QDFA Antenna Countdown, 7 Days Remaining

Only a week now until OJ Sagdahl and I will test run the QDFA antenna. All primary and secondary main supports are up. Four 200-ft lengths and one 300-ft length of twin-lead wire have been cut and ends tinned. Two 17Ah batteries for the QDFA preamp are fully charged. The four 60-ft horisontal lengths of the deltas are cut and will be mounted today.

And I got my second Perseus SDR a week ago.

KGIM 1420, KSDN 930 Aberdeen SD Verification

Back when I was hunting veries, I sent a report to KGIM-1420 and KSDN-930, both in Aberdeen SD. They didn't respond at the time, a pity since I had heard KGIM only once and KSDN wasn't particularly easy to hear either. A follow-up yesterday was quickly taken care of though, and a friendly confirmation was emailed back. Both reports are from October 2007, although I heard KSDN on a number of occations in 2008 and 2009.

All those who believe that Aberdeen is a junction in South Dakota should know that the city is the county seat of Brown County, has a population of around 25,000, is susceptible to flooding and that a hospital is the largest employer. If we dare to trust Wiki. Of course, Aberdeen is also a junction in South Dakota.

Running The QS1R SDR On a Netbook

In previous postings, I have tested the SDR-IQ and the Perseus on the MSI Wind netbook, to see if relatively low powered netbooks can run processor-intensive software like software defined receivers. I was able to run two parallel instances of the SDR-IQ, and I also managed to run the Perseus in 1600 kHz recording mode when I overclocked the processor. All well so far.

So what about the QS1R? It appears to demand a little more from the processor than the Perseus, but the graphical user interface is responsible for most of the load. Once the GUI is minimised, the QS1R runs very light indeed. Otherwise, CPU load is around 20-40 %. I tried to lower the Panadapter Update Rate to around 5 fps and that may have caused a slight reduction in CPU load.

The GUI is scaleable, which is an advantage over Perseus with its fixed window size. Perseus is a little too large for the Wind's 1000 x 600 display, while QS1R can be adjusted to fit.

Since a recording function is not yet implemented, I don't know how well the Wind will handle a QS1R in recording mode.

The picture shows the QS1R in demod mode. The conclusion is: There is no problem running a QS1R with a netbook like the MSI Wind or similar. The processor is an Intel N270 Atom running on 1.6 GHz. This, and newer Atoms with higher speeds, are found in many netbooks nowadays.

QDFA Project Underway

The Quad Delta Flag Array (QDFA) which was tested with great success in Grayland, WA is going to be tested in Kongsfjord too this autumn. Supports are being bought and built, and 10 (!) 5-metre fishing rods were bought at a Finnish travel fair a few weeks ago. And the "Central Processing Unit", the phaser, arrived from the US today, thanks Dallas. Two 17Ah batteries are on their way, to power the 30mA Norton push-pull preamp (see photo). And lots of other stuff needs to be purchased, not least proper weather insulation. The Arctic winter only meters on shore is a hostile environment. Needless to say we are very excited as to how well this new antenna design will perform here.

The air photo shows the layout of the QDFA as a red line. My cabin is on the right.

July 11 Update! A high-res picture of the phasing box is here.

Kongsfjord - An NDB Hunter's Heaven?

I did some NDB DX some years ago, but I've done very little of it lately. I know that great NDB DX is possible though, since I've heard several 50-watt Papua New Guinea stations in the 1600-1700 kHz range.

One of my January Perseus files contained a recording down to 200 kHz, and well-known NDB DX-er Roelof Bakker (NED) offered to compile a log of what was heard on Jan 10 at 1430 UTC. The list left me stunned. I need to take up this part of the hobby...

Roelof made a very detailed report which you will find here. All from a few seconds, one late afternoon when conditions weren't really out of this world. One can but wonder what can be heard with a little more effort. Thanks very much for your detailed report, Roelof!

