|
Categories: all aviation bicycle gadgets misc motorcycle theater Fri, 13 Jan 2012I am ridiculously pleased at this moment.
The new printed circuit board works perfectly! Well, very close, and close enough that I could slap this thing into a case and call it done! Initial smoke testing (apply power, see if anything goes foom) came back negative, no smoke in evidence. Swapping the ATMega chips around proved to be more troublesome than I'd expected, but careful perseverence paid off. At first, the LCD wouldn't display anything, and I had this overwhelming moment of, "Oh jeez, where do I even start?" I don't have an oscilloscope any more, so I was limited to a multimeter for testing. Fortunately, the first thing I tested ended up being the problem. The LCD's RW line (read-write? something like that) was at +5v, when it should have been at ground. I checked the layout, and, sure enough, that was the second LCD pin that I had incorrectly routed to power instead of ground. I clipped the wire, jumpered it temporarily to ground, and voila everything worked! Final tally: two mistakes total, and both of those easily worked around. Not bad for my first-ever PCB design! I've already corrected both pins on the 1.1 version of the design (which will also include an ICSP header). I'm also pondering how this thing is going to get cased, and thinking about going with a premade case for simplicity's sake. That would also mean that anyone else who wanted to build the project would have one less step involved. Overall, obviously I'm completely pleased. I didn't expect this thing to work right out of the box (to the extent that it does). Aside from two jumpers on the LCD, it's working exactly like I thought it would. When does that happen? One interesting factor that's pretty surprising is the power consumption. I figured this board would be pulling down 100-150 mA, but even at full backlight, it's only drawing 40 mA at 7-15VDC in. There must be some power-hungry component on my cobbled-together system that's not present in this board (the USB chip is the only thing I can think of), but whatever it is, it's great news! At 40 mA (call it 50 for a bit of safety-factor), a set of 3AA batteries would theoretically power this thing for 50 hours! (Not really, of course -- once the voltage drops below about 4.3, it'd probably stop working.) Any vehicle's electrical system will take this load without even noticing it. Just shy of .03 watts! That's not a typo; point-zero-three. I can't wait to get this thing in a case, although that is a completely different challenge, and one which is roughly commensurate with making the electronics in the first place, depending on how I go about it. I'm just amazed that it works, and so pleased I can call it "good enough," and move on to the next phase! Posted at 22:11 permanent link category: /gadgets Thu, 12 Jan 2012
The Penultimate in ADI Progress
After a fairly annoying day, I came home with a freshly-printed transparency of the top paste layer for the ADI board -- that is, what would soon become my solder stencil. I'm using solder paste on this project, and this allows me to apply it in a single swipe. The advice from Sparkfun on creating a solder paste stencil is that you send out your stencil files, and have a production house do it. They also say, several different ways, that this is nowhere near the critical step you might think it is -- in other words, that close is good enough. I took that advice to heart, and decided that I would give it a try in the most homebrewed way possible. I opened up the circuit board in Eagle, and selected only the tPaste layer to view. I printed this out, onto a standard laser jet transparency that I picked up at an office supply store. The box claims that it's a 5.3 mil thick sheet. When I got home, transparency and new X-acto knife in hand, I very carefully cut just inside the tPaste lines (and only cutting out a smaller square in the middle of the CR2032's minus pad). This resulted in a somewhat amateurish, but serviceable-looking stencil. Granted, I wouldn't want to do this for hundreds of SMT devices, but for my three devices, it wasn't bad. After that, I just swiped away with the paste (also from Sparkfun) and a utility razor blade. Voila! I did it twice, rejecting the first one as being a bit too messy on the BMP085, which has miniscule pads. The second attempt wasn't any better, so I just very carefully separated the adjoining solder paste blobs with the razor knife, and called it good enough. I borrowed a hotplate from a friend who's treaded this path before me, in order to do the heating. The idea (as documented by Sparkfun here) is that you apply the paste, drop your components in place, then stick the whole mess on a hotplate or skillet. You heat it up until the solder liquifies, then you pull it off and let it cool down. That is, more or less, what I did. I had checked beforehand with my handy IR thermometer, and found that this particular hotplate had a hotter spot in one area, so I determined that I would stick my board there. I cranked that sucker up to HIGH, dropped the board on the hotspot, and watched it carefully. Within about a minute, all the solder had gone liquid, and I pulled it off.
