Heads up – 7up!

Finally got to replace the burned off driver in one of my 7up lights. Made a good looking wire harness with a small shrink wrapped push-button and stuck the whole thing to my helmet.

2013-11-18 23.33.31

No more gaffa tape to hold the modified MR16 bulb from DX like I have been used to! It looked terrible anyway….

MR16 bulb

I used one of these big fat silicone rubber bands to fix the light on top of the helmet. It provided a good tight fit but I don’t hope the rather sharp edges on the handlebar mounting bracket will prove too much for the budget rubber band…

They’re around $2 on DX so buy a whole bunch while you’re at it anyway.

Tomorrow morning we’ll see how well it performs in general traffic. I’ll probably blind the occasional driver.

2013-11-18 23.34.49Only one thing to do now – clean up all the mess in the kitchen and get straight to bed!

See ya!

Final assembly of multiple 7’up lights

One long night of cutting, glueing, soldering and – unfortunately – burning off expensive electronic hardware resulted in two finished lights.

One with an ordinary panel mount 12mm push button and the other with a micro switch hidden in a small length of shrink wrap. The latter will probably be the easiest to fix to the handlebars.

It doesn’t take long to create quite a mess on the dining table and I’m astounded every time by the time consumed to do even these trivial little tasks.

The work table full of bits

There are two more lights in the pipeline. One has the driver thermally bonded in place whereas the other – well that one needs a new driver due to last night’s unfortunate mishap.

Game on!

Be careful when handling poorly isolated wires…..

Fuck fuck fuck!!! Tonight when I was fixing the 7’up light to my handlebar one of the exposed soldering pads on the control switch just touched the positive connector coming from the battery.

A brief flash of light and that was that. Say good bye to at least $50 of hardware, hours of glueing, soldering and assembly. Now I have to order a new MaxFlex from Taskled. That will be at least a couple of weeks of waiting – provided that they are in stock and Danish Customs don’t snatch them at the airport…..

Note to self: Be very careful when using heat shrink. Be sure to cover EVERY FUCKING MILLIMETRE of exposed copper wire!!!

Lesson learned the hard, stupid and expensive way 🙁

LED PCB modifications

The design of the 7’UP lamp necessitates routing the wires from the driver to the LED’s through holes in the bulkhead and the LED PCB. By using the bulkhead as a template the holes were marked with a felt tip pen. I did a Delrin fixture to hold the LED PCB in place while drilling. This also protected the LED domes, which are made from silicone and are quite fragile.
2013-06-14 17.13.49

Holes marked with blue felt tip pen

2013-06-14 17.14.22

Drilling away from the back side

2013-06-14 17.14.37

Removing burrs by finalising the drilling from the front.

2013-06-14 17.15.19

The finished LED PCB ready to mount on the bulkhead and driver assembly

New custom components ready

I’ve been playing around with the lathe at the job and came up with these little babies:
2013-06-10 21.43.52

The tool necessary for holding the work pieces in place was made from a length of Delrin and two 33x3mm o-rings. It held the acryl pieces in place like a charm:
2013-06-10 18.49.28
2013-06-10 18.48.46

I just did a whole stack of lenses so there’s enough for everybody involved.
2013-06-10 21.36.40

Hvad sker der i hall sensoren?

Den mĂžrke tid er kommet med overraskende hastighed. Det tid til at komme i gang igen efter en lang sommerpause.

Sidst havde jeg en masse uforklarlige hendelser nĂ„r jeg prĂžvede at programmere vores LED driver. FĂžrste troede jeg der var lĂžse forbindelser og dĂ„rlig lodninger. For at analyser percist hvad der skete nĂ„r der blev trykket pĂ„ hall sensoren, har haft et ocyloskob koplet til hall sensoren. Med oscilloskopet har jeg sĂ„ kunne mĂ„le om hall sensoren trak signalet “lav” nĂ„r magneten blev trykket ned.

Programmering af setting forgÄr ved en kombination af korte klik (<0,3s) og lange tryk (>0,3s). Jeg har haft en bekymring at bevÊgelse at vores tryk knap tog lÊnger tid end 0,3s og det der skulle vÊre korte klik blev til lange tryk og det var det der drillede nÄr jeg prÞve at programmer driveren

I videoen giver jeg fÞrst 10 kort klik. De er alle ca. 0,2s, minder end de 0,3s, sÄ det er ikke korte klik der bliver til lang tryk.

