From VMSSERV@arecibo.aero.org Mon Mar 29 14:01:51 1993 Received: from arecibo.aero.org by cs.tut.fi with SMTP id AA16774 (5.65c/IDA-1.4.4 for ); Mon, 29 Mar 1993 14:01:44 +0300 Message-Id: <199303291101.AA16774@cs.tut.fi> Date: 29 Mar 93 02:44:00 PST From: "VMSSERV Mail Server" Subject: UWO9301.TXT To: "rko" Status: OR +++++++++++++++++++++++++++++++++++++++++++++++++++++++ + SPEEDX's + + Utility World Online Edition + +=====================================================+ + Volume 1, Number 4 January 1993 + +=====================================================+ + Send all comments and contributions to: + + Chuck Yarbrough, Editor + + Johnsons Mobile Court, Lot 29 + + Route 6 + + Statesboro, Georgia 30458 + + United States of America + + Internet: SOCOMCY@GSVMS2.CC.GASOU.EDU + + Compuserve: 70544,2566 + +=====================================================+ + Copyright 1993, SPEEDX. Permission granted to use + + this information as long as SPEEDX and Utility World+ + is cited as the original source. + +++++++++++++++++++++++++++++++++++++++++++++++++++++++ Happy New Year! We're back at it again with a full year of ute news and logs from you guys (and gals!). This month we are going to look under the hood of the Hoka Code3 multi-mode decoder. But first, as you know there are some new developments on the international front, particularly in East Africa. We start there. Somali Ute Freqs The United Nations last month started a massive airlift of troops and material to the East African nation of Somalia. We have a special report by Larry Van Horn, Ute World columnist for Monitoring Times SW Radio Magazine. Take it away, Larry. The following was posted by Larry to the GEnie electronic information service. Where do I look for DX? First off, go to the GHFS frequencies. Watch for use of discrete frequencies (See my September USAF article). For those of you not MT subscribers (shame on you) here on the primes again (please don't ask me to rekey the discretes): 4725 6738 8967 8993 11176 11243 13201 15015 17975 Now as far as AFI-3 and INO-1 here are those to watch out for: AFI-3 Family Addis Ababa 3467 5658 6574 8870 11300 13288 17961 Aden 3467 5658 11300 13288 Benghazi 3467 5658 11300 13336 Bombay 11300 13288 Bujum Bura 8888 11300 13288 13294 13306 Cairo 3467 5658 6574 11300 13288 Dar Es Salaam 5658 8870 11300 13288 Djibouti 5658 11300 Hargeisa 5658 11300 Jeddah 5658 11300 Khartoum 3467 5658 11300 13288 13336 Mogadishu 5658 11300 Nairobi 11300 Riyan 5505 6624 8847 8959 11300 13336 Sanaa 8918 11300 Seychelles 3467 5658 11300 13288 17961 Tripoli 5658 11300 Watch 11300 in late afternoon and early evenings, my favorite aero frequency this time of year. Right now lots of activity, good place to watchfor Mogadishu. You might catch it also around 1200 UTC on long path with a little luck. Evening should hold until around 0300 to 0400 UTC. INO-1 These are a little rougher to hear, but traffic coming from West Coast via Diego Garcia will be found here. Long path evenings best bet. Antananarivo 5634 8879 13306 Beira 8879 Bombay 5634 8879 10018 Cocos 3476 5634 8879 13306 17961 Colombo 8879 13306 Dar Es Salaam 8879 13306 Johannesburg 3476 5634 8879 13306 Lusaka 5634 8879 13306 Mahe 5634 8879 13306 Mahajanga 8879 Male 5634 8879 13306 17961 Mauritius 3476 5634 8879 13306 Nairobi 6559 13306 Perth 3476 5634 8879 13306 17961 St. Denis/ Gillot 3476 5634 8879 Seychelles 3476 5634 8879 13306 17961 Toamasina 8879 Be sure to watch NAT families especially NAT-A and E for mil activity during non-Africa propagation periods. These should be ripe anytime with mil traffic headed that away. Watch for Coronet flights, Coronet + color calls (I would be especially interested in hearing about those), and also watch some of the Desert Shield/Storm AMC discretes that popped up. I don't have them handy right now, but I covered some of them in the features and columns during the period. Thanks Larry for that most useful and timely information. Larry is retiring from the Navy after 23 years this spring and will be 'moving in with Bob' at Grove Enterprises mid-summer. He is currently working on a new edition of Grove's Shortwave Directory, something to look forward to! Larry invites all ute DXers (SPEEDXers included!) to give him feedback and info on what they/you want in a guide. You can contact Larry at Grove Enterprises, P.O. Box 98, Brasstown, NC 28902-0098, USA. Special Discussion: HOKA Code3 I decided to devote a little space this month to a discussion of the Hoka Code3 multi-mode decoder. I will provide you with my thoughts on the machine, but I felt that more than one opinion might be informative. To that end I solicited a review from several Code3 