Sensitivity Measurements - Bjarne's Receivers

I’ve owned a few receivers over the years. I’ve measured the MW sensitivity on most of them, and on one I borrowed.

Receiver	500 kHz	1000 kHz	1500 kHz
Telefunken E-1501[1]	-108 dBm	-112 dBm	-112 dBm
Racal RA6790/GM	-108 dBm	-109 dBm*	-109 dBm
EAC R-390A	-116 dBm	-116 dBm	-112 dBm**
JRC NRD-525	-112 dBm	-114 dBm	-115 dBm
Icom IC-R75[2]	-111 dBm	-111 dBm	-110 dBm
Icom IC-746Pro[3]	-112 dBm	-112 dBm	-111 dBm
Icom IC-703[4]	-110 dBm	-113 dBm	-113 dBm
Icom IC-7000	-102 dBm	-109 dBm	-111 dBm
Icom IC-R1500	-96 dBm	-96 dBm	-96 dBm
Etón E1	-112 dBm	-114 dBm	-115 dBm
SDR-IQ	-101 dBm	-101 dBm	-101 dBm
Perseus[5]	-105 dBm	-103 dBm	-104 dBm
QS1R/QS1R ext.preamp[6]	-93 dBm/-105 dBm	-94 dBm/-105 dBm	-95 dBm/-106 dBm

*Measured on 1100 kHz. On 1000 kHz there was a tone (spurious signal?) that prevented measurements.

** Surprisingly low. Measurements on 1400 and 1600 kHz showed -114 dBm sensitivity.

All measurements made with 400 Hz tone, 30% modulation on AM with AGC off and "wide" bandwidth (nominal 6 kHz except RA6790/GM; nominal around 4.5 kHz, and Etón E1, nominal around 7 kHz).

And the winner is… not very surprisingly, the EAC R-390A, but with the NRD-525 a very close second. At most locations, sensitivity is not always an issue. But at my location, sensitivity certainly is. The 525 is no more in my possession, but the R-390A will stay.

The SDRs lack a few dB to be up front, but they have other virtues. There will never be a perfect radio.

---

[1] Attenuation pad removed below 1.6 MHz

[2] Fully Kiwa-modified

[3] Attenuation pad below 1.8 MHz removed, and preamp enabled

[4] Attenuation pad below 1.8 MHz removed

[5] Preamp on/off, dither off, preselector off. Other settings give less sensitivity. 500 kHz reading actually 800 kHz.

[6] A Wellbrook, 11 dB (nominal) preamp was put in the signal path to compensate for the QS1R not having an internal preamp (unlike the other receivers)

Another Kongsfjord Perseus File

This Perseus file is recorded on Jan 10, 2009 at 1430 UTC. This is rather late in the day, with  the eastern half of North America already sunlit, but some of the more dominant west coast MW stations can still be heard, together with several Pacific Canadian NDBs in the LW band. The half hour is full hour in India and Iran, and there should be a few stations from that region as well. Iran 702 and India 1566 were certainly very strong.

SpectraVue Beta Reads 1600 kHz Perseus Files!

With the exception of Winrad, SDR software has been proprietary to the hardware. SpectraVue for the RF Space SDRs, and Perseus for the - you guessed it - Perseus SDR. Finnish DX-er Tarmo Kontro alerted me about SpectraVue 3.0 beta 19 which is in fact able to read 1600 kHz Perseus files! The beta 16 could read 800 kHz Perseus files, but I was hoping Moetronix would develop Perseus support further. And indeed that happened! Above is an image of SV 3.0 beta 19 running a 1600 kHz Perseus file from October 2008 (incidentally just as I heard a KVXR ID on 1280). The software, although in beta, appeared to be very stable.

