Today I went up to Everett to take my General class license test, to upgrade my license to the next higher class. It was 35 questions multiple choice, and I passed it pretty easily. Then, I was encouraged then to take the Amateur Extra class exam, even though I hadn't studied for it and didn't know all of the material. Amazingly enough, I did know enough of the material to pass the test! So I got two upgrades today for the price of one. Everybody there at the testing session was amazed that I actually passed the exam - me most of all.

So now I guess I can do more than just listen on the HF bands. It is kind of nice to look at the band plan and not have to worry about which frequencies in the band I can and can't use... since I have all amateur radio operation privileges! Now I'll have to see how that Icom IC-706MkIIG really works...

This is a portable packet station, complete with Windows Server computer, TNC, and mobile radio.

It's waterproof, portable, and pretty durable. Here's what's inside and how I made it.


The radio is an Icom IC-228H, a 45W 2m mobile radio. I interfaced its microphone input to the Kantronics KPC-9612 TNC using the gray cable as shown in the picture. The mic connector on the radio also has audio out for the TNC, so it is convenient for a single cable. The computer has an onboard ethernet, but it is not used since I have the USB WiFi card, shown on the left. Power is supplied by the battery on the left (note power switch on the top) and distributed to each of the components through Anderson Power Pole connectors and the distribution box shown.

Here is the tiny single board computer, an AMD Geode running completely fanless and on a 12V DC supply. I installed an actual 12V regulator to prevent any power spikes or over-voltage conditions from reaching the board. The board consumes less than 700mA on normal operation, even with the ram and flash drive. The 7812 regulator is easily capable of delivering up to 1500mA and the input voltage is never likely to be over 14V so it doesn't even need a heat sink. I soldered a 0.1uF capacitor across the output to absorb noise.

The CPU runs at 500MHz, pretty slow by today's standards. However it gets the job done. It has 1GB of ram on the back side of the board, and an 8GB flash drive that it uses instead of a hard disk. It is completely solid-state.

Here is the internal battery, with its own fuse and switch. The system is designed to run off an external power source, with a power cable with power pole connectors that plugs in to the distribution box. This internal battery will either run the system or charge when the switch is on.

The station uses a Kantronics KPC-9612 TNC. I added a USB WiFi stick instead of using the onboard ethernet on the motherboard. The power distribution box shown rear left distributes power to the 5 devices using 12VDC.

Here is the Icom IC-228H radio, an older single band 2m 45W mobile rig. It feeds audio and mic signals directly to the TNC.


Packaging it all up together, here is the top-down view of the box.


Here is the station, complete and running. The radio is hooked to the roof antenna, and it is tuned to 145.670MHz. The callsign is KD6GKD-4 for the RMS node, if you are in the area and wish to connect to it.

Any questions about it? Feel free to leave a comment and I'll reply.
73 de KD6GKD

"CQ Field Day, CQ Field Day from kilowatt seven romeo ... K7R!"

That phrase is now permanently burned into my synapses. After operating or logging for almost the entire time from 11am Saturday to 11am Sunday, that phrase has been said literally thousands of times. I did take a nap between 2am and 5am Sunday, but basically that was just because the noise on the 80m band was so loud and my back was thrashed from sitting at a wooden picnic table for hours and hours. We operated continuously either calling CQ or scanning and pouncing on others calling CQ. It was interesting some times that we could hear two parties clearly who couldn't hear each other very well at all.

From what I heard, the propagation wasn't very good. Most of the contacts I made were along the west coast. Alaska, BC, Washington, Oregon, California. We did get some southwest and east coast stations too, but mostly on the west coast. The antenna was a sky-loop, suspended up about 60 feet, from half a dozen ropes to tall trees all around the perimeter of the lawn. It was a pretty amazing sight for a temporary antenna. Thanks to Icom for letting us borrow the 756 Pro III radios, those were pretty cool. I learned a little about the rig, but mostly just enough to operate for the contest. It sure would have been fun to make some foreign contacts for sure though. Here is a video from a guy in Seattle (K7HV) who made contact with one of our stations as his second contact.

As a team, we made lots of contacts in a lot of different modes. I personally operated the SSB station on 20m, 40m, and 80m for the day. We were operating 5A (5 simultaneous transmitters), completely independent of commercial power, on a total of three generators. I don't have the final numbers yet but I think the team logged over a thousand contacts. We used the N3FJP logging software which is user friendly but unfortunately relies on continuous network connectivity which is hard to keep running 24/7. I hooked my laptop up to the 27" wide screen LCD TV so the operator could get a clear picture of the software with the callsign and location and make sure both the operator and the logger got the data correct. That part wasn't a hardship...

The generators were fairly reliable until we realized that one was putting out only about 70V so we had to shut it down. The second one ran fine until 9PM when its solenoid failed, and the crew had to figure out how to repair the thing in the heat and growing darkness... They got the job done and got it back on line within an hour. Since we had the radios running direct from AC power from the generators, not only did we lose all wifi communications, we also lost contact time in the contest. Next time we will run all the radios off of car batteries, charged by an Astron charger running off the gen. We probably could have run the radio for 4 hours just off the battery I figure. My generator has a direct 12VDC output as well as the AC, so I could use that for a much more reliable and direct approach to emergency power.

