Newbie

[Jonster]

Hello to all you modders ... :welcome: ... I've been gleaning info from here for a while and thought I should sign up.

My mates all have super-bright Lupine or Hope bike lights and I need to out-do them!:laughing:
Yes - I'm being driven by a desire to go "one up" and No, I'm not ashamed.:thumbsup:

Basically I have a good heatsink and am thinking of getting the following from DX or KD ...
x2 MC-E M bins
x1 5-mode driver
x6 18650 batteries
x1 charger
x2 optics

For a newbie is there any essential reading in terms of how I go about wiring and any good starting tips!!

Cheers dudes,
Jon

 

[DM51]

Welcome to CPF, Jonster.

As your question is about a new build project, I'll move your thread to the Homemade & Modified section.

 

[J_C]

You don't mention what level of electronic circuit experience or knowledge you have, or if you have a multimeter (highly recommended).

Overall the process is pretty straightforward. Harder part is deciding on a battery holder(s), casing and/or mounting system for the bike, the type of connectors and switches, waterproofing.

The connection itself is depending on how you set it up, what driver you use. Presumably with one driver you'd wire the LEDS in series, and with 6 cells in your battery pack you'd want at least a buck type regulating driver board (buck as-in it drops a higher battery voltage to a lower summed LED forward voltage).

Making it rugged enough is the key, simply soldering battery output to driver input and driver output to LED input is straightforward, though I suggest you make temporary connections on a bench and complete the circuit by measuring current with a multimeter so you can see if the battery drain rate and LED drive current is what you expect it should be, what you aimed for or at least in a safe range.

Otherwise, despite which LED you use there are some other DIY bike light threads on CPF, your process is not so different than those, so those topics should be helpful.

To really 'out-do' a professionally made light, half the work is making a housing for it that is attractive enough to be considered eye-candy. We can't help much there, depends on what you want, your budget, machine shop type skills, etc.

 

[Jonster]

Hi J_C

Yes I do have a multimeter but my level of electronics experience is relatively low. I am however a mechanical engineer so pick things up relatively quickly.

Regarding a housing I picked up a heat-sink which I believe is from a machine tool capacitor. I plan to halve it and put an LED in each, mounted on a turned aluminium backplate. I may try and get it anodised first. Here's a pic ...

The stuff I'm more jittery about is ordering the right kit making sure it works when I get it.
The P7 vs. MC-E is a biggy but I think the MC-E sounds more reliable??
Also i'm not sure whether ordering from Kaidomain or Dealextreme is the right thing for a first-timer as it sounds like people have to mod their boards sometimes?

Anyhow its going to be a fun experience!

Jon

 

[DM51]

Could you resize that pic, please? It is 1024 x 768 pixels, and the maximum allowed under CPF Rule 3 is 800 x 800 pixels. Photobucket has the facility to do that for you when you upload there.

BTW, that looks like one of Mrs. DM51's hair-rollers. In her case, they don't seem to be very efficient as heat-sinks, lol

 

[Jonster]

Hi DM51,
Pic cropped and shrunk ... cheers for pointing that out.

If I cut some holes in my helmet and grow my hair a little the "hair rollers" approach might make a nice ultra-light way of mounting the lights!:laughing:

Cheers,

 

[yellow]

the main question is:
how will Your battery pack look like?
From there everything else is more-or-less plug and play.

When You read a bit here on CPF (and in modified, Bike light and led-lights especially) You will learn enough to to build a good, reliable light.

Why not start with a single quad-led light? 2 quads is way overbuilt, imho.

PS: for bike use, I think a reflector is much better (but I think a refelctor is better with most every use, so ...)

PPS: that heat sink looks cool!
Might be difficult to machine, but really cool.

 

[Jonster]

Cheers for the nudges ... I've found some great threads on the bicycle section.
I am steering towards a hipflex driver which looks like a good reliable starter. The KD and DX stuff looks just too unreliable from what I've picked up.

The reason for having two lights is that it allows me to spread the light on the trail to give a good general beam in front and pick out a tighter spot up the trail.
If going for just one then I'd look to mount it on my helmet - which isn't a bad shout for the future.

The battery pack will probably be a sexy acrylic affair cut to fit in the front triangle near the headset. Otherwise an old water bottle may do as a starter. Not really put a lot of thought into this.

