Dead Silent, Dead Battery, Dead Stupid

The 2m section of my Icom IC-706mkII has finally died. Unfortunately, it picked a wrong time to go out: the moment I hit the interstate to head up to the Belton Hamfest. Helluva time to be without a voice, but oh well. Luckily, I had my Baofeng UV-82HP on hand and a connector for the magmount.

The fan mod I did on the Icom was too little, too late. The transistor was probably already on the ragged edge of viability, and I had the habit of running it at full power (20W) for the local repeaters (tsk-tsk), but even with power dialed back, it still had strange issues. I think the 706 just didn’t take too well to being a mobile rig in this Texas heat.

The final behavior is that when PTT was pressed, the power indicator would go to maximum, the transmit light turned red, and no actual signal was emitted (no kerchunk returned). With the dome light on, I could detect a definite power sag with the light dimming when I pushed PTT, so that tells me the transistor was likely just shorting the rails and slurring the signal enough to not really be readable by anything.

My best hope now is that the HF section is still OK, since I seldom used it in the car. I’ve pulled the radio; it’s back in the house. I’ll check it out and see if I can still use it as a backup/portable HF rig.

In the meanwhile, I’ve accelerated my plan to replace it with the Yaesu FT-8500 I bought months ago (I mean, who’s really got the time?). Well now I have all the impetus to get it installed, and boy have I. Thankfully, the extension kit fits the car, and I didn’t have to make any modifications to fit it to my gooseneck mount. Easy-peasy.

However, programming this in bulk is going to be a bitch, since Yaesu’s menuing system is deep. I just wish I could figure out a PC programming solution. But that’s another battle for another day.

The current battle, though, is one that is sneaky, and there are very few references to it online. But, once I figured out the magic words, I found a few newsgroup posts where other hams with the same radio talked about the FT-8500 forgetting its most recent frequency and mode if it is powered down more than 30 seconds. Every time you power it on, it starts in VFO mode with a pair of frequencies that aren’t in any sort of memory, and the UHF section has primary status, with squelch open on both sections. The only way out of that is to either dial another frequency, or press “D/MR” to switch to a memory channel after manually clicking the VHF knob. Complete bullshit, really.

Well, what the user manual never tells you is that the stupid radio has a hidden CR2016 coin cell behind the faceplate. Yeah, it has a battery. Not for the memories: those are stored in EEPROM. It stores the most recent state (frequency, mode, main channel status, etc.) between power-ons.  Without it, you have to manually power it on when car power is restored, manually select the VHF section, and manually select a memory channel to make it operational.

All this time, I had thought it was either a defect, a usability quirk in design, or a secret option setting that wasn’t in any published docs. Nope, it’s just that fucking CR2016 battery with solder tabs that has drained in the X years since the previous owner used the radio.

I gotta take the radio back out of the car, disassemble it, desolder the battery, take it to the battery store for a replacement, have them weld new tabs on, and then take it back home to replace, reassemble, and reinstall. Pain in the ass.

Radio is hard.

Winlink 2000 – Email on the Wind

Since I have my ham rig already set up for soundcard modes, I decided to give WinLink 2000 a try. WL2K is an email-over-the-air transport system that can take advantage of many modulation modes and frequencies, and allows a remote station with no Internet access to send and receive email messages complete with attachments. It’s useful for ship/boat stations, emergency communications during disasters, mobile/roving stations, camping/hiking stations, as well as a great thing you can do to get better at long-distance digital modes for fun and practice.

There’s a moderate amount of work that needs to happen up-front in order to use it. The Winlink service requires a valid amateur radio license from your country’s licensing authority. You will need to install some client application in order to compose/read mails and orchestrate communications. Once you sign up and make your first connection to the Winlink system, you will receive an email with your on-air password, which is then required to make subsequent connections. Your email address will be <callsign>@winlink.org and your inbound messages (to you) are filtered by a whitelist that is fairly restrictive. You can edit your whitelist on the Winlink web portal if you’re connected to the Internet, but otherwise any email address that you send to through the system will get added to the whitelist automatically; all other inbound mail is rejected.

On my home workstation, I installed and configured Winlink Express (published by the Winlink group). Winlink Express requests that you install the VOACAP (published by the Voice of America) propagation estimation software if you want to use its automatic link quality estimates. You’ll also need to install whatever software will be the modem between Winlink and your soundcard/radio (I used Direwolf in this situation).

The system can use several modes. The first, and simplest, is to use a basic telnet connection on an Internet-connected workstation to transfer messages. Your client software talks directly to the Winlink servers via the Internet. This is how I set up my account and got the basics worked out.

Another mode is Packet, which is your basic AX.25 packet radio mode, either 1200 or 9600 baud, “connected mode”. This is primarily used on VHF/UHF frequencies. If you don’t have a hardware TNC (Terminal Node Controller) that runs in Packet KISS mode, you can definitely use the Direwolf software TNC and tell Winlink to use it at a local network address and port instead of a hardware TNC at the other end of a serial connection. Direwolf requires some configuration from defaults to use your audio device. This is a good exercise to help you learn how to stitch multiple software components together to perform the work.