I actually think the heat was up too high (the plate itself got up over 320°C), as the board got a little scorched-looking on the bottom. I think it could have been set to a lower temperature and still done the job just fine. But scorching aside, the board looked great, and testing revealed no solder bridges on the BMP085, which was my only real concern. The downside to this approach is that I now had a handful of components stuck to the board, and any real chance of doing a phased installation, with intermittent smoke checks, was gone. So I stuffed the whole board and soldered it all up. It looks great, but I haven't taken the final step of applying 7-30 VDC to the input terminal to see what goes foom.
In documenting the LCD connector, I did discover that I made one mistake: the pin 12 connection should be to ground, not to +5v. It's easy to work around, just don't attach a wire from pin 12 to the LCD, but it's a little silly. That fix will go into the next version. My next step is to take my fate firmly by the horns, apply power, and see where the smoke escapes. For now, I think I'm going to go to bed while I can still ride the good feeling of making progress, without the disappointment of destroying components. Leave that for another day. Posted at 23:59 permanent link category: /gadgets Wed, 04 Jan 2012I finally had some time this evening to sit down with my new ADI board and see if everything actually fit (it looked like it did on paper, but you can never tell for sure).
It actually all looks really good, with only one exception -- the potentiometer in the upper right hand corner doesn't fit into its space very well, the holes for the leads are a bit too skinny. I can make it work for the first version, and be perfectly happy with it, but I'll tweak that on the next version (which I've already started modifying, and will include an ICSP header above the processor, in that blank space toward the LCD connector). Now I'm just waiting on getting a hotplate to do the surface mount parts (which I'm hoping to document for the ADI page). Well, I still have to confirm all the connections with a continuity tester, but I don't expect any problems there. I may be only a day or two from having a working one-board ADI! Update: Three things, actually. First, by straightening out the legs of the potentiometer, it slid right in, so no changes required there. Secondly, out of curiosity, I piled all the parts onto the balance to see what they weighed. Including too-long leads on most of the components, and a paper package around my 2032 battery (but no solder), the entire ADI weighs a mere 71 grams. Of course, a case and some wire is still required, but that's pretty light. Finally, I did a quick sanity-check of the connections on the board, and they all look good. I still have to do a proper checklist, but things are looking very positive for this board actually working like I want it to. Posted at 20:16 permanent link category: /gadgets Tue, 03 Jan 2012I knew this day was coming: the PCB has arrived! Only I was in for a surprise when I opened the package:
I ordered one, but I got two! Woot! Double woot! I guess they must have been shy a few square inches, so they doubled up at least my order, probably a few others too. BatchPCB aggregates a bunch of small designs into one big board, which is the size preferred by the fabrication house. I guess mine was the right size to fit into an otherwise to-be-wasted space. Hooray! I know how I'll be spending my evening... Posted at 13:55 permanent link category: /gadgets Thu, 08 Dec 2011Progress indeed! I just uploaded my first PCB design to BatchPCB, and $27 later, I should be getting a shiny new PCB in the mail in about a month. I updated the ADI page with the new circuit board layout files, and this puts me on hiatus until some time in January, when I'll get to experiment with surface-mount soldering and other such fun things. Who knows, the board might even work as expected! (I'm not holding my breath for that to be the case.) I suppose in the mean time, I can work on things like the Construction Guide, or updating the web page to be a bit more useful and less "Hey, look at me talk a lot!" Posted at 11:27 permanent link category: /gadgets Wed, 30 Nov 2011Things have not all been quiet on the Air Data Instrument front, although I haven't mentioned any of my progress here. Well, I change that now! It's not done yet. But it's getting much closer. I decided a few months back that I wanted to come up with a printed circuit board of my own, to make a much more compact version of the ADI. The whole "stick a bunch of breakout boards together" method is fine, but it's kind of bulky, and suffers from some pretty obvious weaknesses when it comes time to package it all up. I've finally (I hope) come up with a schematic and a PCB design that I think will work. I've ordered all the parts I need to finish a couple of ADI boards (so I have spares for when I release the magic smoke -- I have no plans to offer anything but documents to anyone else). Next step is, once the parts arrive, to print out a copy of this board at life size and place all the parts to see what interferes, what doesn't line up, etc. Then, if it all looks lovely, I can send the PCB design off to be fabbed, and if it comes back functional, I can confidently publish it as being finished, which would elicit a huge "Woot!" from me. Of course, that leaves the whole "putting a case around it" problem, but that's an issue for a different day. For now, I'm content to say, "Oooh, pretty PCB."