users--only one took me up on the offer. That brave soul is Ed Deasy of Charlottesville, Virginia. Ed has used the Code3 for several years and, from his discussions seems to be quite expert with the unit. Take it away, Ed. Code3: A Detailed Review Ed Deasy, Charlottesvile, VA I first heard of the Code3 RTTY decoder from an advertisement in Monitoring Times more than two years ago. At the time I had been curious about the RTTY modes mentioned in the Klingenfuss Guide to Utility Stations, some of which were not available on U.S. made RTTY decoders. The Code3 seemed to cover all of the Klingenfuss modes, and quite a few others. With some trepidation I mailed my mega-dollars overseas to someone I thought understood my order on the phone. What I got was much more than what I expected. In fact, the function of many parts of the software were beyond my comprehension. Unfortunately, I also got an extremely complete German manual, and a halting, sketchy English translation. I couldn't afford much transatlantic "quasi- German-English" tech support, so I had to tough it out. Fortunately, Hoka Electronics, who make the Code3, had their manual re-written by Hoka's U.K. office and now is quite understandable. Hardware description: Code3, other than software, consists of a small box with two cables and one jack, and no adjustments. One cable is for AC power, one is for a 9-pin RS-232C serial port, and a RCA jack for audio input. Unlike many serial port adventures, I just connected everything together, installed the software, and it worked. The Code3 works primarily by analog to digital conversion. The audio passband is "sliced" into digital values, and then processed to determine timing and placement of tones. This is the basis of most of the Code3's "analysis" modes, they are all mathematically derived from the digital information. Software considerations: The software is copy protected, with a numbered install routine. The word that come to mind is "onerous," but I can understand the company's need to protect their product. Other decoders have their key code in ROM; you can't use the program without the hardware. The Code3 hardware is quite simple, so the company has found it necessary to make the software "locked." The trouble with locked software is that it can be inconvenient. Also, if you should have two hard disk failures you'll need to return the software disk for renewal. Features of the Code3 A unique feature of the Code3 is derived from the Analog to Digital signal conversion. The digital information is stored in a buffer during receive. This information can be stored as a disk file (about 500k). Hoka U.K. says this is rather like DAT tape. The entire audio passband is stored in digital form. When you play back this file through the decoder it is just like being tuned to the original station again. Thus, you can store an unidentified mode for later identification. Furthermore, if you have a fast PC, like a 386, the playback is much faster than real-time. The software kicks up the speed in the decode and analysis modules to synch with the file dump. The speed of the playback will also depend on the module--some are more numerically intensive than others. It may only take a minute to listen to 3 minutes of saved digital info. In most cases it's more convenient to use this feature to identify or analyze a signal than to do it in real time. Since there are more than 24 different modes and 11 analysis modules to check the signal out on, the signal may be gone before you've found out what it is. I should also point out there are two analysis modes that can act as decoders for modes not covered above. They are "character analysis" duplex and simplex. These modules search for bit patterns of known RTTY alphabets in the data flow. Many parameters exist within these modules for customizing. In practice, only rarely does this actually work. Modern systems are often quite complex, with interleaved inverted characters, making alphabet analysis a fairly complicated procedure. Nonetheless, I have been able to log several keying systems not covered in the decode modules using this analysis module. Analysis and Tuning: You tune while viewing a spectrum analysis of the passband. You simply tune the peaks corresponding to the tones until they are equidistant from the center frequency. A baud readout, accurate to .00 (!), shows in 10 to 30 seconds, with both synchronous and asynchronous speed equivalents listed. I don't know what the maximum speed reading is, but I've measured rates above 400 baud. This particular baud readout will be effected by noisy signals. The "correlation mod signal" module should be used in this