Saturday, March 24, 2007

Updating PC-HFDL for New System Tables

We've had some confusion over how to copy the new pchfdl.dat file into PC-HFDL's configuration folder to update its frequency list (ARINC System Table #31 / 0x001F hex).

The file linked from this blog is identical to the one available on the PC-HFDL Yahoo! group. That's where I got it. I thank this group for making it available.

Not everyone can access the files sections of Yahoo! groups, since a Yahoo! login is needed, while a mailing list can simply be read in e-mail. Therefore I have been posting copies of these files to the Utility World web site.

Here is the procedure for installing pchfdl.dat as the new frequency file for this excellent program. This is how I have done it for several system table changes now, and it has worked every time:

How to update the PC-HFDL System Table with pchfdl.dat:

Background:

This file updates the internal system table used by Charles Brain's commercial version of PC-HFDL, a program for the Wintel PC which decodes ARINC's High Frequency Data Link signals.

The HFDL system stores its frequencies in a common list called the System Table, which is changed every few months as HF propagation conditions change with the seasons. The system table number, which is transmitted by the ground stations, is incremented by one.

Ground stations transmit the new list until everyone has it, and no aircraft has a table number mismatch. This usually takes only a few days. PC-HFDL cleverly grabs these transmissions to update itself, and stores them in a file called pchfdl.dat. When PC-HFDL detects a new version of the System Table, it automatically updates this file.

An issue exists where, if PC-HFDL is not running or there is no clear reception at the right times, it will not receive the transmissions needed to update this file. This causes the display of frequencies in the squitter screen to change back to numbers rather than kilohertz, making the program much harder to use. Therefore, the user is told how to update this file manually.

Procedure:

1. Get a copy of a pchfdl.dat file that someone else has made with the program since the new frequencies were transmitted. There is usually one on the Yahoo! HFDL group within a few days, which I change on my site at the same time. This file should be treated as a binary, whether or not it is. If it is read on the screen, it will look more or less like total nonsense, but the program knows what to do with it. Download the file to your disk by right-clicking on the link, then choosing "save target as," "save link as," etc. Make sure it downloads to the desktop or some other place where you can find it quickly.

2. IMPORTANT! Close PC-HFDL before going any farther!!! You won't break anything if you don't, but the frequencies won't update.

3. Find the directory that PC-HFDL runs from. This is usually C:\Program Files\PC-HFDL. Change to this as the active directory.

4. In this directory, there should be a subfolder named configs. Change to this as the active directory.

5. In configs, find a file named pchfdl.dat. It should be around 24K. Rename this file something like oldpchfdl.dat, or pchfdlnn.dat, where nn is the number of the former system table.

6. Copy the pchfdl.dat you have just downloaded to this subfolder. It should now contain your old data file, your new pchfdl.dat, and three other files. It's safe to open these others, but don't change them! One of these, systable.txt, is a human-readable version of the System Table that is written out once pchfdl.dat is processed. Another, pchfdl.txt, lists the ground stations. The third, pchfdl.ini, is the binary initialization file that stores the previously used configuration of the program.

7. Go back to the desktop or start menu, and run PC-HFDL. Find a nice loud ground station with the squitters and click the box. The numbers should once again be replaced by frequencies.


Good listening!

Monday, March 19, 2007

New HFDL Frequency Table 31 is current

I got busy for three days and look what happened. ARINC changed its HFDL System Table of frequencies used for its GlobalLink system.

This is done periodically, usually when the season changes. The new System Table is number 31, or 1F in hexadecimal.

The pchfdl.dat file on this column's web site has been updated. Close down PC-HFDL, then find the directory it runs from. Change to the configuration subdirectory, find pcfhdl.dat, and rename it. Copy in the new file, then start the program. The numbers should change to frequencies again.

I have done this procedure on my own computer, and the file works. Thanks to the HFDL Yahoo! group for putting this up.

Meanwhile, here's what's being transmitted in New York's squitters at 2218 UTC. Obviously other times of day are different:

