mountain_lion

LEDs can be connected in series or parallel, but control is much easier in series. The LEDs are driven at constant current using a driver such as the Buckpuck. One Buckpuck can drive several LEDs in series with the same constant current supplied to all. If you connect in parallel you can use a lower supply voltage (i.e. less batteries), but if you use a single driver, the current is the split between the LEDs i.e. a 1000mA driver will supply 500mA to each LED if you have 2 in parallel (assuming the 2 LEDs have identical impedance). With the Buckpuck, it is easy to control the output current (potentiometer between control and reference pin) so you can dim all the LEDs instead of switching some off.

The Lupine Betty is a very nice and super bright light, but not many of us can justify the US$995 price tag! I like DIY and have done several versions over the past 5 years. Now busy with another set, this time using LEDs.

Also see gearreview

Have a look at the MTBR LED reviews

I am considering getting the Maha 800S charger. Ebay seems to be the best source. Have not found it locally in SA at all.

Light & Motion manufacture great products ($$$), but the Arc is HID. I believe HID lights are history. Their lamps are expensive and fragile. LEDs are starting to equal and even exceed HID in terms of light output and do not have the lamp problems of HID. Rather get a decent LED light. mountain_lion2008-04-24 14:39:08

OK, I did some tests comparing light output of the 1W Luxeons at various load currents. Since I don't have a lux meter, I used my camera. I placed the light (3x 1W LEDs with Trilens Narrow Beam Lambertian collimator) on a table and pointed it at the ceiling (white). The camera was mounted on a tripod, taking spot measurements at the center of the light spot on the ceiling. I recorded load current, F-stop & lens opening with the camera set to 400 ISO and used the formula on this page to get a lux value (not accurate at all in absolute terms, but fine for relative comparison). This confirmed what I guessed: At first light output follows current almost linearly, but from approx 600mA light output starts to taper off and increase in current does not yield much increase in light. In fact from 800mA to 1000mA LED current, my camera reported the same light reading. Bottom line: There is not much benefit in running 1W Luxeons at a load current of more than 600mA. So driving 3 LEDs with a Buckpuck driver from 8 x 1.2V NiMH cells will do fine.

These 2 PDF documents from http://www.speedtraps.co.za/info-page.php might be of interest: Guidelines_regarding_speed_measurement.pdf Guidelines_regarding_laser_speed_measurement.pdf

Mux, was this a fixed speed camera or a mobile speed trap? If mobile, it should have been maned and the officer manning it should be able to produce the calibration certificate for the device as well as proof that he has been trained to use said device. Worth stopping at these if you suspect that you have been fined and checking these. For fixed installations, you can also ask for the calibration certificate if you want. I guess that often these devices are not properly calibrated. Your GPS should be more accurate, but I agree that using a GPS in court might not be easy...

I have the speed/cadence sensor mounted on my road bike because there I am interested in cadence. However I mostly use my Edge on my MTB without the sensor. Reception is very good and it is very seldom that I loose signal, even under tree cover.

WH, my observations are based on work bench tests only. I have not constructed a usable housing / mount for my light yet to allow field testing on the bike. I have to admit that by eye it is difficult to see the difference in light output produced by the 1W Luxeons at 600 mA and 900+ mA. So the point may be moot. Running at 600mA might actually be more efficient producing less heat loss, i.e more lm/W. I will try to do some comparative light measurements later. This method looks interesting if you do not have a lux meter. Getting the housing and mounting sorted out is my next priority. Fine tuning and more experimenting can be done later. Doubt if my light will look as neat as yours; I am not importing a round heat sink (at least not for now). BTW, What capacity batteries (mAH) are you using? Have you tested the run time? i.e. how long can you run before the light starts to dim? At 1A load most NiMH batteries seem to maintain their rated voltage for most of the discharge cycle and then starts to drop quickly near the end. So voltage drop should not be a major problem. See battery shootout on candlepower forums (updated recently). 1.2V x 8 = 9.6V is close to the limit for driving 3 LEDs in series. Proof of the pudding is in the eating though; If it works good enough then it is good enough!

I Googled a bit and found this site specialising in radio controlled toys http://www.rclines.co.za They stock LiPo batteries and chargers but as others have said you pay $$$ at the hobby shops! Buying a LiPo battery and charger will push the price of a home made light very close to that of the cheaper commercial offerings. Looks like my 12V sealed lead acid batteries or else NiMH AA's will have to do, unless a cheaper source of LiPo's can be found...

OK, my collection of lenses and LEDs have arrived so now I can start experimenting. I also found this and like the idea: http://www.freelights.co.uk http://www.reelight.com I might even try making something similar for my hybrid. Now what did I do with that coil wire I had a few years ago...

You can try to do a complete refresh to ensure that all is up to date on the page. Ctrl-F5 works in most browsers.

I doubt if you do. You do get BMX tubes with the "car type" value. Then you can take out the core. You Will have to drill out the value hole in the rim though (you need to enlarge the inner hole in any case). I tried doing the same DIY conversion a year ago. I could only find ones with the car valve. Perhaps this is part of the reason I could not get it to work. My rims whre just too narrow at the value to fit the thicker part of BMX tube around the valve and the tyre bead. Bead just did not want to seat at the valve. Plus I was using standard tyres which had not been sealed before. I gave up and got a Stans rim strip. This worked, but I still struggled to get it to seal at the valve.

OK, 8:30 I might just make, if it does not rain... Weather for Paarl Forecast: Sunday, 6 April 2008 Min: 13?C Max: 21?C http://www.weathersa.co.za/weathersa/images/icons/cloudrain.gif Weather: Cloudy with rain at times (60%) Wind: fresh westerly Sunrise: 06:59 Sunset: 18:33 Moonrise: 07:12 Moonset: 18:29 Moonphase: New Moon Jonkershoek is a rain magnet...

Anybody know how you can download from cycling.tv to view later off-line? Apparently Veoh player can download from cycling.tv. http://www.veoh.com/downloadFlow.html I installed it, but cannot find the epic coverage using Veoh. Also just tried to view stage 2 "live" on cycling.tv http://www.cycling.tv/channels/free2view but although Epic stage 2 is listed as "now showing" something must be wrong with their links since all I get is track cycling ?? Will try again tomorrow.

the crash video? yes. Riding a MTB, no! It is sunny in Stellenbosch, time for another ride...

BTW, to embed the youtube player in your post, just use TUBE in [ ] followed by filename and then /TUBE in [ ] where filename = 4blahxuuhxg for the crash video above. [TUBE]4blahxuuhxg[/TUBE]

I did say young & reckless... and I have one shoulder with snapped ligaments from a few years ago because I was reckless, but not young!

If only I could ride like this! Guess you have to be young and reckless... [TUBE]pTb-_8aGNaA[/TUBE] mountain_lion2008-03-24 04:10:08