"When the going gets tough, the tough get going." When one ham decides to tinker with packet radio, he kept on going and did not give up until he had achieved what he set out to do.

Varon Smith, WS6V, monitors the 9600 baud VHF TelPac station passing its traffic to the PMBO. To the left of the computer screen sits the Yaesu FT-7800 (top); the Alinco DM-330MV power supply (left) and low-profile Kantronics KPC-9612 plus are under the small hutch. A second FT-7800 transceiver and a duplexer later made the station simultaneously 1200 and 9600 baud capable. [Courtesy Denis Franklin, W6EW]

The author (center) sits at the Airmail station, demonstrating Airmail-TelPac functions to EmComm enthusiasts from several Bay Area ARES and RACES organizations. Left to right are Michael Borkon, KF6CRZ; Tom Talezaar, KG6MAC; Peter Hoffman, W6DEI; Doc; Mark Nowell, WI7YN, and Varon Smith, WS6V. [Michael Hole, KG6DER, Photo]

Here are the HF and VHF Airmail stations, consisting of the Dell laptop in the foreground, the SCS PTC II Pro Pactor3 controller (center right) with the VHF module installed. If the Airmail HF Terminal window is selected on the screen, traffic is sent via the Yaesu FT-897 (center left) with computer-selected PMBO and frequency. If "Packet Client" (VHF/UHF) is selected from the Airmail modules menu, traffic is passed via the Yaesu FT-7800 at top left. Set up for the 9600 baud VHF test (using 5 W and dummy loads), the controller and Yaesu can be operated at 1200 baud if connected by the spare data cable stowed just to the left of the Heath scope. [Courtesy Denis Franklin, W6EW]

Only requiring an additional Yaesu FT-7800 to make it 1200/9600 baud capable, this TelPac/PacLink station is shown in its present configuration, with the TelPac software running on the screen. The Kantronics KPC-9612, lower right under the hutch, is dual ported and the ports are defined in the AGWPE software for the TelPac and PacLink programs. The 32 A Alinco DR 330 MV power supply is ample for the two radios rated at 8.5 A on transmission. The TNC draws on the order of 25 mA at 12 V dc. [Courtesy Denis Franklin, W6EW]

After more than seven months effort, I was about ready to give up on 9600 baud Winlink 2000 (WL2K) VHF. Just in time, I finally found the adjustment that made it all work, at least from one side of my shack to the other. I went from discouraged to elated faster than you can say TelPac!

Judging from my own experience, and what I've learned from others since I undertook this project, using 9600 baud packet in a given emergency communications environment remains an open question with both technological and strategic considerations.

Without a foundation of expertise in communications software and a broad general experience with packet radio, getting a WL2K VHF station operating at 9600 baud can be a formidable undertaking. Even if 9600 baud WL2K communications are accomplished, there remains the question of whether high speed will, in practice, result in more emergency traffic throughput than the much larger number of existing 1200 baud stations might provide.

Once Upon a Time...

About 18 months before beginning my at-sea project (see the February 2005 issue of QST, pages 47-50), I made an enthusiastic recommendation to my ARES/RACES teammates that we find a way to incorporate Airmail/WL2K HF e-mail communications into our backup plans for the Alameda County California EOC we serve. ARRL Pacific Division Director Bob Vallio, W6RGG, and ARRL Pacific Division Vice Director Andy Opel, N6AJO (our RACES Chief and Assistant Chief Radio Officers, respectively), told me that the League was already looking into the VHF/UHF version as an adjunct to EmComm projects around the country. Accordingly, I determined to master the complexities of WL2K VHF and learn enough about packet radio to get a demo station up and running, both for my own amusement and to show our county officials what Winlink could do.

I had a hunch that the only way for me to truly understand this would be from hands-on experience. Moreover, I had some gear I thought might be dedicated to the project, including an AEA Packratt 232 controller, an Alinco rig with a built-in TNC, a Yaesu FT-5100 (with external data jack) and even a Kenwood D700 with built-in TNC. But it turns out that the WL2K system is more computer than radio. Ask any ham who has used computers for a while -- virtually nothing new will ever work with anything you already have on your desk. It's a rule. So I discovered the many incompatibilities that can (and do!) happen from the very beginning of the project.

