Wednesday, August 26, 2009

You, um, expecting trouble? BIG Batteries...


Through a contact at an SF Bay Area web hosting company I recently obtained for free a set of used -- but still very strong -- UPS batteries; the Deka/Unigy 31HR5000. This hosting company offers their customers a 100% uptime guarantee which means that they can't wait for their UPS batteries to fail; they are used for a number of months and then they're replaced. Since the hosting company has to pay someone to come haul them away they're happy to see them repurposed.

These batteries aren't the typical deep-cycle marine variety you see on Field Day. They weigh 95 lbs each. Fresh from the factory they're rated at 135 amp-hours, which is about 1,800 watt-hours. My main HF rig (Kenwood TS-2000) consumes about 200 watts worst-case (when keyed on 2 meter FM, set for 100 watts) which means that using one of the Deka batteries I could leave my rig keyed-down on 146.520 FM @ 100 watts and it would remain on the air for about 9 hours.

I now have 3 of these monsters running in parallel on my home station. That's over a full day of continuous talk time. Left in receive-only mode my station will run off-grid for about 8 days.

Total cost? About $40 for some 4-gauge jumpers and misc hardware.

Bring on the zombie attack. I'm ready.

Monday, August 10, 2009

Secret History of the Silicon Valley


I'm really enjoying Steve Blank's article series on the Secret History of the Silicon Valley. A very well-written series that has also been presented live at venues such as Google TechTalks, etc.

Most interesting for me is the confirmation of something I've believed for a long time; the "Silicon" Valley might might just as well have been named the "Wireless" Valley.

Frankly, I prefer the latter...

Sunday, May 31, 2009

-- --- .-. ... . -.-. --- -.. .


For a while now I've been forcing myself to spend time studying Morse Code. It's no longer required for amateur radio but I felt I needed to have some proficiency in it, and it's something that I've never been able to quite grasp. I've always believed that I simply don't have an "ear" for code; some people can hear it and some can't. Now I think that's probably true for fast code but I believe now that with some effort an "effective speed" of 6 words-per-minute can be achieved.

The method I've been using might seem unusual. I started out by obtaining W6TJP's Code Quick audio CDs. This is an innovative method for learning which teaches you to associate a "sound-alike" and a humorous image with every Morse character; for example when you hear "DAH DAH dit dit DAH DAH" the sound-alike is "Coma, it's a coma" the image is a bear laying in a hospital bed, and thus the character is "comma" (coma). W6TJP claims that this method ties the Morse Code sounds into your brain's language center, and I can believe that it in fact does. For slow code, to pass a basic test or decode repeater IDs it works and it's good enough.

The problem is that it takes time to mentally process the sound-alike, the funny image, and then recognize the character. So there's an upper limit to the "effective speed" you can reach with this method. I should make a note here about the difference between "effective speed" and "character speed". Character speed is a function of how long the dots and dashes last, and their timing relationship to each other. Effective speed is a function of the duration in pauses between characters. You can send Morse Code at a character speed of 18 wpm, but at an effective speed of only 6 wpm; this gives the receiver time to process each character before the next is sent.

Thus once you've mastered the sound-alikes, and you want to increase your effective speed, you need to "unlearn" the sound-alikes and learn to hear the code directly. One method for doing this is called the "Koch Method" where you start out with two characters and after you reach 90% correct copy you add another character. This type of method typically requires a computer to handle generation of the audio and "grading" of what you type in response. The best trainer I've found for this is "Learn CW Online" at www.lcwo.net. The reasons LCWO is so great are (1) it's free and (2) unlike a lot of websites done by hams it's very well designed.

I started out with LCWO doing fairly well; the number of characters was small (the first 4 taught are K, M, U and R) so if I got stuck I knew the right answer had to be one of those. As the lessons progressed I no longer had that luxury; the answer could have been any one of 40 characters (26 letters, 10 digits, plus comma, period, slash, and equals sign). And yet I also found that I was actually more accurate than I thought I would be! As you increase the effective speed you can't dwell on each character; you have to make a choice and move on. If you get stuck you'll likely not only miss the character in question but also the next few after. So in some cases I'll hear a character but won't be sure, and will just type what I think it is. At the end of the session I'll be thinking "Well, I must have really flubbed this one" but in reality I only missed one character out of 40. So clearly there's some kind of subconcious connection being formed here between my ears and my fingers which is bypassing the rational/analytical part of my brain.

I'm not sure I'll ever be one of those 50+ wpm code guys who hears entire Morse Code sentences in his head, but with luck I might be able to actually hold an on-the-air conversation using Morse Code some day.