case. Oscilloscope: This is a real oscilloscope, totally synchable, and with a storage feature. This feature costs extra. Correlation Mod Signal: This determines baud-rate to .0000. You can get a readout in this module when the noise level is too high to get a consistent readout in the standard tuning screen. Correlation Bit: This "exclusive Xor" analysis of bit patterns reveals CRC (character repetition) or other repeated patterns in data. Since it is statistical, it will work with very weak signals. Uses a graphical screen. This soon becomes an indispensable aid in identifying modes. Mark-Space analysis: This determines whether the signal is synchronous or not, even in high noise. Uses graphical display. Speed Preset: In high-noise, this modules does a statistical analysis to indicate which standard baud rate you might have. Speed-Bit analysis: This shows bit length graphically, which can be measured in Milliseconds using on-screen pointers. Bit analysis: This is the same as databit mode on M-7000. There is also a 'Save as ASCII' feature, which, oddly enough, costs extra. Performance: There are two issues here; how well the decode module is implemented and how well the unit actually senses tones in the audio passband. The decode implementations on the Code3 are in many cases better than the M-7000. There is no need for auto-synch or manual synch on synchronous modes. It doesn't matter whether the station is idling or sending traffic, the Code3 synchs up immediately. ARQ-E is very poor on the M-7000, does not easily synch on the 4 CRC version of ARQ-E: Code3 does this mode correctly. Baudot also seems more readable at lower signal levels than M-7000. The M-7000's TDM decode is very good on maintaining synch on very weak signals, if they are idling. The complexities of evaluating performance can be seen by taking an example from a recent monitoring session. I was trying to get an I.D. on a very weak 200 Baud TDM two channel station. I had both the Code3 and M-7000 on the signal. The M-7000 was maintaining synch on the idling signal much better than the Code3, which was holding synch for stretches, but then dropping it. When the actual traffic started, the M-7000 lost synch during a fade, so I had to hit the manual synch button to try to regain synch. I lost all the traffic on the M-7000 because it refused to re-synch on the traffic signal. The Code3 also lost synch during the fade, but regained and held synch long enough during the traffic to get an ID. On other modes, such as Sitor-A, the situation would probably been reversed, with the M-7000 catching more of the traffic. When it comes to sensing tones in the audio passband, I would rate the M-7000 higher in certain conditions. In wide shifts combined with high noise levels, lets say, an 850Hz shift signal below 10 MHZ in typical evening conditions, the M-7000 will pull the tones out of a surprisingly high level of noise. The general signal to noise tolerance of the M-7000 is higher, although the differences narrow substantially as shifts get narrower. The Code3 seem particularly sensitive to static and noise bursts in wide-shifted signals on the lower frequencies. It's easy to understand these performance differences when you consider how the "front ends" of the two decoders differ. The Code3 scans the audio passband, so, in the above example, an 850 hz."window" has to be open in between the mark and space tones. This gives it multi-tone capabilities, and perhaps makes other modulation schemes decodable in the future. The M-7000 and other "dedicated filter" type demodulators must be programmed (or auto-tuned) with the shift, after that there are only two very narrow windows of filtration at both the mark and space frequencies. This is very effective in reducing off-frequency noise. For instance, it's possible to copy a wide shift RTTY signal on the M-7000 even if a heterodyne is between the tones. The Code3 must scan across the het., so it chokes. Hoka Electronic is aware of this, hence, the Code30, which has a automatic DSP prefilter, and numerous other features. The DSP filter constructs "windows" in the proper places, rather like the dedicated filter approach. In this case, the "window" will configure tightly around a multi-tone group. However, this product costs much more than the Code3. (approx. $1200) I can also compare the Code3 to the M-1000, which I have used for years. The signal to noise tolerance between the Code3 and M-1000 is very similar. One obvious observation here: the M-1000 is NOT comparable in this respect to the M-7000. In other words, I have found the M-7000 outperforms the M-1000 in most situations. However, the M-1000 seems to have an