Considering that the documentation for various WL2K software modules has been produced piecemeal, and that both software and hardware are modified frequently, it is surprising how helpful the instructions remain. Omissions of key facts, however, seem to follow Murphy's Law: That which you need most will be missing.

After reading about 200 pages of Web site descriptions of the system several times, five minutes afterward I still couldn't remember the difference between TelPac and PacLink. But I did learn how to download and install all the software modules, along with AGWPE, and did so onto two computers, one running Windows Me, the other a Mac iBook running Virtual PC with Windows 2000 as the operating system (OS).

Talk to the Expert

The PC also had a copy of Airmail working on HF, and was at that time attached to the Packratt 232 and Yaesu FT-5100. But when I clicked on "VHF Module" in the Airmail menu, nothing happened. After days of re-reading the instructions and still not getting anywhere, I hollered for help, calling WL2K guru Steve Waterman, K4CJX.

Steve told me that first of all I should run 9600 baud, which I accepted provisionally on faith. The FT-5100 would not run 9600 baud, he said, nor would the Packratt 232. None of the software would work with my Alinco or Kenwood "built-in" TNCs. Moreover, he explained, TelPac and PacLink may not run reliably on an operating system other than Windows XP Home Edition. On the other hand, the WL2K software was known to work with the Kantronics KPC-9612plus, he said, and the Yaesu FT-7800 would operate at 9600 baud.

Okay, so much for economizing. I believed that this demonstration project was really important to our local EmComm efforts and decided to purchase whatever was necessary to get it to work. Of course there may be other choices that also work, but Steve was recommending software and gear he knew would work.

In another blow to my wallet, Virtual PC had quit running when I had upgraded my iBook operating system to Tiger, Mac's latest. Moreover, the Virtual PC I had bought did not include the XP operating system. At a loss of what to do, I bought a new Dell PC with the anointed XP OS and upgraded the OS in my other PC to the same. Then I did fresh downloads of all the WL2K software modules onto both machines. From Gary Wood at Farallon Electronics I learned that Airmail was looking for a VHF module in my SCS PTC II pro (Pactor 3) modem, which wasn't there. He installed the latest DSP model (that handles both 1200 and 9600), and I replaced the Packratt with the upgraded SCS modem. All of a sudden, Airmail could find the "VHF module."

Next came a second-hand Kantronics KPC-9612plus modem and two Yaesu FT-7800 VHF/UHF transceivers. We later discovered the KPC-9612plus to be defective when the software would not recognize it. We replaced it with a brand new one while the original went back to Kansas for some TLC.

During a couple of days tracking down the pin-out diagrams in several manuals, I soldered up the data cables and hooked everything up. In the "Airmail" station (the "end-user" station in the field), my old Dell laptop was connected to the SCS modem with VHF module, and that to an FT-7800. The TelPac station, designed to receive packet and route it onto the Internet, consisted of the new Dell laptop, cabled to the KPC-9612plus, which in turn is cabled to the other FT-7800 transceiver.

Prepared, I took Steve up on his kind invitation to call him any time to get help with the software setup and troubleshooting process. He ran through all the entries in the setups for Airmail on the "field" station and TelPac and AGWPE on the other. He explained that although TelPac could run directly on the KPC-9612plus, AGWPE was designed to talk to a multitude of different kinds of TNCs, while presenting a friendly face and easy connection to the TelPac software. So in the end you can just turn on the station and start AGWPE, which in turn will auto-start TelPac.

The main distinction between operating at 1200 baud packet and 9600 baud, aside from the speed difference, is that the setting of the deviation for 9600 is both more critical and more difficult. Namely it can't be done "by ear" by listening to tones because the sound of packet at 9600 is a rough hiss of white noise. Having no clue as to what XMITLVL stetting of the KPC-9612 would result in the desired 2.8-3 KHz deviation from the Yaesu, we started by guessing at the settings for the both modems.