SAN FRANCISCO - CALIFORNIA UTC
8927 KHz 5508 KHz

MOLOKAI - HAWAII UTC
21937 KHz 13276 KHz 11312 KHz

REYKJAVICK - ICELAND UTC
8977 KHz 6712 KHz 5720 KHz

RIVERHEAD - NEW YORK UTC
11387 KHz 6661 KHz

AUCKLAND - NEW ZEALAND UTC
13351 KHz 6535 KHz

HAT YAI - THAILAND UTC
10066 KHz 5655 KHz

SHANNON - IRELAND UTC
8843 KHz 6532 KHz

JOHANNESBURG - SOUTH AFRICA UTC
8834 KHz 4681 KHz

BARROW - ALASKA UTC
11354 KHz 6646 KHz

KRASNOYARSK - RUSSIA UTC
6596 KHz 5622 KHz

AL MUHARRAQ - BAHRAIN UTC
8885 KHz 5544 KHz

AGANA - GUAM UTC
13312 KHz 11306 KHz

CANARIAS - SPAIN UTC
11348 KHz 6529 KHz

Thursday, March 15, 2007

Give Us This Day Our Daily SK01

17436 PSK31 in progress at 1708 UTC. Began like clockwork at 1700 with 10x 11111, missed some of callup due to operator stupidity (mine, not theirs), middle message is NNTWI, 3rd GTWGI. The signal is strong s7-9 but with frequent deep fades. The audio is way overdriven for PSK31 with heavy distortion causing missed characters at times.

UPDATE 1751 UTC: The signal cut abruptly without the usual 10x 22222 at 1730, back up at 1733 in mid-message, still going at 1052, much later than normal.

Wednesday, March 14, 2007

SK01 Makes Both Skeds

17436.0 kHz at 1700, started promptly in PSK31 with the 10x 11111 starting signal, then callup and message in modified M8a format, using the letters. Still in progress at 1720.

UPDATE 1732 UTC: SK01 abruptly changed to V2a at 1732, finished transmission in voice, "Final, final" at 1744 and gone.

Digipan has trouble with very long PSK transmissions, being designed for short amateur exchanges. Will try it with SkySweeper next time.

Cuban PSK 3/14

17478 5-figure group message in progress right now at 1605 UTC, weak into California. Appears to be the modified V2a sked.

Tuesday, March 13, 2007

Cuban PSK Continues

Today, two schedules.

1600 UTC 17478 kHz, modified V2a format. Repeats group 11111 at message start, 22222 at end. Three messages.

1700 UTC 17436 kHz, modified M8a format with the letters!!!! These are essentally meaningless in a Varicode direct printing mode, where they save a few bits per character, as opposed to cut-number Morse, where they save a huge amount of time. Presumably, Cuban spies and/or their software are used to them.

Both times in PSK31 mode, and audible worldwide. The ENIGMA 2000 designator is SK01. Apparently the PSK is here to stay.

Sunday, March 11, 2007

Digital SSTV Saga Ends

After another day of watching the grass grow, I have decided to go back to utility and leave the digital SSTV experimenters to it. I received a couple more nice pictures, with missing segments, from an amateur in Mexico, who was sending dramatic pictures of women in body paint obviously applied as some sort of art show. These were great stuff, and make me want to look around Google Images for more examples of same.

To see real content is, of course, is rather refreshing, considering the one-note samba that is American analog SSTV. You have the guy that wants us to get Jesus, the one who never met an airplane photo he didn't like, and the one with "YL" in his call who sends pictures of, uh, YLs in rather scanty clothing.

Come to think of it, that third guy can keep on with what he's doing. That scanty clothing thing is fine with me.

Must... think... about... radio... .

Digital SSTV, of course, is not a very good name for it. Digital file transfer would be better, since that's what's going on, and the data could really be just about anything. That's the advantage and disadvantage.

The advantage is that, if everything is received, you have the same file that left the transmitting station, not a grungy decode of a scan of its contents. The disadvantage is that Digital Radio Mondiale in its present state is an audio broadcast mode, and better ways exist to transfer files. There is no repeat procedure for missing information transfer units (packets, segments, whatever), because in audio broadcasting there does not need to be. Segments failing the CRC check are simply discarded, and you don't hear that part.

Of course, the receiving station of a file transfer can always ask for, and get, the missed segments. They've sort of automated this process, with bad segments being logged and requested manually by clicking a button for a "BSR" (Bad Segment Resend?) transmission. This is fine, but obviously the retransmission will probably not be the same segments that people listening on-channel missed. For that, you need redundancy, like in FEC mode. In digital SSTV, real redundancy consists of transmitting the whole file more than once, but that means very long transfer times.

There is also the problem that DRM might not be the best mode for amateur equipment, which at least in the US is not allowed to exceed 1500 watts peak envelope output power. DRM's peak to average ratio is awesome, something on the order of 10 dB. In amateur service, this means that your typical 1500 watt PEP linear cranked to the max will be reading about 120-130 average watts at "full legal output."

Somehow I doubt that we will be seeing whole screens of Internet SSTV widgets showing reception worldwide, as we've had in the analog mode for some time now. (A great one is Worldwide SSTV Cams.)

Now, back to utilities...

Friday, March 09, 2007

GULP!

Waterfall, 14232.975, 2331 UTC:




2342:




2343:




MY FIRST PICTURES!!!!!!

Two days of sitting at the receiver have been rewarded. At 2226 UTC today, W9CY sent a file named UP7.jp2. Amazingly, something appeared in my rx window. I darn near spit my coffee all over the keyboard.