For at komme ind i settings skal man give 3 korte klik og 1 langt tryk. De 3 klik gĂ„r fint, men i det lange tryk kommer der to afbrydelser hvor hall sensor signalet gĂ„r “hĂžjt”. Det skal det ikke! Afbrydelsen svar til at der kommer et ekstra klik hvilket betyder at driveren gĂ„r i programmerings menuen eller helt ud af settings menuen.

Er der nogen, med elektronik kendskab, der kan forklare hvorfor den gĂžr som den gĂžr?

Evil eyes – a CAD rendering of a twin P60 drop-in lamp

Using a milling machine with a 4, 6, 16, 24 and a 26,5 mm mill this design is a possibility.

It is made from two pieces of aluminium each 60 by 35 mm, two P60 drop-ins from Kaidomain, lenses and o-rings from Dealextreme and a 12 mm switch. The XM-L LEDs on 20 mm stars are mounted directly on the finned rear section, thus giving the best heat transfer possible. The main body could be finned also.

O-rings between housing and lenses provide sealing at the front. Between the two parts of the housing I’ll use liquid gasket.

Below is an animated assembly.

Lamp-explode

The drop-ins are $9,21 a piece including LED and LDO regulator. The lenses and o-rings are less than $1. Get your hands on some scrap aluminium, and you’re looking at a pretty cheap 1800 lumen lamp. For added power – just scale this thing to have more drop-ins….

Have fun!

What’s driving us crazy => Koldlodning

Snorre, Hans og jeg har sidet her til aften og nĂžrdet med 4’er lygten. Vi fik taget en video af lygten og dens mystiske respons pĂ„ vores clicks (korte tryk) og press (lange tryk), med henblik pĂ„ at sende den til Taskled og prĂžve at fĂ„ en forklaring derfra.


Efter at have fumlet rundt i nogle timer, fandt vi en kold lodning pĂ„ ledningen til Vcc. Irriterende at en sĂ„dan lille fejl kan holde os hen i sĂ„ lang tid 🙁 Derudover drillede det at vandringen pĂ„ magneten nok ikke er helt lang nok i forhold til positions tolerancen. Det betyder, at feltet ikke med sikkerhed skiftede “nok” til at Hall elementet kunne registrere det. Noget vi mĂ„ have eksperimenteret mere med og fixet pĂ„ nĂŠste rul 🙂

Driving me crazy

Peter og jeg har rodet med at fĂ„ programmeret driveren…og det volder en del problemer.

Manualen til driveren ligger her:

http://openproduct.dk/wp-content/uploads/2010/11/maxflexuni_v4.00.pdf

og af den fremgĂ„r det hvordan man kommer ‘ind’ i programmeringsmode og hvordan man efterfĂžlgende kan sĂŠtte forskellige indstillinger, sĂ„som maks. strĂžmstyrke etc.

Vi fik samlet en lampe for ca. 14 dage siden og brugte en aften pÄ at prÞve at programmere den. Vi kunne godt komme ind i programmeringsmode, men efterfÞlgende kunne vi ikke fÄ manuelens anvisninger og lampens feedback til at passe sammen. Det var som om vi enten blev smidt ud af programmeringsmode (en masse blink kort efter hinanden) eller ogsÄ for vi vild i de mange menupunkter.

Vi havde en teori om at det kunne vĂŠre hall sensoren der sendte uklare signaler (vi brugte en magnet foran hallsensoren til at trigge driveren). Derfor prĂžvede vi i sidste uge to ting – 1. samle en ny lampe for at udelukke at det kunne vĂŠre en fejl i lampen og 2. bruge en trykknap for at udelukke at det kunne vĂŠre hallsensoren der drillede. For begge lĂžsningers vedkommende har resultatet dog vĂŠret det samme som fĂžr – altsĂ„ manglende sammenhĂŠng/forstĂ„else mellem input og output i driveren.

Umiddelbart er nÊste skridt at vi kontakter driver-leverandÞren for at fÄ nogle tip, men hvis der er andre idéer er de meget velkomne!