The third mode is Pactor, and is the mode of choice for anyone who happens to own an external Pactor TNC. This is a high-efficiency mode with extra error correction, signal processing, and high bandwidth (for MARS bands or if the FCC allows during disasters), but the external modems are expensive and outside the budget of the casual ham.

The built-in mode is Winmor (published by the Winlink software team). It is fairly robust and has a lot of handshaking and error correction, and the Winmor software TNC runs in a graphical dialog window to show you the signal quality, current action, and status of the transfer. Winmor uses QAM (Quadrature Amplitude Modulation) at certain times of the conversation with various constellation densities and bits-per-second speeds to fit what’s best for the HF link condition.

Winlink supports many more modes that I haven’t experimented with yet.

The Winlink system relies on a worldwide group of volunteers who run stations that listen on a published list of frequencies for remote stations that wish to connect. They will rotate through their frequencies until they hear a request. They then do a song and dance with the remote, and forward any mail between the Internet and the remote station. It is essential that you download the latest published list of RMS stations before you go offline. You can also request the latest list over the radio, but you’ll need to know at least one station you can reach for that to work.

Aside: you can also request help files, propagation reports, weather reports, etc., automatically with a click with Winlink Express; the Winlink base station will provide a copy on connection.

One thing that tripped me was the option to connect to a Winlink station versus connecting to a “radio-only” system. Winlink architecture allows stations to connect radio-only or P2P, allowing them to shuffle email around without an Internet connection. My assumption is that if a radio-only station gets a message and accepts it, it will use the next available time slot to start up an HF link to the next reliable station and pass it off until it eventually gets to where it needs to be. Not quite.

On the times I tried to connect to a station to transfer, it got through the handshake, my client saw that the outbound mail I had wasn’t destined for that host, and both sides closed the connection without transferring anything. In that situation, I should’ve used “Connect to Winlink” instead of “Connect to Radio Only”. But OK. Lesson learned.

I was highly successful with Packet, because there’s a station W3MRC here in Austin which is a gateway. I was able to transfer mail at less than 10W with a 2m antenna. However, transferring on the HF bands is significantly more difficult. The concerns with Winlink on HF are the same as any other HF mode: propagation is important, and so is your radio, antenna, feedline, etc. If they can’t hear you in the noise, you’re gonna have a hard time.

So the software will show you a list of stations (a “channel list”) sorted by estimated link quality. Your best bet is to find something with a high estimate. But keep in mind the frequency you’re using, the time of day, and the skip zone (ionosonde data is useful here). It’s no good trying a station 500km away if you’re on 40m at night and your skip zone is 1500km. I tried to reach stations in Ohio, California, Baja Mexico, Florida, from my Texas apartment on 40m, but my crappy dipole just couldn’t get enough energy into the atmosphere for anyone to hear me. So my best success was to reach station K0SI in Missouri (which is in the skip zone, so I dunno about that), which heard me well enough, even with the fading, to do the job.

I’d really, really love to experiment more with this, to get better at it and learn how to set up a great antenna for it (groundplane vertical would be nice), to learn how to set my audio filters to really narrow the audio down to the middle 1KHz of audio in the passband, and to find that favorite collection of remote stations that are highly reachable in case of emergency.

One final note on this rambling post: watch your radio power. This is a digital mode, and your radio’s duty cycle could approach 100% depending on transfer action. You could overheat your final amplifier or exceed your radio’s ability to cool itself. Luckily, my Yaesu FT-857 limits or shuts off the output power if it overheats to save itself. But unless I’m watching the needles on my antenna tuner, I wouldn’t know it was not transmitting. If your side suddenly stops being heard, the other station is left in the lurch and in a bad state, making it unusable by anybody until its timeout expires. Don’t be that guy. Drop your power to something reasonable; I had to set mine to 40W, but the audio level from the soundcard is only driving it to 30W peak, and still that’s almost too hot. Be ready for fan noise. A great antenna would be your friend in this case.

Overall, this has been a positive experience. I hope to use it more, and for more legitimate purposes than sending myself test messages. It’s a super-useful system, and it’s during these relaxed times that we practice and get better so we can do it reliably during stressful times.

Icom IC-706mkII Fan Mod

My mobile rig is an Icom IC-706mkII HF/2m unit that’s been doing me well for months, semi-permanently installed in the car. Lately I’ve been noticing signs of its deterioration, getting signal reports that I’m either cutting out or the audio has a sporadic “electric shock” sound. I’ve been trying like hell to track it down, thinking my custom power cables were either sub-par, that my car’s electrical system (battery, alternator, regulator) were getting edgy, or that my final transistors were starting to fry. All of those options were scary to consider.

The symptoms exhibit more strongly when it’s hot outside and I’m chatting away on a local repeater on the afternoon commute. Eventually, the radio starts freaking out and I have to sign off.