Posted at 08:56 permanent link category: /gadgets Wed, 16 Nov 2011My history with gadgets is long and varied. I've owned pocket-size computers for as long as I reasonably could, starting with the venerable HP 200 LX, which I purchased in early 1995. This tiny (for its time) wonder (for its time) ran DOS and sported a CGA screen, running for 20 hours on a set of AA batteries. It had a set of built-in apps which, if not perfect, were at least pretty good. And it had a full (including numeric pad) keyboard, which was quite typeable. I actually still have that wee beastie sitting in a box in the basement, but the lack of a backlight keeps me from seriously considering it anymore. I graduated to a Compaq iPaq running Linux (deep, deep nerd), then to a Palm Tungsten T5, and finally, after many years of holding out, to an iPhone 3GS. Now, I find myself in possession of a Motorola Droid X, which was the white-hot coolest smartphone in the world for about three weeks in late 2010. Up until the iPhone, none of my pocket computers (which I have variously called "Palmtops," "PDAs," and "smart phones") were particularly connected. The iPaq came closest, with a backpack that doubled its sleek, curved volume and would take a full-size PCMCIA card, including a wifi card. But really, I never had a connection until the iPhone. And, as things have progressed, the first really "cloud-enabled" device I've had has been the Droid. The problem, for me, is that "the cloud" is a term which is roughly synonymous with "blind trust in people who have no vested interest in your privacy." Granted, the data connection to the Internet is paid for, but the services of Google's cloud are, only in the most tenuous way possible, tied to any sort of financial transaction. Thus, Google's interest in my privacy is largely academic, or at least reputational. So, I have been unhappy about the idea of putting my data into "the cloud." Some of it, sure. Facebook, whatever. I don't put anything I actually care about there -- what goes up here (which is imported into the cloud) and there (which is the cloud, and with a decidedly antagonistic view toward your and my privacy) is, more or less, fluff. I do have data I care about, though. Obviously financial data, social security numbers and the like, sure. But I care about my schedule. It does a couple of things: first, it tells me where I need/want to be, so reliability is high on my list. Second, it tells anyone who cares to look at it exactly where I am to be found, and when. Without engaging full-on paranoia mode, there's a lot of value to certain other people in such data. For instance, people who are interested in liberating you from whatever material possessions you might not be actively guarding at a given moment. I'm sure there are many other more-or-less-frightening uses for someone's schedule. So it was with a great deal of trepidation (as I wrote about before) that I started using the Droid X, with its inherently cloud-based everything. I refused, in fact, to use the calendar, instead buying a classy paper calendar book to carry around for my personal use. But I wasn't happy with that. I don't want to carry around extra crap. So I was very pleased when the calendar silently upgraded itself to suddenly include a "phone" storage option for the calendar. This was several months ago. I gleefully ditched the paper calendar, and shoved all my temporal data into the phone's little memory banks. For a few months, all was good with this little world. My schedule was safely under my control, and unlike a paper calendar, the phone would actually make little beeping noises at me to remind me of things I'd scheduled for myself. But you can see where this is going. A few weeks ago, for no apparent reason, the alarms stopped feepling and jingling. Then they'd go off 23 minutes after the event had started (and not the 15 or 30 minutes before that I'd requested). Sometimes they didn't go off at all. It got to be very discouraging. Then, tonight, I went to look up plans around Christmas, and discovered that, to my horror, events I'd placed in September were showing up (sometimes having moved, sometimes having been duplicated) in December. You may recall that I said earlier that reliability is one of my watchwords. If I put an event in the calendar, it damn well better stay there. And you know, until this Android wheeze, it did. The iPhone, the Palm, the iPaq running Linux, the 80186-based HP 200LX, they all kept my schedule exactly as I'd entered it. Without fail. It wasn't always pretty, I didn't always love the interface, but the data was always there. Well, without belaboring the point any further, Google has failed me. They may be striving to do no evil, but neglect is about as bad, as far as I'm concerned. The paper calendar came out tonight, and I'll be transferring everything that may or may not be correctly recorded in the phone's little scattered memory banks, and acquiring a fresh 2012 calendar on the morrow. I'm no longer interested in being subject to Google's whims, evil or not. Posted at 21:34 permanent link category: /gadgets Tue, 26 Jul 2011I've been idly working on the ADI project when I've had a few moments, but I ran into something strange in the last few days. I'd make a change to the code (at the time, trying to make the metric VSI display more useful), and all of a sudden, the ADI would start rebooting at random times, or get all wacky on me. Then I recalled a discussion I'd had with my friend Clint about the practical limits of an Arduino. One of those limits (and one I hadn't considered before this conversation) is RAM. The ATmega328, the processor inside the Arduino Uno, has 2k of it. 2048 little bytes. And every string you load up, and every static or global variable, takes up some of that space, before you can even start running. I didn't even need to do the math (although once I did, it was pretty silly): I have a splash screen with 17 characters times 5 lines. I have a menu with 7 items, each of