additional weakness on high baud rate synchronous codes. I've found that 192 baud ARQ-E3, and 200 baud TDM that I can easily synch on the M-7000, will not synch on the M-1000. It can't just be a matter of baud rate, since the packet module on the M-1000 is fine. So, while the Code3's signal to noise tolerance is similar, it does not have the shortcomings of the M-1000 mentioned above. Comparison of important features: The Code3 also uses a slightly different manner of printing text to the screen than the M-7000 and M-1000. If an error is received, a "tilde" [~] character is printed. That means when copy is really broken, paragraph spacing and columns still show reasonably correctly on the Code3. This also makes broken copy much more readable. The Universal equipment, with the exception of Sitor-B, lumps whatever is received together with no spaces. So, broken copy is much harder to make sense of. Of course, there are also some features on the M-7000 which are omitted on the Code3. The Code3 does not have the M-7000's selcal function, which allows certain defined elements in the received traffic to turn on the printer, and turn it off. I don't miss this on a software decoder; it's much better to save to disk than to print out. In fact, the M-1000 is most advanced in this respect, the selcal can control "save to disk." This does add to convenience, keeping disk file sizes small. Rating the Code3 The problem of comparing the Code3 to "decoders" is that is a bit more than just a decoder. The tuning and analysis features have no comparable parallel on any other decoding equipment that I'm aware of. While daunting at first, these modules soon become "eyes" on the spectrum. Many of the modules reveal the actual bit structure of the received signal; they become interactive lessons in RTTY structure and function. This brings new descriptive powers to new keying systems you can't actually decode. If you come across a 125 baud signal, you'll be able to say it has a 8 character repetition rate, that it's idling with pseudo-random characters, etc. For true code "hackers", which I am not, you can go further still. If you can very accurately describe a keying system, using the analysis features of the Code3, apparently Hoka will consider writing a module for it. I have no idea what this might cost, though! They do seem to be actively working to expand the base of systems. Here's a good example of the difference really knowing what you are receiving can make. I often see the statement on Shortwave topics that most RTTY in the spectrum is encrypted 75 baud. The problem with the statement is that it's incorrect. The majority of these 75 baud signals are synchronous RTTY systems: Baudot is an asynchronous RTTY system with start and stop bits. If you have a decoder which indicates whether the code you're receiving is synchronous or asynchronous, it soon becomes clear that only a very small percentage of those 75 baud signals are, in fact, Baudot. The rest come out garbled on your decoder because 'they are not Baudot', not because they are encrypted. Even if they were sending in the clear it would print garbled on a Baudot decoder. On the other hand some folks just want to print out press Baudot and faxes and couldn't care less about statements on code structure, etc. More power to them, they're still sane. The Code3 is more suited to "enthusiasts" who can go on endlessly about such things as I did in the very last paragraph. There are many economical options for beginners in RTTY these days, including the nearly shareware "Hamcom". I hope this information will be of help to prospective RTTY decoder buyers. As you can see, a comparison of equipment and software of this complexity is complex in itself. It can help, where overwhelmed with features as in this case, to carefully consider what you really need in a decoder before making a decision. Thanks Ed for that fine review. After using the Code3 for about 6 months myself, I have to agree with most of what Ed said in his article. I would like to add a few comments, though. I think the last paragraph in Ed's review is worth noting. If you've never decoded a RTTY signal before, you will be overwhelmed by the flexibility and sheer decoding power given to you by this unit. The learning curve is fairly long when learning how to use the Code3. I am (he said humbly!) fairly proficient on all the M-7000 modes, yet I have only managed to decode fewer than half of the modes contained in the Code3! I HAVE learned much more about RTTY and data structures than I ever would have by using the Universal machines. While the Code3 is a great analysis tool, I do have some questions about the implementation