As it turns out, my radio is overheating. The internal fan’s not doing its job.

I took it inside for examination. With the covers off, I put the radio on CW mode, half power, into a dummy load, and held down the key. The fan would only run while I had the key or PTT down, and never outside of that. Even when it was key-down, the fan took a long time to kick over; the driver circuit would attempt to start the fan, but didn’t have enough voltage to push the fan blades except after a few kickover attempts. So it never gets cooling if my talk times are less than 15 seconds (and the heat builds up over the QSO). I was afraid the fan was dying, but when I removed it and drove it with 12v, it blew like a champ (thankfully). So something is wrong with the fan-control thermal circuit.

I found a few references on various radio boards where others have had the same symptom on their own Icom IC-706mkII (and the mkIIg as well). Apparently, the original Icom IC-706 was designed so that the fan would blow constantly. When they designed the mkII and mkIIg, they added a fan circuit to limit the noise and current drain. Unfortunately, once that circuit gets marginal, it stops being useful and actually contributes to the radio’s demise.

Among the references is a rework involving the addition of a 200Ω 1W resistor between L50 and J2 (on the IC-706mkII, at least) which will provide a constant voltage to keep the fan moving at a slower speed. The benefit is that the fan controller won’t need to start the fan; it just ramps up to the right speed.

Two 100 ohm 1W resistors in series for 200 ohms (identified by screwdriver tip), wired between Vcc and the fan.

I soldered a pair of 100Ω in series and shrinkwrapped all connections so they don’t contact the radio circuit (I left the body of the resistors uncovered for cooling), then flew the rework over the board between solder points. It doesn’t appear to wobble or vibrate much, and there’s enough cooling inside that hot case to keep it from frying.

Orange wire (identified by screwdriver tip) soldered to coil L50.
Orange wire (identified by screwdriver tip) soldered to hot side of fan connector J2.

It’s been a week since the rework, and the radio’s still doing OK in the car. I’m still getting spurious reports of noise, but I think my radio’s got an EMI sensitivity when I drive near electric utility substations (EMI/RFI has always been a problem with my 706). But otherwise, it’s doing alright.

It sucks when we have to modify a production device due to engineering mistakes, but thankfully we have the public resources to help us find our way and stay on the air.

Pinch To Talk

Behold my score from Austin Summerfest:

MFJ-564 iambic paddle

I’ve had an iambic paddle on my ham radio grocery list for a while, and now I have one. I guess this means I have to get better at Morse Code to be able to use it frequently. Right?

All shined up and ready to go out.

Picked it up second-hand for about forty bucks. To make it mine, I removed the 1/4″ plug and soldered my own 1/8″ plug to make it compatible with my Yaesu FT-857. Then I tore it down and gave it a complete spit-shine with alcohol wipes, eyeglass cleaner, and a lot of polishing.

I noticed this paddle is representative of the mid-grade build quality of some MFJ products: the heavy metal base is merely chrome-plated, and some of the base has small pock-mark oxidation. The nylon insulators and pivots are a little worse for wear. Some of the adjustment screws could use a dot of threadlock to keep them from walking out of their loose tolerances. But, overall, it’s still solid and highly usable.

Piiiiinch For Looooong and Shoooort Beeps Aaaauuu-to-maaaa-tic-lyyy

To operate any paddle, you simply press sideways in one direction to make a “dit” and the other to make a “dah”. What makes this “iambic” is that there are two paddles, one for dit and one for dah, and if you squeeze both at the same time, you get an automatically-generated string of “dit-dah-dit-dah-dit-dah“, much like the iambic meter you find in poetry.

The paddle itself doesn’t generate beeps. It’s just a specialized switch. You need an external keyer to take the paddle inputs and generate an output for the radio. Luckily, most radios have a keyer circuit built-in. Mine has a bunch of options like key speed and ratio between dit and dah lengths (1:3 is common).

For the record, there is no hard rule for which direction on a paddle is dit or dah. That’s up to the individual to sort out what is comfortable. Many right-handed operators choose to put dit on the left and dah on the right. It can always be reversed by crossing the connectors or using the radio’s menu if it has that option.

Now that I have all I need to send code (a paddle and a straight key), my last obstacle is myself. Knowing how to send is half the equation: I need to learn how to copy code, to hear what someone else is sending and transcribe it (speed experts do it all mentally and even recognize whole words by sound). That’s the hard part, and there’s just no way for me to do it other than with lots of listening practice.

Wish me luck.

Fat, Thick, and Dirty

I finally did it. I soldered a pair of PL-259 connectors (screw-on type) onto a 28′ (8m) length of scavenged RG-8 cable. And it works. It really, really works.

PL-259 soldered connectors RG8 coax
PL-259 connectors, kinda soldered, and they actually work.

Of course, they’re really, really dirty.

PL-259 soldered connectors RG8 coax
Dirty solder. So, so dirty.

And it took me 2 hours to f’n do it.

But it works.