Sunday, October 23, 2011

Another DX Test - Another Blackout

Although propagation hasn't been outstanding, it has at least opened up for trans-polar signals the past week or so. But it's a rather weird coincidence that most times a DX test is announced, the conditions will go rock bottom.

As they are today, during WGBW-1590's test going until 0700 UTC. Not even a faint trace of a signal. Even the near-24/7 companion KBRW-680 is gone.

The reason is that the proton flux began to rise yesterday, almost to the level of a proton event. Increased proton flux levels will always kill propagation in polar regions.

I hope they will set up a new test sometime, though.

Saturday, October 22, 2011

KONG21 Antennas And Receivers - Some Reflections

I had four SDRs with me to this DX-pedition: The Winradio G31DDC and G33DDC, the RF Space NetSDR and the Perseus. The Perseus has been with me for a few years but the G33 and the NetSDR were new this year. Actually, the G31 was never put to use. I've been just as busy testing and discussing software as DX-ing this time, actually a rewarding experience since I've had some fruitful discussions with  Simon at SDR-Radio.com and Ian at Radixon (Winradio).

Most of the time, if not all, the Perseus has been connected to the 50-degrees beverage, the G33 to the 340 and the NetSDR to the 310. We also had the QDFA up, and another receiver (G31) available.  But they were never put into use. The QDFA wasn't working properly, and the reason wasn't discovered until Saturday afternoon when the rest of the crew had left. A rodent had short-circuited the feedline.The cut was so small I didn't discover it the first time I checked the feedlines. And although conditions weren't outstanding, we never found time to investigate deeper. Until Saturday afternoon.

The QDFA feedline is now fixed, and will be available for remote DX for the four of us when conditions improve. Even today the effects were stunning compared to the 340 beverage (the QDFA is 350 but a wider front lobe) as semi-local Murmansk 657 was 27 dB weaker on the QDFA, while KBRW on 680 was a dB stronger. Similar effects on 1035 where China and Estonia were fighting for glory on the 50 degree, but the QDFA nulled Estonia almost into noise level.

Nighttime conditions were poor throughout the week so we probably wouldn't have heard a lot more anyway, but it's a bit embarrasing to know that we could have lost a lot of good DX because we failed to spot the faulty feedline.

The G33 has been a real workhorse. It's been 100 % stable, and by and large is an excellent receiver. The NetSDR was tested with both Spectravue and SDR-Radio software, the latter with beta versions. Most of the time operation has been smooth and trouble-free although I had trouble with scheduled recordings a while until a fix was made. A write-up on the NetSDR will be released soon.

Unfortunately, my impression is that the Loran C noise has been worse than ever this fall. Some frequencies are almost impossible to listen to. I hear rumours that the Loran C network will be closed in 2015. We'll see.

Finally a picture of the signature dish here in KONGsfjord, namely the King Crab, in Norwegian KONGekrabbe.  The raw, frozen crab was supplied by Berle Fisk AS and Svein Olaf, and was immaculate. The best restaurants in Oslo aren't even close in matching this quality.

Friday, October 21, 2011

Date And Time Stamp On Winradio G31DDC Playback

Are you among those who hoped that Winradio would equip their G31DDC recordings with date and time stamp? So that you could see the time and date on playback, and not having to guess? So that you could move the file to another hard drive or folder without the created time/date being lost? Well, you know that it never happened, and probably never will happen

The solution is surprisingly simple, although it involves a bit manual labour on the keyboard. The key is the G33 software, which can play back G31 recordings.

The G31 has a frequency stamp (Insert freq) which keeps track of the center frequency. The G33 on the other hand has a date, time and frequency stamp (Insert FDT). The frequency bit is identical on the two. What separates them is the date and time bit. So I thought: What if I inserted the date and time info manually to the G31 file, and played it back with G33 software?

Worth a try wasn't it, so I changed the original file which was Rec-0.657MHz.ddc to Rec-0.657MHz_11-10-08_16'23'00.ddc, opened the G33 software and pressed Play. Smooth as silk. I then tested two linked G31 files and changed the file names according to G33 syntax and played them back. Smooth as silk. (The time displayed in the playback bar is UTC, i.e. local time - 2 hours)

Of course: If you have a lot of files, it will take some time to change the file names. And linked files may take longer than separate files, because linked G31 files use the -$001.ddc syntax while G33 files use -^001.ddc syntax. But the linked files will have the same date and time stamp as the first file. If you have made a habit of writing down when a file starts, the time should be fairly accurate. And if you haven't made that habit, now is the time.