which has a 17 character title. I have at least two data-smoothing buffers (and was adding a third when I realized what was going on), of 32 entries at 4 bytes per entry. I haven't bothered to count the global and static variables, but there are probably a dozen or two. Some of them are big, complex structs taking up many dozens of bytes each. The data buffers alone account for 384 bytes. The strings take up around 200 bytes. I'm well over a quarter through my RAM, and I haven't even started to run yet. So, I pushed some of the strings out to flash (of which there's 32 kB), and shrunk the data-smoothing buffers down to 16 entries each. Suddenly, everything Just Works. It is a lovely and delightful thing when that happens. Interesting side-effect: the smoothed displays are still stable enough to be useful, and now they react faster. Woot! Posted at 14:24 permanent link category: /gadgets Mon, 25 Jul 2011I had a long idle time this weekend (this is a good thing -- it means the theater wasn't breaking), and I decided to work on the new ADI page. It's not done yet, but if you wanted to get a glimpse at what I'm preparing, you can see it in-progress: The Dangerpants Labs Air Data Instrument Posted at 11:42 permanent link category: /gadgets Mon, 18 Jul 2011I'm working on a show (which we call Patty), for which we wanted to have a magical glow coming out of a box. People were suggesting those little three-LED puck lights, and I decided we needed something a bit more dramatic. In progress (and as I write, completely wired, missing only some structural support):
Posted at 00:09 permanent link category: /gadgets Mon, 11 Jul 2011Last night I put what I'm going to call the finishing touches on the Dangerpants Labs Air Data Instrument: I finished the real-time clock implementation. With that, I think the code for the ADI is pretty well done. I only know of one bug, and one feature I might want to add: the feature is software-controlled contrast, and the bug is that when I let it sit and run for 24 hours, it stopped responding to the SET button. Since most use-cases involve turning the ADI off after 4-8 hours, it's hard to see this being a real problem. I still need to put together proper documentation (both hacker-oriented and user-oriented), but the firmware at least is in pretty good shape. I'd like to clean up the code, mostly in separating out some of the stuff packed into the main source file, but that's a relatively low priority. I think the highest priority right now is to come up with a wiring diagram, followed closely by making a more-polished project page that's actually hosted on dangerpants.com (I decided that "Dangerpants" was easier to spell and remember than "Obairlann," despite most of my efforts going into the Obairlann site). Of course, I also need to work up my lawyer-ese in the documentation, so no one sues me because they crashed their ultralight into a mountain because they were staring at a faulty ADI altitude display. I'd hope that no one would be silly enough to take a designed-as-a-novelty device and strap it onto a vehicle where it could be responsible for life or death. But hope, as we have demonstrated over and over again, is not actually enough. The hardware I've got now is getting a little silly. The real-time clock module I'm using is from Futurlec, and although it's perfectly functional, it's a bit huge (nearly the size of the Arduino board itself), and is packed onto the bottom of the Arduino. It's tempting, for the finished product, to actually design a purpose-built board, but I have a feeling I'm going to build the mess I've got into a project box and call it good enough. None of the components are very big (the coin cell on that DS1307 board is actually the biggest single component after the LCD), and a custom board would allow the whole thing to live on about a 2x2" square, which would be cool. The next update here should be when I have more documentation to show off, or maybe a fancy new project page. Might not be for a while, as theater is about to fire up again, and that has a way of taking over all my time. Posted at 12:57 permanent link category: /gadgets Sun, 26 Jun 2011I made some progress with the Air Data Instrument recently, so I decided I should make an updated video. The first video is here. Nothing too special, but keeping up with the super-cheezy video/sound effects. Posted at 21:46 permanent link category: /gadgets Wed, 01 Jun 2011I just got back from taking my Airdata project on its first practical test. Surprisingly, it worked almost exactly like it should. First, I turned down a piece of very thick-walled tubing I had in my metal bin (which had a 1.95mm hole down the center, conveniently close to the 2mm nominal ID of the silicone tubing I got) so I could use it as a pitot tube. This was then affixed to the windshield of my motorcycle using a very high tech mounting arrangement:
Yes, that is in fact gaff tape. It was the easiest quickie mounting solution I could come up with. Making the real pitot tube is going to be an interesting challenge. I was pleased to see that even that short tube was enough to get clear of the turbulent air around the fairing. Anyway. The ADI itself was mounted (with the use of more gaff tape) atop a tankbag, which was also used to store all the excess cable and tubing that went along for the ride. Power was via a standard Powerlet connector I installed on the bike years ago, which ran to a standard cigarette lighter outlet, into which was put an el-cheapo USB power converter. Then I used a standard USB cable to the ADI to provide it power. This is far more convoluted than the eventual proper installation will be, but it worked. The pitot tube pressure was routed via silicone tubing to the ADI's ram-air pressure input (all ten feet of it -- why cut it if I don't need to, right?), and that was all that needed to be connected. On top of all that, I strapped the GPS to the handlebars, so I would have an accurate idea of what my groundspeed was during the test.