of one mode--the Facsimile decoder. While I have been able to decode weather faxes, the tuning is difficult for someone used to the Universal implementation. Most of the decoding is automatic and the resolution doesn't match the M-1000. I would also like to see Hoka beef up their documentation so that non- Computer Scientists could understand the fundamentals of data structures and, more importantly, how to decode them. This point is the most frequently mentioned point of Hoka users, "its just not very user-friendly". The graphic interface is just as friendly as any other decoder, but the features/modes documentation leaves many gaps for the user to 'discover' on his/her own. Perhaps Hoka NAm could expand the manual? Ed is right--the Hoka U.K. manual is MUCH more understandable than the Dutch/German version. From an operational standpoint, the Code3 is easy to install and easy to navigate. The windows-like interface is simple, but it does not support monochrome VGA displays unless you run it under Windows 3.1. If you are still a DOS user, you must have a color monitor. The Code3 will not install on a hard disk with modern disk compression software. (ie. Stacker, DR.DOS 6.0, Xtradrive). Since I run Xtradrive, I had to partition an uncom-pressed segment of my disk (2-3 Megs is enough for Code3). It works fine now. One problem, which I think Hoka NAm has solved, is the power cord on the A-D Converter box. Mine came with a European 2-pin plug which, even with an adapter, didn't stay plugged in. I eventually replaced that plug with a 'clip on' US-style power plug. Now it works fine. Another minor 'gripe' I have with the Code3 on the operational level is that I must give up my COM2 (serial port) to keep it plugged in. I run Windows all of the time and must have a mouse on COM1. I also have an internal modem installed, so every time I want to use the modem I must physically unplug the Code3. I would like to see an internal PC card version of the A-D converter. Despite the above comments, I really like the 'feel' of the program. This decoder has rejuvenated my interest in RTTY decoding in the months since I first got it. As Ed said, it allows you to make the jump from simple 'reactive decoding' to actually understanding what is going on within the signal you are receiving. Indeed, as Ed said you will probably become a 'radio hacker' if you get this unit. If you have grown a bit bored with the same old modes on RTTY, or tired of all those '75 baud blackholes', then you are probably ready for a Code3. Code3 costs $495 US and can be obtained from: J&J Enterprises, 4001 Parkway Dr., Bossier City, LA 71112. The order line phone number is (381) 631-3081. Well, I'd better shut up and get on with the loggings or we'll be out of room. Here's what you've heard this month (in an abbreviated form, alas!): January 1993 Utility World Logs Frequency Mode UTC Callsign Station/Remarks/Contributor 5754.00 RTTY 0830 RFFGCEM UNID/ARQ-E 72 baud/Jansen-NED 5847.50 RTTY 0830 RFFDBCM UNID/ARQ-E 72 baud/Jansen-NED 5909.00 CW 1517 BMB Taipei, TAIWAN/wx and ID/Berri-CA 8474.00 CW 2300 HCG Guayaquil, ECQ/marker/George-MA 9057.00 RTTY 1547 DFZG MFA Belgrade, YUGO/75 baud Baudot/Jansen-NED 14372.00 RTTY 0955 SNN299 MFA Warsaw, POL/Pol-ARQ 100 baud/Jansen-NED 15703.00 RTTY 1543 YZJ6 Belgrade, YUGO/TANJUG FF Nx 50/400/Wolfson-OH 16065.00 RTTY 1709 IRO30 Rome, ITA/ANSA EE Press 50/500/Wolfson-OH 17535.07 RTTY 0945 V5G MFA Bucharest, ROU/ROU-FEC 164.5 baud/Cooper-UK 19808.00 RTTY 1202 Lagos, NIG/SWED-ARQ 100/415 (Norwegian Embassy, Lagos!)/Hall-RSA 20004.50 RTTY 1446 CLP1 MFA Havana, CUB/100 baud Baudot to Iraq/Jansen-NED 20198.30 LSB 1656 AFE70 Cape Canaveral, FLA/STS52 Shuttle Launch/Lukas-NY 20558.20 RTTY 1625 5AQ88 Tripoli, LIBYA/JANA EE Nx 50/400/Wolfson-OH 26725.5 CW 1538 NAR Key West, FLA/marker/Lukas-NY ------------------------------------------------------------------- Sorry about the meagre logs this month, but our page allocation is firm. I'll concentrate on the logs next month. Keep 'em coming to the address in the masthead. 73 until next time! ABOUT SPEEDX (The Society to Preserve the Engrossing Enjoyment of DXing). Speedx is a non-profit volunteer hobby organization founded in 1971. SPEEDX is a member of the Association of North American Radio Clubs (ANARC). The Club publishes the monthly bulletin _Shortwave_Radio_ Today_, 48-pages of up-to-the-minute DX news from both Broadcast SW Stations as well as Utility News. If you would like more information regarding SPEEDX or apply for membership, please send all inquiries to: Mr. Bob Thunberg, Business Manager SPEEDX P.O. Box 196 DuBois, Pennsylvania 15801-0196 United States of America