muscle.  If it continues to be a struggle I expect some RPi action to commence shortly…!

Stretch-Wire Lighting Interference

Have spent some effort this week replacing a number of halogen spotlights with low power LED bulbs. After trialling a few different types of bulbs and replacing 12V AC transformers with 12V DC power supplies I’m (finally) really pleased with the result – a pleasant, versatile lighting scheme using only 1/10th of the power of the halogen bulbs.

P1020375

During the week I also noticed a high level of broadband noise appear on my USB SDR dongle I’m running in the loft. It made sense that this might be linked with the changes in lighting. Initially the interference appeared to cycle over a period of a few seconds. That was quickly traced to the two transformers running a pair of 12V stretch-wire lighting rails across our dining room. Although the transformers were running the new LED bulbs, the slow cyclic nature of the interference suggested the transformers weren’t happy with the very low load presented so I’m quite thankful for it!

I’ve now replaced the transformers with 12V DC power supplies specifically for LED lighting. It hasn’t cured the interference (it’s now constant when these lights are on!) but at least I’m happy the new lighting is working correctly and I can control this source of interference so I can live with that. The loft SDR receiver is still a new toy and I suspect I had this interference from the wires when they were running the old halogen bulbs, I just hadn’t been looking at the radio while they were on. Here’s a screenshot of the lighting wires wiping out the VHF aeronautical band, the waterfall shows lights off (bottom), one rail on (middle), both rails on (top):

LED light interference