When I pulled out, my first test was to hit about 15 MPH, and hold steady there. The GPS and the ADI agreed on my speed. One reason I was excited to get out for testing tonight was that it was a still night, with no noticeable wind blowing. I increased my speed to 20, then 30 MPH, and still, the GPS and the ADI were pretty close. I rolled down to Aurora, where I was able to hit slightly higher speeds, and it seemed pretty accurate up a bit above 40 MPH. Then I made my way to the freeway, where I could safely hit about 70 or 75, to test the higher-end speed accuracy. I was a bit surprised to learn that on the freeway, it was less accurate. I thought at first I had a tailwind, but when I turned around and came the opposite direction, I saw the exact same effect: about 10 MPH discrepancy between ADI and GPS speeds. I had other theories: the bubble of still air behind the windscreen could shift down at higher speeds, so the static-air port was getting some pressure (which would reduce differential pressure, making it read slow); other cars' movement could be causing the air to move along with the mass of cars. I couldn't shield the static-air port effectively while riding, so it was hard to tell what exactly was going on.
Then I got back onto deserted residential roads, and did a few brief super-legal spurts of speed, where I saw that the ADI and GPS were tracking each other perfectly. So, it was probably all the other cars on the freeway that was causing the error. I've always wondered how much air movement all those cars caused. Now I know: about 10 MPH. Interestingly, the altimeter and vertical speed indicator worked quite well, too. I had earlier come to the conclusion that vertical speed was probably a worthless piece of information, but it actually looked like it was providing useful data as I descended a hill, going down to -300 or so FPM. I still want a clock on there, but I'll have to think about how to have both. So, I'm very excited to see my first practical test of the ADI. It passed, honestly, with flying colors. Now to begin the interesting task of integrating a real-time clock module, and figuring out how it's all going to be put together in the final version. Posted at 22:50 permanent link category: /gadgets Wed, 13 Jan 2010I work with a small theater in Seattle, Annex Theatre. Recently, we had an event where we needed to play a DVD. We have a really snazzy projector up, all HD and everything, and we were feeding it with a $60 flip-open travel DVD player. When this cranky piece of crap player refused to go beyond the first chapter on the DVD, I'd had enough. I talked to our managing director, and figured out how much we could spend, and started looking at options. I quickly settled on the Pioneer DVD-V5000 and DVD-V8000 models as being likely, but then, buried deep in the search results at B&H, I found the HHB UDP-89. This player looked like the answer to all my desires in a DVD player. It handles pretty much any format on a CD-sized disc with the notable exception of Blu-ray. It has highly reputable audio hardware, including balanced stereo out (which doesn't appear on any other DVD player I could find). It has the necessary front-panel buttons so it can be operated without a remote. It's only 1U high, which means it will take up less space in the equipment rack. Of course, it's also a $1000 player, but it looked like a good investment in the future, and would match our fancy projector well. It finally arrived yesterday, and I spent a few minutes plugging it in and getting it temporarily situated in the equipment stack (the rack is still a bit of a pipe dream, at least for the stuff we have to access frequently). When I pulled it out of the box, I had a very favorable first impression. This is a dense unit, and I suspect that if I open it up, I'm not going to find that density supplemented by weights (unlike some cheap consumer equipment I've seen). The front panel is very plainly a thick, milled piece of aluminum, very attractive and speaking favorably of the quality inside. Once I got it powered on and had spent a few minutes playing with it, though, that favorable impression evaporated somewhat. The buttons feel quite cheap and plasticky, although they emit a loud click when pressed, so at least there's no question of "Did I press that button or not?" The power-up process takes a shockingly long time, something like 10-20 seconds before it responds to any control inputs. It's not a big deal in the long run, but it's pretty surprising for something that's a new design. You have to wait for it to boot up before it will even stick out its tongue