Practice Makes (Almost) Perfect

With Steve on the phone, I conducted the first test by composing a short message in Airmail and sending it through TelPac to my regular e-mail address. When I hit the green "Send" button, I was rewarded by an immediate exchange of "shhhhups" between the two stations (these were on desks on opposite sides of my shack). But no traffic passed and eventually Airmail hit its limit for numbers of tries and went silent. Theoretically, Airmail would have automatically passed any outgoing traffic to TelPac, which would have connected to my DSL line and sent it to a WL2K PMBO station. Fortunately, I had elected to activate AGWPE's port monitor windows, revealing that TelPac had heard the Airmail station and repeatedly told it to send the traffic, which it never did.

As you know, a diagnostic problem becomes logarithmically more complex when there are multiple variables. This may explain why it took me another six weeks of read this, then test this, read this, then try that to prove it wasn't the cables, the receivers or the other modem that was failing to pass along the "go ahead and send" message from the KPC-9612plus. All of this while waiting for a back-ordered deviation meter that arrived only after the project had been completed.

I found it hard to read the Kantronics 279-page manual on my computer screen, so I took the CD to Kinko's and had it printed. Finally, on page 236 of the KPC-9612plus's manual I read, "For port 2 the drive level is...80 mV p-p to 4 V p-p... Default drive level is about 1 V, corresponding to a XMITLVL command parameter setting of 64."

I realized that the default was at the low end of the range. And I remembered that at a setting of 100 I had once gotten one sentence to pass, although it took nearly 30 minutes; we had decided the delay was due to a super-high error rate. I immediately tried setting the level to 150, 200 and 250. At 250 the error rate was again extremely high. Backing down the XMITLVL to 175 I finally got traffic moving at information superhighway speeds, just eight months after starting the project. Well, to be fair there had been a one-month trip to sea on an oceanographic ship and another month spending the holidays with my grandkids, but it still took a long time.

Since there were no other 9600 baud WL2K stations in the San Francisco Bay Area at the time, I was only able to talk to myself. Total cost at that point: More than $2600, half of that for the new laptop; however, the demo stations must have successfully proven the value of adding Winlink VHF/UHF packet to our RACES capabilities. Subsequently we received a Homeland Security "interoperability" grant and were able to replicate my demo stations at the County Emergency Operations Center (EOC). This has provided the county with the beginning of a "backbone" of 9600 baud EmComm stations. In addition, we added another Yaesu FT-7800 to each of the Telpac stations. Now operating at both speeds, we can also interconnect with the growing number of 1200 baud TelPacs in the Bay Area.

The Moral

The point of this cautionary tale is that for most hams not already active on 9600 baud packet, that capability may be a long way down a twisty road. If the ham is not a communications programmer or already skilled at packet applications, there will be very little guiding documentation that does not leave out just the one piece of simple information needed. To attempt the project, you really must have the help of an accessible Elmer. Eventually you're going to need an FM service monitor or deviation meter of some sort, so you might as well arrange for the use of one from the beginning.

Using WL2K HF and VHF for emergency communications is still a wonderful idea. Here's my current understanding of how the modules can work:

Airmail VHF is a packet radio program that itself looks like an e-mail program. It will connect to and pass outgoing and incoming e-mails (including binary files such as diagrams and photos) automatically from a field station lacking an internet connection to any TelPac station or PMBO of the WL2K system, and then into a working part of the Internet.

TelPac stations require an Internet connection. They receive packet radio traffic on behalf of the PMBOs where the traffic is immediately and automatically relayed. During the same connection, they send any waiting messages back to the sending station. These stations receive Airmail connections and pass traffic in an automatic mode, or they can be contacted by ordinary packet stations, passing message traffic to and from the PMBO via the Internet.

PacLink is just a little more complicated; completing its installation will be my next project. Simply put, it can, using a mail program like Outlook, automatically send and receive e-mail for an individual, or, when connected to a Local Area Network (LAN), like the EOC, it can become an additional emergency mail server for key computers on that network. It will send the mail traffic out of the EOC in any of several ways, switching automatically according to a specified order of priority. For instance, it can use the Internet connection if present, and if that's lost, it can switch to sending traffic by packet to a TelPac or PMBO station outside the disaster area, which will still have an Internet connection.