A word of caution: Always, always, always back up the files you want to change. G33DDC software is available here.

Kiribati Strongest Skywave Signal

Would you believe that Radio Kiribati 1440 could be the strongest skywave station in Arctic Norway? Well today it is. A nice -90 dBm on the 50 degree beverage is only matched by KICY Nome AK 850. Even some of the Russian groundwave signals are weaker.

Strange conditions. Very strange indeed.

Wednesday, October 19, 2011

The Winradio G33DDC - Impressions

As usual, I cannot resist the temptation to write about the receivers I have. Some initial impressions have been posted earlier in this blog. You will find a more detailed report in the Links section, or here. Comments are appreciated.

Tuesday, October 18, 2011

Tuesday Morning (well, sort of)

Wow. What an uneventful night, followed by an uneventful morning. Now we're in the middle of an uneventful day. Only the usual dominants are heard, and not at very good signal levels either. Which is a bit strange, since there is no indication that the solar weather has turned worse, on the contrary. While we're waiting for things to improve, we check recordings from prior days and some nice stuff is found such as KPAY-1290 and KQEQ-1210.

My own October log is still here.

We are looking forward though, to Arnstein's special dish, Bue's Uruguayan tender loin beef tonight.

Monday, October 17, 2011

Monday Morning

An uneventful night, but a bit better in the daylight hours. 910 produced KCJB first, later on KWDZ. Graveyard channels mostly dominated by Idaho, Oregon and Washington stations. KBLA Santa Monica CA had a wonderful signal on 1580 a while. We had hoped for similar Asia/Pacific conditions as yesterday, when Kiribati-1440 had a great signal prior to s/off 0930 UTC, but the band has been very, very quiet. Only now, around 10:30 do we hear the first Japanese stations. Luckily, the sun is still relatively quiet, and we can only hope for it to stay that way.

I have had the chance to play around with the new Winradio G33DDC with the latest software revision a few days, a "joint venture" between me and Radixon.  We have indeed made some progress, and I hope it will continue.

Sunday, October 16, 2011

A Brief Sunday Report

Things are going as planned, with reasonably good conditions. Even New Zealand was heard today, and so was Kiribati 1440 - which is predicted to be a common guest in the morning before it closes at 0930 UTC. Cool and windy on Friday and partly Saturday; rather warm for the year today with around 10 Celsius, but increasing wind and rain is heading our way now.

Below a couple of pictures of last night's ordeal; a King Crab Carbonara main course and some selected cheeses with port for dessert.

OK conditions towards the western part of the US today, and some very low powered Japanese stations also showed up.

From left: OJ Sagdahl, Arnstein Bue, TJ Bråtveit

Friday, October 14, 2011

Bjarne's Updated KONG21 Log

Some of the less common stations I hear will be listed in this update. Some explanations:

G33: Excalibur Pro G33DDC
G31: Excalibur G31DDC
Perseus: Microtelecom Perseus
NetSDR: RF Space NetSDR

50: 500 meter beverage at 50 degrees
310: 225 meter beverage at 310 degrees
340: 340 meter beverage at 340 degrees
QDFA: Quad Delta Phased Array at 350 degrees

Mostly, one antenna will be connected to one radio; the 310 antenna is connected to the G33 etc.

New Voice From The Pacific - Kiribati 1440

Actually not new - Kiribati's 846 kHz outlet has been heard here on a number of occasions. However, in 2008 the transmitter broke down after havning had serveral technical problems. On March 28 2011 Radio Kiribati started broadcasting on 1440 kHz, replacing 846.

To my surprise, Radio Kiribati was heard with a nice signal from 0929 UTC today, until sign-off at 0950, although with occasional interference from Japan. Nice ID at 0936. The MW band was exceptionally quiet at the time with very few stations audible.