Now, this seemed to be good news also for other users. Netbook owners! The Perseus software is a fixed window 1024x648 (?) pixels, while most netbooks only run 1024x620, or even 1024x578 pixels. With 620 lines, running Perseus is sort of OK if you move the window a bit upwards and allow the task bar to be hidden. But SpectraVue is scaleable, leaving you in better control over the task bar and the desktop. I tested a 1600 kHz file on my MSI Wind running at turbo mode (1.98 GHz instead of the standard 1.6 GHz), and it was running very smooth, although with a high (90%) CPU load. As was pointed out by Anders Hultqvist though, a "chopped" Perseus file (using Chuck Hutton's excellent "WavChopper" utility) will not play at the utility's current state.

SpectraVue 3.0 beta 19 is [Update: Was!] available from this site. Only the latest beta is downloadable, and there is no guarantee that following betas will support Perseus. Also, I am quoting the usual caveat: If you are not comfortable un-zipping and dealing with bugs and program crashes PLEASE do not download these Beta versions as they are continually changing and can be unstable or non-working.

Update: Well, Spectravue was "open" just for so long. The current beta on the link above is 20, which has a new limitation: "Limit Perseus File playback to SDR-xx owners". So there you go. No SDR-IQ or SDR-14, no Perseus playback. This posting has been changed accordingly.

Update: No, I will not distribute the beta 19. No more requests please.

New Kongsfjord Perseus File

Since the first Kongsfjord Perseus file was downloaded 188 times before I removed it, many seem to have found an interest in hearing how arctic dx is like. So I thought I'd keep you lads busy a bit more.

This file is from Feb 26 at 0600. Conditions towards North America were rather undefined with signals stretching from Ontario/Ohio to the west coast. A few Mexicans too. Again, most Europeans are Russian stations. Note 1485 - NRK Svalbard with regional programming originating from Tromsø before national news on the hour. 1485 is my indicator station as it is located in the signal path. Again, this file will only be available for a few days or perhaps a week.

The file is zipped from original 686 MB to 488 MB. Enjoy!

Kongsfjord DX Files

Update 16 April: The file has been removed. It is still available on this blog.

I have uploaded a 2-minute 1600 kHz Perseus file to  this location. It is 907 MB, so be sure you have a reasonably quick internet connection before you try to download. It will only be available for a limited time due to storage restraints.

The recording was made on November 1, 2008 at 1300 UTC. Most stations from the Americas are western US and Canada, from Montana/Idaho/Utah to the Pacific coast, Alaska and at least one Hawaiian. The European stations found are mostly Russian.

I was queried about the background noise. The general noise level is low, but I also have a "ticking" noise which is my local powerhouse, the Loran C. It is 250 kW, 14 km away as the crow flies, and in the same direction as North America. The frequency is 100 kHz. This is an extremely challenging adversary, if you excuse the expression. In order to avoid overload to broadband receivers like the Perseus (and many others), a 100 kHz notch filter is crucially important. Stefan Wikander made one for me many years ago, and I've copied that design (and a
modified design by Dallas Lankford) for all my antenna feedlines. Search elsewhere in this blog for details.

To avoid excessive Loran C noise bleed, maintaining the rest of the equipment at a 100%
quality level is equally important and takes a lot of time. For me, one poor connection is detrimental.

Setup: The beverage was fed with RG-316 coax to a self-made antenna connector made up of a cannibalised SE-6 switch box (an accessory to the Sherwood SE-3 sync detector). From the antenna selector to a Wellbrook 10 dB preamp w/ bypass, then to a 1:8 Wellbrook splitter/13 dB preamp w/ bypass. Bypass is important. During nights, often no preamps are engaged. During the lightest part of the day, often both preamps are engaged. I don't remember the preamp settings at the time of the recording, but at least one preamp was on to compensate for the loss
in the 1:8 splitter. Probably both.

Some of the Russian stations are quite strong, such as 657, 1134, 1449 and 1521. These are "local" stations, located on the Kola peninsula. Other Russian stations of potential interest are the VOA station on 810 (Kurkino) and one on 846 ID'ing as Radio Moscow Oblast. Location Electrostal?