Strategic Considerations: 9600 versus 1200 baud

I'm not an EmComm expert, but as a ham for 40 years I was the project director that set up the paramedic system (including communications) in a Bay Area county for a population of half a million people. I helped coordinate our medical communications with nine other counties (more than five million people). And I was one of thousands of rescuers in Sioux City, Iowa, for eight days after the crash of Flight 232 in 1989. I've come to believe in the principle that what works in an emergency is a system that is being used every day.

I also believe in the slogan "Keep it simple!" I like to think it is wise to go with what you've got, and I think there's a lot of underutilized 1200 baud packet equipment in a lot of ham shacks.

Of the roughly 600 TelPac stations listed on the WL2K Web site, I count only about 20 in the US who run at 9600 baud. (Steve, K4CJX, who knows more about all this than I do, says that there are more than that in Tennessee alone and he may be right.) All the rest appear to run at 1200 baud. And the race goes not always to the swift.

To convert to 9600 baud is, by amateur standards, expensive, and by any standard, difficult to accomplish. I may not be the sharpest knife in the drawer, but my learning curve is at least average. Plus, I'm retired and have lots of time to work on such projects, but as you have seen, getting a 9600 baud WL2K station up and running has been no walk in the park for me.

To make it even more difficult, before adding 1200 baud, my station was useless to most packeteers in the Bay Area. While it makes sense to construct some "backbone" WL2K networks at 9600 baud, in so doing we must not get rid of the much larger resource -- the slower speed packet hams in a disaster area. Remember, even without the Airmail software, such hams could send and receive ordinary packet messages through a TelPac station to and from e-mail addresses, if the TelPac operates at 1200 baud.

At the conclusion of this project I asked Steve whether, in the light of these considerations, it would be better to run my TelPac node at 9600 or 1200 baud. He responded, "You need to do both." This, it turned out, is entirely possible with one KPC 9612plus, since it is "dual port." By adding another transceiver on 1200 baud, both speeds can be run simultaneously.

I went on line to Winlink 2000 for Disaster Recovery/Emergency Preparedness on Yahoo Groups, where the gurus hang out. Rick Meuthing, KN6KB, directed me to the example file that would set TelPac to support both the 9612plus's ports, one at 1200 baud and one at 9600. But it turned out this can also be done in the AGWPE port set-up, and for me that was the easier way to go.

We now have an additional dual-speed TelPac station at the Alameda County EOC, as well as two complete Airmail HF/VHF/UHF stations set to 9600 baud on the VHF/UHF side. One will be used at the EOC or deployed where needed in a laptop case, while the other will soon be installed as a complete mobile Airmail station in the new Sheriff's Communication Team van.

Work has just been completed on yet another VHF/UHF Airmail station set up at my home. Operating at 1200 baud, it is being used to locate and log the new Bay Area TelPac stations that are appearing at the rate of one every few weeks. In due course, we hope to see the kind of dense distribution of Winlink TelPac using 9600 baud among our own county stations. We are prepared to be available for 1200 baud stations and to make use of every possible TelPac and PMBO resource no matter what their speed.

Dr Denis "Doc" Franklin, W6EW, was first licensed as WA6HGB in 1967, the same year he began medical practice. Although his time was limited by his career, the high moments of those years came from being of service during disasters, phone patching for maritime mobiles and his physician friends while they donated overseas service, and providing needed medical advice to isolated victims of accident or illness. As a Navy MARS member, Doc helped out around the RTTY gateway station operated by his mentor, Don Johnson, W6QIE (SK) during the Vietnam War. After retirement, he wrote on medical and scientific topics. When not at sea, Doc is a member of the Alameda County Sheriff's Communications Team, a RACES group. He has recently returned from an Antarctic icebreaker, where he helped study sea ice; this is chronicled in the February 2005 issue of QST (pages 47-50). Doc currently lives in Oakland, California.

Previous: FCC Denies Petition to Require Exam Questions in Languages Other than English

Next: ARRL Seeks to Force FCC to Produce Documents on BPL Complaint Dismissals