KONG21 DX-Pedition Starting Today

Another year, another KONG DX-pedition. This year, it starts today and ends on Saturday, October 21. TJ Bråtveit and Arnstein Bue are scheduled to arrive at 1100 UTC while OJ Sagdahl takes a later plane and arrives at 2100. Conditions are ok, not exceptional but not discouraging either. A number of Perseus, Winradio G31s and G33, an RF Space NetSDR and a QS1R will help us get through the coming week. And if they can't help, we are well supplied with 10 kgs of King Crab, reindeer tender loins, fresh salmon loins, ox tender loin from Uruguay and carefully selected red and white wine, champagne and various other beverages.

Speaking of beverages, the antenna farm is the usual: a QDFA and three beverages directed at 310, 340 and 50 degrees.

Just below freezing today so the roads are icy and a bit slippery, but we're expecting close to 10 Celsius on Sunday.

Thursday, October 13, 2011

Winradio G31DDC LF Sensitivity

The following may be of interest for a very small percentage of my readers. However, there was a discussion on a Winradio refelctor on this issue, a debate coloured by many assumptions and few facts. I never did bother to measure the G31's LF sensitivity, but today I did. With the usual parameters (AM mode, 6 kHz bandwidth, a 400 Hz tone and 30 % modulation), I measured:

500 kHz: -104 dBm
400 kHz: -104 dBm
300 kHz: -102 dBm
200 kHz: -102 dBm
150 kHz: -102 dBm
100 kHz: -100 dBm

I'd say that's fairly good. The G31 doesn't have a built-in preamp, like many conventional receivers have, so a direct comparison may not be fair unless an external 10 dB Norton push-pull preamp is placed in front of it. If so, the G31 will play games with anyone.

Sunday, October 09, 2011

Finally Some Transpolar DX

The sun hasn't been kind to us lately. The contrast between now and a year ago is startling. But there's not much we can do about it except search for the few openings there are. Like this morning at 0400 when east coast stations were well audible.

A check of the overnight scheduled G33DDC recordings was not encouraging, but at 0400 things opened up a bit and I noted WFLR Dundee NY 1570 which is a rare guest, along with WBAE Portland ME 1490 (actually a new station for me at this location) and a few others. Some of the Ontario stations like 1580, 1380 and especially CJMR-1320 had excellent signals, but then they are not the ones we want to hear.

October 9 was the first day of the season with a thin white layer on the ground. It's expected to stay cool but above freezing the week to come.

Sunday, October 02, 2011

G33DDC DDC Recordings

OK. About time to explore the functionality I missed in the G31. First a few words on on-the-fly recordings: This function works well but one should take care in activating the "Insert FDT" button as well as the "Split" button - the former to ensure you have frequency, date and time associated with the file, and the latter to avoid monstrous DDC files.(For the record: No, a 2 GB DDC file is not monstrous).

You can activate a buffer to make the recording start up to 3 seconds earlier. While the idea is very good, the limited buffer size makes it less useable than if you could select up to say 10 seconds.

Scheduled DDC recordings is an important factor because it adds flexibility to the DX hobby. A well designed scheduler is the key factor. This is how the Scheduler window looks like:
The jpg should be self-explanatory. What I'd like to add is that "Start time" is always the PC's local time, not UTC. Take care in selecting the correct DDC bandwidth (default is 24 kHz), and find a suitable argument for the Stop recording function.

The file is automatically frequency, date and time stamped. There is support for multiple tasks, especially suitable if you want to record different frequency spectra and other bandwidths

A scheduled recording is played back as shown below. If recording has been set up to split, the Scheduler will automatically start a new file linked to the previous one, and on playback will do so seamlessly.
Note that although the timer and the file name refers to the PC's local time, the date and time line to the left of the playback bar will display UTC, if UTC has been selected to be the primary time display (Options -> Time).

Sadly, the playback bar is still difficult to manouver. Inspiration for improvement could come from as different sources as Perseus, SpectraVue, SDR-Radio and Winrad. They are all much better.

G33DDC files will be huge. This is because unlike just about every other SDR on the market, the G33 makes 32-bit recordings up to 3 MHz bandwidth. The widest spectra (3.2 and 4 MHz) are recorded in 16 bit due to limited USB 2.0 transfer rate. Thus, 2 minutes, 40 seconds at 1.25 MHz will take up 2.2 GB. In comparison, the Perseus is 24-bit for all bandwidths except 1600 kHz (16-bit), while the netSDR is 24-bit up to 1 MHz and 16-bit on 2 MHz.