Here is NV6H Digital SSTV First Light:



I wonder what it is. Looks a bit like a Chrysler Air Raid Siren. I think there was something on the orderwire about its being in Hawaii. I missed about 20 segments. Obviously the Irfan View partial feature is working. All in all, not much worse than most of the analog SSTV ends up looking in California on the typical wire antennas used for ute DX.

A mere nine minutes later, the miracle repeated itself! Another picture appeared in my rx window! It's called dq2.jp2. I only lost 11 segments this time. Here's the awesome result, as seen on my very own computer:



This time I know what it is. Thanks to W9CY on the orderwire, I know that it is a Dairy Queen in Illinois!

Conditions seem quite good today. There is no telling what further miracles of human communication await.

UPDATE 2313UTC:

OH MY GOD! There is an AMERICAN EAGLE in my WATERFALL!

The possibilities for conceptual art are endless...........

Digital SSTV Saga Continues

14233 remains active whenever the band is awake to propagate it.

The DRM has drifted slowly back on-channel, last one being measured at 14232.995, accounting for known receiver errors but not unknown computer clock ones.

One file received toward the end of the UTC day, when signal strengths tend to peak on 20, was reported as received perfectly - no bad segments, no CRC errors, full file integrity. It was a picture, according to the voice orderwire. Nothing happened. Irfan View did not start, DIGTRX did not display anything, and no file was saved anywhere. It is presumably in bit heaven.

Wonder what it was.

Seeking verification, I downloaded WinDRM, which worked, but no joy there either. Another try of a newer version of EasyPAL Lite crashed even harder than the last one, not even making it through its first initialization, while giving streams of errors about writing to nonexistent memory addresses before hanging altogether and being killed from Task Manager. Obviously some underlying runtime module is wrong, missing, or corrupt. Probably not EasyPAL Lite's problem, but radio is about communication, not troubleshooting your operating system.

I note from the Yahoo! group that this is not exactly the turnkey program of all time anyway.

While the quest continues, good old grungy Scottie 2 and Scottie DX look better all the time.

UPDATE 2125 UTC:

Discovered the local loop-back test mode. I am now quite good at sending jp2 files to myself. Software appears fine, and the Irfan View preview works.

Still no decodes here. Most stations are in the Midwest. As most people who live here know, short wave listening in California is often like watching grass grow.

Thursday, March 08, 2007

Cuban PSK Second Day

For the second day, March 7, the voice schedule on 17436 kHz at 1700 was transmitted in PSK31 instead, using the slightly altered format. The message was a repeat of March 6, this time with a little Radio Havana Cuba program leaking in.

20 Meter Digital SSTV Continues

This activity is pretty much continuous during Western Hemisphere daylight hours, up from 14233 kHz, going as high as maybe 14233.35 at times. They continue to transmit the callsigns in the waterfall. Despite a continuous voice chatter (some kind of orderwire, plus a guy talking about his '53 De Soto) no signals have been anywhere near strong enough here for a decode.

UPDATE 1930 UTC: Currently the USB voice orderwire is on 14320.0, but the correct tuning for the digital signal is 14233.24 +/- around 200 Hz. Mention was made (following the '53 De Soto discussion) of another program called EasyPAL. This apparently implements another widely used mode called HamDRM (Presumably Digital Radio Mondiale). Unfortunately, it crashes and burns if you are using hyperthreading on dual-core Wintel machines with XP Service Pack 2. You have to turn this off in the BIOS, and I didn't feel like doing that, since I use the feature a lot with other processor-eating apps to still have a usable computer for things like radio, while they're thinking away.

UPDATE 2036 UTC: Still nothing even close to decodeable signal levels. Apparently the mode needs a good steady S8 or 9 for perfect reception, and not too much lower for any decode at all. Unless someone local transmits, that is not going to happen on 20 meters at this time of day at this point in the cycle. This explains why I seem to be the last hard core radio hacker ever to hear of digital SSTV. Like so much of HF amateur radio, it is just not that rewarding without the use of large rotary antennas on tall towers, of a sort that will get you sued in most decent US neighborhoods.

This is in contrast to analog SSTV, which will decode even when below the noise, though you don't get much. At least you know your software is trying.

So it goes in the exciting world of digital radio. I knew there was something to be said for R-390s and straight CW.

UPDATE 2205 UTC: Downloaded a new DRM DLL for DIGTRX. For some inexplicable reason, the entire folder is set read-only, perhaps because the program is running. Resetting it only causes it to be set back. An effective work-around was to rename the old DLL while out of DRM mode, then copy in the new one. Now DRM mode of DIGTRX does not crash with hyperthreading enabled. (Of course I could always have tried exiting the program...)