to accept a disc. The disc tray itself is thin and springy, which at first seems cheap, but then I realized that its springiness was its salvation -- it could be thin without being brittle and breakable. Once I got a disc in (a CD, as there was a rehearsal going on that I didn't want to disrupt with a video projection), it quickly read it and was ready to go. I was pleased to see that the auto pause function was given its own, clearly labeled button. This will make it the favored CD player for production use, almost certainly. I quickly realized why the auto pause button was so prominent: when you pause the playback, then skip to the next track, it will automatically start playing again, if auto pause is off! This is counter to every CD and MD player I've used in the last decade, and is going to screw up every first time operator of this player. The information display is pretty minimal, a small two-line dot-matrix display with a few dedicated icons (CD, auto pause, play triangle and pause bars were the ones I noticed, though I'm sure there are others). There is no time button on the front panel, most likely in the interests of clarity, although I'd expect to find one on a pro CD or MD player. Still, a time button (for switching between at least elapsed track time and remaining track time) would be a welcome addition, even if it is a bit useless for a DVD player. I suspect the function is available via the remote control, although I didn't check. Fortunately, the player is quicker to respond once it's playing a disc, as compared to start-up, but it's still a bit sluggish. I pressed a track skip button and it took a second to catch up -- a potentially fatal flaw when you need to skip several tracks quickly, particularly in a tense production moment. You'd better be counting button presses, because the display lags. I suspect Sony avoids this by making sure that the display is always quick to respond, even if the underlying system and mechanism is taking a moment to catch up. Granted, this first impression doesn't actually say anything about the UDP-89 as a DVD player, but I was interested to see my own reaction to it. As there's little out there in terms of user review, I figured I'd share my reactions. I expect I'll have more to say later. Just to ensure that I'm not leaving you with the wrong impression, I was overall quite favorably impressed. The build quality, with the exception of the plastic front panel buttons, seems to be top notch. The slow UI is something that could easily be corrected with a firmware update, which is a procedure clearly outlined in the manual, although no updates are currently available. I'm looking forward to powering up the projector and putting the UDP-89 through its paces with a couple of different DVDs. If I can scrounge together enough cables, it'll be fun to hook up to the theater's 5.1 system and get real surround sound going. I'll try to come back to this topic in the future, and keep you updated on our snazzy new DVD player. Posted at 10:45 permanent link category: /gadgets Mon, 23 Mar 2009I just received my new toy: a Canon G10. It's either a porky compact camera, or a slim SLR replacement, depending on who you choose to believe. Reviews are copious and mixed, although generally very positive. The stupidest thing? 14.7 megapixels. Yes, this camera has almost 50% more resolution than my SLR does. That's just goofy. Anyway, of course the first thing I did was start snapping random pictures, and I thought I'd post one here in all its terrifying glory:  Click for the Ludicrous Size (3.2 MB) image My immediate take on it, as a replacement for my venerable Canon S410, is that it's nowhere near the same class, and so it's hard to compare. It's much bigger and heavier, but it also has clever things like an image-stabilized lens, and, of course, more than 3x the resolution. But the main thing, the aspect that really nailed it for me, is that it includes full manual control. If I want to shoot at 1/200th of a second, and only 1/200th, I can. If I need to keep the lens wide open at f2.8, I can. It's an excellent replacement in almost every way. Pity it's so (relatively) big, though. I will almost certainly have refined opinions after I get some time to play with it. Oh, but I already have my first gripe: the shutter release has a very mushy feel about it, making the half-press point hard to find. The S410 beats it hands-down on that count, which is pretty weird considering how much higher-spec the G10 is supposed to be. Posted at 11:52 permanent link category: /gadgets Mon, 18 Dec 2006
You wanna know who killed the electric car?