A 32-bit file contains more information than a 16- or 24-bit file. The question is of course if the added information can be used to hear more stations, like a considerably higher dynamic range, sharper selectivity etc. It would be nice if that was the case. Until someone comes up with proof I find it hard to believe that a G33 file recorded in 32 bits would let me hear more DX stations than a Perseus file recorded in 24 bits.

Saturday, October 01, 2011

A Brief G31DDC vs G33DDC Comparison

After having tested the two Excalibur SDRs, the G31 and the G33 "Pro", it's time to draw some preliminary conclusions.

Those who have followed my blog, and postings to the Winradio G3 reflector, will know that I have found the G31DDC to be an excellent receiver, but with a serious drawback: The missing ability to date- and time stamp DDC recordings. In fact, the files do not contain any date and time information. This is of course a software issue and not a hardware one. Winradio decided to wait a year, until they released the G33DDC, to implement this crucial function.

So, if you want to use a function found in all other important SDRs (including the low-cost SDR-IQ), you need to "upgrade" to a "pro" version of the Excalibur. There was a heated debate on the topic on the Winradio G3 reflector a while ago, which ended with us "whiners" basically being told (in a more diplomatic language) to shut up or get lost. Mind you, this was not posted by a Winradio representative.

Anyway, I couldn't resist the temptation to buy the G33 (thus proving that Winradio's business philosophy works, unfortunately). A record-breaking shipping time of only seven days from Grove Enterprises allowed me to test the SDR this weekend.

As you can see from the display screendump above, the G31 and G33 GUI is basically the same. The only difference is the added functionality in the G33 hardware and software. The G33 software can play back G31 DDC recordings. More surprisingly, the G31 software can play back G33 recordings, although without date and time, and limited to the 2 MHz DDC bandwidth of the G31 software But why would you use G31 software for G33 playback.

The G33DDC is as sensitive as the G31, a uniform -106 dBm or 1 uV was measured which is very good for SDRs. The G33 even has a 10 dB preamp, which should allow for even better sensitivity. At first I found no audible or measureable effect and thought it could be a software bug in version 1.61. However after testing with the G33's built-in measuring tools, I found the preamp to increase sensitivity to around -110 dBm, making it with a good margin the most sensitive amateur-grade SDR in the market. For those (few?) who need it.

So then: How do the two Winradio SDRs compare? To find out (and to save myself from writing and my readers from reading a long and boring blogpost), I set up a matrix with functions I have found interesting for the DX-er in general and the MW DX-er in particular. The grading is subjective. I may have overlooked something. Please comment if you find other important differences between the G31 and the G33.

Value for money? Well, sort of. For me, the G31 was a disappointment in two important parameters, but others have different views. The G33DDC is USD 1,650 from Grove Enterprises. A UK quote is GBP 1,333 ex. VAT (comparably a lot more).  A Perseus bought in the UK costs only GBP 583 ex. VAT (exactly the same as the G31DDC), leading me to the conclusion that in Europe, the Perseus is much better value for money. In the US the situation is somewhat different with the Perseus selling for USD 1,000.

The RF Space NetSDR is USD 1,450, but the netSDR is hardware expandable and basically a different receiver altogether (and under a continous software development).

In conclusion, the G33DDC is what the G31DDC should have been. But revenue is everything.

Winradio G33DDC Excalibur Pro

Actually I had decided not to buy it. Then I discussed the matter with my good friend Takashi Kuroda who had bought one to replace his G31, and I decided to place an order at Grove Enterprises. The delivery time was a stunning seven (!) days.

I've just unpacked it and connected it to an aging dell laptop running a 2 GHz dual-core Intel T6400 CPU. Incidentally, the stated MSR is "2 GHz Dual Core"... I was pleasantly surprised to see that even at a 4 MHz sampling (and recording) rate, CPU load would not exceed 30 % (filter length set at 5000).

No antenna connected yet. I will go to Kongsfjord this afternoon. More as it happens.