Most of the machining will be from bar and the heat sink will then be lopped in half to shroud this.

I've got all these thoughts going round my brain but basically need to just start finalising the electronics and ordering bits!

 

[J_C]

I'd start with how big, expensive, and heavy you want the battery pack to be, then proceed from there deciding how powerful and how many LEDs and housings you can use and still have enough runtime.

Great heatsink, a good start at a cool looking enclosure and a great shape for getting airflow through the fins on a bike.

Keep in mind that even with a tight beam, say about 7 degrees lens, there will be enough sidespill to illuminate a wider area sufficiently closer to you. I don't think you really need two different beam patterns, if you opt for two lights I would make both tight beam and expect plenty of sidespill, but personally I'd rather one on the bike and one on a helmit.

KD and DX drivers can be a lottery, but their comments section helps a lot to know *if* you're getting what was described and what, if any, modifications might be helpful. Their drivers are so inexpensive it's not a big loss except for the time waiting to receive the parts and time dealing with any potential issues.

If you have plenty of spare time and want to save a buck it may be worth the gamble, but since this is going to potentially be a really nice light I can see the justification for getting a theoretically better driver... but first thing first, nail down the battery configuration you want.

 

[Jonster]

Cheers J_C,

Regrading the battery config if my maths is correct I will need upto 3.4V per MC-E plus around 0.5V for a driver so (3.4V x 2qty) + 0.5V = 7.4V
If I pop two 3.7V 18650 batteries in series I've got my 7.4V but no margin for error - although would i need some headroom for power drop off as it discharges?
I was then going to use three pairs of 2400mAh to give myself hopefully around 6hours run time at 2.8A (assuming all nominal run temps etc).

The qty6 18650 batteries would fit nicely in 1 quad and 1 pair holder which could be mounted in a fabric pouch in the front triangle of the bike. I may even put the driver in with it?

Looking at the driver I've decided to ditch DX as they don't do an M bin MC-E and go with KD. For $8 they do a 5-mode board (sku S007542) which is probably worth a shot and if it doesn't work I've not lost much. Only thing a little scary is that its rated at 3A when I only want 2.8A and don't want to fry my LED's! Also it doesn't state whether you can run a pair of MC-E's or whether its one board per.

I'll continue to ponder and research - but at some point I will press the button and get an order on!!!

Cheers,
Jon

 

[J_C]

Depends on what driver you can find, whether it's just buck (dropping voltage) or boost-buck which would allow running the LED at full power still after the battery voltage drops below that of the LED forward voltage.

Considering the light, I agree 2 or 3 parallel series of 2 (4 or 6 total) is better, IMO 3+ hours is the least runtime a nice light ought to have but opinions vary on that.

Since the driver is being matched to the light config, I'd sooner put the driver in the light housing unless you think the space for it would be a problem, or if it used electrolytic capacitors that you didn't want subjected to as much heat as there might be in a sealed (at least waterproof) light housing, though better drivers don't use electrolytic caps and it's one of the things that could be modified on a lesser driver by swapping in polymer, tantalum, ceramic, etc caps. It should work fine in the battery pack instead if you want it there, and unlike some light configurations you aren't making a significant difference in current over the wire and connectors like someone would if they had a low voltage higher current battery pack boosted later, though any wire thick enough to be mechanically durable as it flexes when you turn the wheel, should be more than enough for the current needed but on that point I advise using high flexibility, high count/small strand wire for longest life. Then again I don't recall reading about anyone having a problem with their wire but I wasn't looking for reports about that so...

2.8A vs 3A isn't much of a difference, keeping in mind many people drive single die ~ 3W LEDs above 700mA in flashlights with far less effective heatsinking than 1/4th of what your housing will have the potential to provide, and riding the bike you have the airflow to help but ultimately you'll have to build it and see how hot it runs. That's the thing with prototypes, you make the best educated guess and adjust if it doesn't work out as well as planned.

Assuming your heatsinking is good (I don't know the dimensions of the 'sink pic you posted), the small increase in current could be seen as a desirable improvement in output, though with some cheaper drivers you end up with an output current that while supposedly regulated, varies a bit with the input voltage.

I've no experience with that driver so I can only suggest seeking some info on it from someone who owns one.