Saturday, November 21, 2026

KONG-HQ Maintenance Day. And a Jaguar That Really Bites!

 Three weeks since my last visit. November's been strange so far. Mostly mild. A few days with down to -10°C. Lots of snow on Monday and Tuesday. On Wednesday we had heavy rain with strong gale winds and 3-4°C. Massive rainfall for an Arctic location. Mayhem on the roads. On Thursday everything froze again. Icy roads on Friday, but with studded tyres and 4WD I felt reasonably safe driving from Vadsø to the KONG-HQ although the average speed was quite a bit lower than usual. A couple of hours after arrival this was the view from my front door.

Moonlit afternoon

Maintenance. 

I have a micro form factor Dell PC that has troubled me since day one. I got a borrowed replacement Intel NUC to handle the Perseus off the 340 beverage, but I had hopes that the Dell would cooperate at some point. I did full re-install of Windows and had it running for two weeks with flawless Perseus + Jaguar operation at home. Nothing else is installed, so I hope that this time, it will do its job. So far, so good. But I have a "new", used Intel NUC in the mail. Just in case.

Other maintenance chores were replacing hard drives. I prefer listening to recordings at home instead of from a remote connection, and so do my friends Ole and OJS.

Killing the dominants - the Jaguar way

I've been using Jaguar to run my Perseus SDRs for - could it be 15 years? Being a MW DX-er, I feel that there is no other software that can compete. The developer Ilpo is a devoted MW DX-er just like we are. Maybe that is the reason it works so well.

Jaguar has been in constant development over the years. Mostly "evolutionary", small steps at a time. Not so with Version 13. With the help of AI, Ilpo and his team has developed some truly amazing tools. Imagine cancelling out the super-strong signal above or below you, like 909 when you're trying to find out what's going on on 910. And then nulling the dominant you don't want to hear on 910 for that elusive catch you never thought you'd be able to hear. That's what's happening now! Of course, it won't always work out that way. But often it does. I have cancelled co-channel  signals from 7000 km away, such as the super-dominant CJCA Edmonton AB 930 to reveal a clear ID from KHJ Los Angeles CA - almost never heard before by me.

The tools are named ACE for Adjacent Channel Eliminator, and CCR (with a nudge to one of our favourite bands when we were teenagers) for Co-Channel Remover

Too good to be true? Listen to this looped recording. The first 15 seconds is CJCA together with a male voice buried deep in the RF mud. The last 15 seconds I had engaged CCR, and CJCA is almost completely subdued and there is a weak but clear KHJ ID from that same male voice.

Airspy's SDR# has offered the Co-Channel Canceller (CCC) a few years. It never impressed me since skywave signals were poorly nullled. In addition the process is a lot more tedious. It's worth mentioning that WavViewDX is also offering the Co-Channel Eliminator (CCE), so things have really developed lately.

Three weeks ago with Jaguar's CCR and WavViewDX' CCE still in early stages, I did an A-B-C comparison, including CCC from SDR# v. 1921. On this clip, a Jaguar file is played back with Jaguar, SDR# and WavViewDX respectively, while jumping in and out of the function. Jaguar runs the first 34 seconds, then SDR# until 1:04 and WavViewDX until the end at 1:44.The station I try to null is CNR6 1089 kHz from Fuzhou, China (600 kW). The distance is 7290 km. My personal experience is that Jaguar nulls CNR6 very well, leaving "room" for weaker stations. SDR# didn't null the signal anywhere as well. WavViewDX nulled the signal very well too, but left some noise and distortion of the weaker signals.

Mind you this is a one-off test and results may vary.

A couple of repair jobs on the beverages were necessary today. Not the best day for working with bare hands: -1°C and a brisk breeze. While I was out I took a couple of photos as usual. The snow was crusty and nice to walk (and drive!) on.

To the left: Starting  point of the 50-degrees beverage

The bay


Today's sunrise was at 10:17 local. Sunset was one hour, one minute later, at 11:18! The Polar Night is just two days away.  I returned to Vadsø a little after 11 and the drive went well on sand-treated roads. We don't use salt on roads here - contrary to the southern parts of Norway.

Sunday, November 01, 2026

Further to the alt-Loran Noise Problem: Testing the Setup at the KONG-HQ

Another mild and windy weekend! Saturday peaked at 8 Celsius early in the evening, along with windy conditions and the odd rain shower.

Some may remember that we complained about noise at our alt-Loran location. Was the noise generated inside the Zarges box, or did it come from outside? We were quite sure that the noise didn't originate inside the box, but opinions were mixed. So, we decided to test on a different location, namely at the KONG-HQ.

I arrived a bit before local noon and my plan, among other things, was to test the alt-Loran equipment in a similar off-grid setup near the house. I already had a longwire site 30 metres from the house, so the idea was to use its sole ground rod and disconnect all the rest. I then walked out 300 metres of single-thread aluminium fence wire in a north-south direction without far-end termination - similar to our 900-metre beverage at alt-Loran. The fence wire and grounding were connected to the alt-Loran transformer and the Zarges box was placed only centimetres away from the antenna. Similar in every way to the alt-Loran setup.

Testing! The aluminium fence wire is the lower one.

I let it run for continuous IQ recording for from early evening to the next morning. I started off using the ground rod. Later on I replaced the ground rod with the radial wire ground we made at alt-Loran. A while afterwards I used the ground rod to ground the Zarges box. Again: Same procedure as when we tried to mitigate the noise problem at alt-Loran.

The setup was noise-free from beginning to end. I did several measurements of the reference signal on 399 kHz which is fully groundwave, and the noise floor on an unpopulated frequency (1615 kHz) The numbers were identical from evening to morning. Since I had Jaguar running all the time I could check the signal levels at any time. Signal level checks after IQ recordings were done with the Perseus software.

So, what remains may be a "polluted" ground condition at alt-Loran. We do know that a buried 22-kV line crosses the beverage around 800 metres from the starting point. It connects to a 22kV:230V transformer station around 120 metres to the east of the antenna. And then there is a high-level RF output to the west. So, we may have some dirty neighbours.

That said, the advise to add a grounding based on radials, and to also ground the Zarges box, did fix the noise problem at alt-Loran.

By the way,  the new fence wire antenna had excellent directional properties. I was even able to log a new country, as COPE Madrid on 999 kHz was almost totally absent, leaving the frequency to a very clear Radio Malta. Through the rest of the evening and night mostly Arabic and Greek on the 9-er frequencies, and weak signals from North America until the disturbance killed everything.

A couple of photos to end this post, first a very early morning impression, still dark. The second one with the sun at a very low azimuth was taken 90 minutes before local noon. We're just four weeks away from the Polar Night.

Early morning


Soon lunchtime

And the coming week I will make my last attempt to fix my erratic Dell Optiplex 3000.