Electric car makers did. I just did some cruising on the net, and confirmed a few suspicions on the ZAP! Xebra:
...And so on. sigh. Most of this is from xebraworld.com, which is nominally an enthusiast's site. Now that I reflect on it from a little tiny bit of distance, it doesn't make much sense for me to drop $12k (or more, I haven't heard back from the dealership on what "other" charges like shipping, paperwork, dealer prep, etc. may be tacked on) on what amounts to a glorified golf cart. Granted, it's electric. Cool. But with a regulated top speed of around 32 MPH, it's not safe for the missions I have in mind, which require at least 35, and preferrably 40 or 50 MPH. Having spent some time figuring out what exactly they're selling for their twelve thousand dollars, it's not really a winning proposition, at least for me. Too many problems, with a too-basic system, for too much money. I keep thinking to myself, "I could build something better for cheaper, and I'd have fun doing it." Not precisely true, in that I don't yet know how to weld, but given a year or so, I'd probably come up with a better vehicle for less total expenditure. Each owner's site I've looked at so far has included the same third-world sounding description of squeaks and rattles (which sometimes correspond to major components trying to fall off the vehicle), and lots of little niggles which could have been prevented by spending another dollar or two in construction, or designing something better. Not encouraging. The troubleshooting guide, which I flipped through, suggests some problems so fundamental as to make me question the assemblers' ability to do anything right. Did the parts get assembled in the right order? Did the welder bother cleaning anything, or just kind of weld over whatever corrosion was on the frame parts? What quality of components are in critical locations, like interconnect cables and cutoff switches? So, overall, I don't think I'm willing to consider such a questionable expense. There are too many open questions about quality, and it's apparent that marketing at ZAP! is not too concerned about truth (which is the point of marketing, but they're usually closer to reality than this). All of which is a pity, of course. I had high hopes for the Xebra, but what I know now suggests that they're charging too much for too little. Posted at 16:32 permanent link category: /gadgets Sat, 16 Dec 2006I've got the review of the Xebra PK in place now: Enjoy, and have fun looking at the pretty pictures. In particular, I'm fond of the picture with me standing next to the thing. It's miniscule. Posted at 18:02 permanent link category: /gadgets I went in to the Green Car Company today, and ended up test-driving three different cars. The first was the ZAP! Xebra PK, my primary motivation for going there. First impression? Tiny! That truck, despite what it looks like in every picture I've come across, only comes up to my chest. It's miniscule. That tininess extends to the cabin. I could just fit in, with the seat rammed all the way back, and even then my shoulder and hair were brushing the walls. The pedals are weirdly offset, with the gas pedal about were you'd expect a brake pedal to be in relation to the seat, and the big brake pedal shoved hard to the left wall. It definitely encourages "two-foot" driving despite the automatic nature of the beast. In a quick (~1 mile) test drive, I got to accelerate on the flat, round a few corners, and try going up a moderate grade. Acceleration on the flat is fine, a bit less than what you'd expect from a truly basic economy car. Corners were non-events, apparently due to the positioning of the batteries below the truck bed. No noticeable body roll, but I wasn't pushing anything. Acceleration up the grade was pretty lame. It hit an indicated 18 MPH, and stayed there until the hill flattened out. This was with myself (about 220 lbs) and the salesman (maybe 180-220 lbs) in the cab, and nothing in the bed. Part of our circuit took us over some speed bumps, and the steepest bump there caused some hard part at the front of the vehicle to touch down with a bonk. I would guess it was the lower suspension mount, but I didn't take a close look under the nose of the thing. The sedan model did the same thing (even being very gentle over the bump), so I assume it's typical of the design rather than particular to that unit. The salesman, who I can now only think of as Smilin' Jim (his name is Jim something, but he was definitely a salesman, and not a granola-eatin' hippie) said the Xebras currently come with 300A controllers, but that they were looking into replacing them with 450A controllers. That would definitely help with acceleration, and with getting up hills, but it could seriously impact range if you've got a heavy foot. He said that as equipped, the Xebras don't have regenerative braking, but that with a new controller, they could have regen. I don't know if this is true, and he admitted to me that he wasn't super savvy about the tech side of things. It'd be worth more investigation. He also said that the top speed (40 MPH, more or less) was controlled by a governor, but said that he didn't know where the governor was. If it's not in the motor controller, I'd be very surprised. It probably doesn't make a lot of sense to get rid of the governor, though, as going faster would