The weather was hot as hell. And I've been there (SoCa), so I know. The thermometer under the awning at my operating station read 95 for almost the entire day Saturday. It didn't cool down to even 80 until after 9pm. There were few breezes and they only blew hot air around. It was really pretty brutal being out on the grass for that long in that heat.

We shut down Sunday at 11am, just as it was beginning to top 90 degrees again. Fortunately, we were all packed up and leaving the park by 1PM. We had a large crew, so cleanup was pretty fast. All in all there were some fun times, but the heat and bugs at night really took most of the fun out of it for me. For sure the next time I do emergency communications, I will be operating INSIDE my trailer instead of next to it.

This post details the construction of the Mobile Packet Server project that I designed earlier this month. This server is a complete Windows Server, Packet TNC, and 40W VHF radio in a complete waterproof portable package. I used what I had on hand, a 30 cal. ammo can.

Next, I cut a couple pieces of 1/4" plywood the size of the motherboard (6 7/8" square) and the power supply (6 7/8 x 2") and match drilled the holes for screws. I used #6-32 flat head screws, and zipped them in to the holes with a power screwdriver into a nut on the top side. I then used some LockTite glue to hold the boards in place. These were the only things that required attachment, everything else was just held in place by the tight spaces.



Then I installed the main board...



and then the power supply and all the power wiring...



I installed the battery, the radio, the WiFi card, and last the TNC, which slipped nicely down between the radio and the motherboard. The battery and radio are on the right side, while the motherboard and power supply are on the left. The TNC goes between them, and it is easily lifted out if needed. The laptop hard drive was stuck vertically up against the battery and the flat ribbon cable folded up between it and the motherboard so it didn't touch anything, and was cushioned and held in place.

Here is the completed design:


I installed the optional external USB ports because it was so hard to get to the ports on the main board which are just below (above in this picture) the serial cable (white cable). With the serial port, the monitor cable, and the keyboard all going into there, it was easier to run the USB mouse off the external port.

This package is still on the air (shown below, in actual operation) since construction, and it is a self-contained unit that can run on its own without any physical connection to anything else for about an hour on the 7AH gell cell battery that is built-in. This includes the 12VDC monitor, which runs off one of the power cables coming from the battery bundle. The only thing required for operation is an antenna for the VHF radio. It has a spare power pole power connection (in addition to the input power connection), and a spare coax power cable to run another TNC or GPS if needed. Complete specs are below.

This RMS packet node for Winlink services the Redmond, WA area on 145.670 MHz FM simplex. The portability of this system gives the operator a significant advantage for emergency communications. The package is rugged, and literally requires only a mag mount antenna to be on the air in about 60 seconds from opening the box. An operator could throw it in his front seat, power it off his 12V power plug, and drop a mag mount 2m antenna on the roof and be able to have a mobile packet node able to be placed at an optimum communications point to support communications personnel.



Computer: Via Epia 600 main board and CPU, 512MB ram, 60GB laptop hard drive, 4 USB ports
Software: Windows Server 2003, and RMS Packet (by W5SMM and KN6KB)
Monitor: 12VDC KDS 17" LCD display, powered by coax connection from battery
Network: 100MB ethernet and 802.11B WiFi
Node Controller: KPC-9612 dual-port TNC, powered by coax connection from battery
Power: 7AH Gell-Cell lead acid battery, and external input - power pole connection on 2' cable.
Spare power pole and coax output connections also wired in to power/charge other devices.
Radio: Icom IC-228H 40W 2m VHF, powered by power pole connection from battery

Issues
In a real-world portable situation, I would probably want everything to be insulated, just in case. Being that it's inside a metal ammo box, anything could happen if it got jolted just right. This is just a proof of concept, but it seems to be working out well.

The board puts out about 15W or so of heat I figure, and it does not escape readily now, being that all the gear is jammed up against it, and it's in a box. The CPU and GPU are fanless, so they depend on air dissipation, which isn't happening in this design. I added a small 12V fan (not shown in the pictures) that connects to the spare power pole connection, and just sits on top of the motherboard at the lower left corner of the box. That fan moves enough air to keep it all cool and happy, but it isn't mounted it just sits there. And there's no fan guard, so watch your fingers. It's loud, and probably pumps about twice as much air as needed down into the case, so I may add a SPST switch across a 470 ohm resistor, in line with the +12 for a high-low switch.

There are no "front panel" controls for power, reset. The DIN connector is buried on the motherboard, and I'd have to figure out where to mount the switches.

There is no power switch. When the battery is connected, everything is live, since everything is wired directly back to the battery (to avoid noise and ground loops). I will have to add a 20A 12V SPST toggle switch someplace, which I can wire in-line with one male and one female spade quick-connect, to turn off the battery (and turn off the unit if not under external power).

The computer is S L O W. Its an old board. The newer fanless Epia boards have a bit more horsepower, but I think pump out a few more watts of heat as well and they cost a couple hundred bux. Perhaps there's a way to get the fan at the bottom of the case somehow. There are new Via board designs such as the Pico and the Nano that might better support this design, but they are even pricier. Plus they would take some custom connectors and some fancy wiring to make them interface with the rest of the system. Perhaps after I win the lottery...

Other than these minor things, it seems to be a fairly workable design. Feel free to post any questions.
73 de KD6GKD