DRM has the advantage that you can indeed watch your software try to achieve a decode, complete with a nice phase constellation to show how things are going. At least the listener knows something's happening.

Still no decodes, however. Dial reading when on-channel in this mode is 14232.95.

More Ham Funny Noises (This time 20M and legal)

I was trying to track down this weird noise on 20 meters. It sounded a bit like STANAG 4285, a bit like MT63, and at times a whole lot like the backscatter radar. It seemed to be centered around 14233.3 kHz. SkySweeper wouldn't do anything with it, and MultiPSK wouldn't do anything with it.

So I'm watching it on the waterfall, trying to figure out the mode - pilot tones followed by multitone 8PSK - and just then a stronger transmission started up. Here's what appeared on the waterfall:


That's right. There was writing in the waterfall, transmitted by just the right modulation of the signal. (This, in fact, is what sounds like the backscatter radar.)

See something new every day.

Other bursts had different amateur callsigns, signal reports, roger/no copy messages, and finally a smeary little identifier of a program named DIGTRX 3.11.

Google is a wonderful thing. I grabbed DIGTRX and ran it, and found out it's a digital file transfer mode popular with amateurs in South America to send SSTV pictures without the analog noise and grunge. That certainly explains why they were on 14233. The letters are called a "waterfall message," and there are also "waterfall pictures." Yes, real monochrome photos, in the waterfall. Little faxes, sort of. This is all just too cool.

Can't wait to get some amateur signals loud enough to decode the content. Terrestrial radio survives because people keep re-inventing it.

Wednesday, March 07, 2007

More 7000 kHz Strangeness

This comes from Brazil, describing unlicensed CB-style 40-meter activity in that country:
Regarding your last post ('7000 kHz Is Getting Weird') I would like to
say, perhaps, brazilian RF users are helping on this weirdness.

Unfortunately it is increasingly common to see truck drivers installing
'27MHz->7MHz' transverters on their CB radios. This would explain the
use of words 'cambio' and the whistle (a 'roger beep' signalling the end
of QSO). It is very common to see unlicensed voice traffic between
6900-7000KHz. Here, they call this space as 'faixinha' ('little band').

This page (
http://www.zamin.com.br/principal.php?id_menu=produtos&categoria=10 )
has CB transverters for 6, 40 and 80 meters (25-50W) costing
US$100-US$200. Although the company's objective is to help licensed hams
operator to start on 6, 40 and 80 meter bands with short budget,
unlicensed CB owners are using this devices to get better coverage, but,
sadly, whithout any licensing and control.

Best regards,

ZZ3HAG Huelbe Garcia
member of DX Club of Brazil

Tuesday, March 06, 2007

7000 kHz Is Getting Weird

For the last several days, there has been a real zoo on 7000 kHz, right at the lower end of the 40 meter band. Some people have heard a STANAG 4285 signal that decodes to the usual French Navy testing loops complete with Voyez le brick and identification as FUV (French Navy, Djibouti). Parameters are 300/L/5N2/ITA2.

It may or may not be coincidental that Djibouti and the African "horn" in general is a world trouble spot you don't hear much about, with several important oil fields, tanker routes, and natural gas pipelines at stake. Military activity is greatly increasing there as of late.

Other signals being heard are hours of conversation in Spanish, using procedures such as "Cambio" for "Over." There is a little of the whistling which seems to be a test tone down there, and a little of the hooty sounding multiband inversion scrambling used by the Mexican Army and/or ???.

Finally there are at least two ALE signals, one from the Italian Carabinieri phone patch net which passes an AMD string of "DIAL4," another with short alphanumeric IDs which are still unknown.

This is not counting a weird artifact that has been transmitting on a wide range of 40 meter frequencies for years, producing a short beep every minute or so. It is a weird dual-frequency beep resembling a DTMF tone but not one. I suspect it's local to Southern California.

It is not known whether any of this is legal on 40 meters. I personally suspect it is not, but it's a grey area. However, hams who tune up or send CW on 7000.000 in an effort to drive all this stuff off must have a lot of faith in their VFO calibration.

Cuban PSK Broadcasts Again

The new Cuban PSK31 numbers broadcast (SK01) is back. It substituted for the normal voice sked on the new 1700 UTC frequency of 17436 kHz. The message format was not the standard V2a, but it seemed close, and much longer than the previous transmissions.

It ended with 10 repetitions of the group "22222," then the equals sign (=, meaning break) and then "tt," possibly an end-of-message signal ("todo, todo"?) , perhaps not.

SK01 will bear watching. It is straight PSK31, same type amateurs use. If you don't have a PSK31 program, Digipan is good, and it's free. MultiPSK is also good, and free. MixW is a nice comprehensive package including PSK31, but it is low priced shareware, not free.