seriously compromise range. The Xebra sedan was about the same experience as the truck, but the seating was more comfortable. Since there wasn't a vertical wall right behind the driver's seat, it could be scooted much further back, and there wasn't a curved-in section to hit my shoulder. If I were looking at one of these as a single-person transport, I'd pick the sedan (it's also cheaper). The biggest problem I had with the sedan was that the parking brake was situated on the left, next to the door, and was nearly impossible to reach. Smilin' Jim looked over at me at one point, and said (shortly after I'd described my sidecar rig to him), "You know, one downside of this is that you'll have to get a motorcycle license." I looked at him like he was crazy, since he'd commented on the fact that I rode a motorcycle in to the shop when I walked in, and we had just been discussing the way I don't have any cars. He had to understand that I already had a license. "I have a motorcycle license," I said. "But I mean a class 3 motorcycle license," he said knowingly. He's plainly caught motorcycle people out with this before. I just looked at him like he was crazy again, and said, "I have a three-wheeler license, if that's what you mean." He had a little head-slap moment and said, "Oh, of course! The sidecar..." Ah, Jim. Finally, I drove a SMART car, which was a blast. Very much more refined than the Xebras, I could easily see getting a SMART as an all-round car, if I were rich and very specifically wanted a highly-efficient two-seater. At $25k for the basic model, and $30k for the convertible, they're beyond what I want. Way cooler than a Toyota Echo, but of course nearly double the price. The SMART accelerates very well, and Smilin' Jim said they get 40-50 MPG in-city, and 60ish on the freeway. That's pretty damned good, for a boxy little car with a 700cc motor. (It doesn't sound big, but my 250cc motorcycle has trouble topping 50 MPG in-city, and gets 60ish in real-world freeway driving.) The SMART also has ABS and airbags, so it's actually got a lot going for it when compared to a motorcycle. Of course, it's not actually a motorcycle, so it's a bit of a ridiculous comparison. Anyway, after I was all done, I asked if I could test-drive the Meyers whatever-they-call it (nee Corbin Sparrow), and Smilin' Jim lost a bit of the smile (but not too much) and said, "I'm sorry, but it's not licensed for single-driver right now." I gave him the "huh?" look, and he explained, "Because I can't go with you," as if that cleared it all up. "Oh, ok," I said, since the answer was obviously "No. Go 'way kid, you bother me," no matter how I might have tried phrasing it. (I noticed on the way out that the Xebra PK we'd taken out had no license plate whatsoever, so Smilin' Jim obviously wasn't too worried about legality.) No worries, he'd been very accomodating. I know I'm not going to get a Corbin Sparrow, although they're very cool, and I'm glad to see that another company picked them up. So, bottom line, I liked the Zap! cars. I'm going to write up a more formal review which will appear on my main page sooner or later, with pictures. I'm only going to review the truck model, with perhaps a smattering of comparisons with the sedan model. Posted at 15:30 permanent link category: /gadgets Mon, 11 Dec 2006Ok, this is just cool: I'm going to have to head over to the Green Car Company soon and look at one. That's a truck I could get into. Posted at 17:13 permanent link category: /gadgets Tue, 31 Oct 2006So, earlier this year, I bought myself a Grundig FR-200 to keep at work. It's a radio which can be powered by a hand crank, making it an excellent emergency-preparedness item. I actually ended up cranking the thing for about 2 minutes almost every day, and listening to the news in the afternoon. Yesterday, I went to crank it up again, and to my surprise, it almost immediately went POP, and cranking was suddenly interrupted every 1/5th revolution by a little grunty feeling. Clearly I had popped a tooth off one of the dynamo gears. Oh well, I thought to myself, that was fun while it lasted. Then, today, I remembered that this had happened, and did a little bit of searching around, to see if I could locate a parts list, and maybe figure out what the specs were on the gears. I thought perhaps I could find a replacement, since nylon gears tend to be relatively standardized. I came across the manual for the radio, and there, on page 7, I was reminded that the radio has a 1 year warranty! Ah-hah! I called the listed support number, and got the grumpiest-sounding guy I've encountered in a long time on the phone. "Uh-huh, this happens from time to time," he grumbled. "Try not to crank it for ten minutes." He took my name and address, and said he'd be sending me a new set of crank gears. Woot! That was surprisingly easy! So, thumbs up to Eton (the parent company of Grundig) for standing behind their warranty based only on a punter's word as to purchase date -- no grilling for receipts, or warranty cards, or anything. Most excellent. Now, of course, I have to figure out how to replace the gear cluster. One problem at a time. Posted at 16:05 permanent link category: /gadgets Categories: all aviation gadgets misc motorcycle theater Written by Ian Johnston. Software is Blosxom. Questions? Please mail me at reaper at obairlann dot net. |
