Dyno powered double-triple xpe - options

[Humph]

Just pulled the trigger on a Shimano 3N72 dynamo hub for my commuting bike ('07 Giant OCR2) and am trying to think of a good way to wire up the lights internals. I know for commuting that a triple would do the trick running off Martin's circuit #10, but there are some steep, fast and poorly lit downhill sections that could do with a bit more light [or maybe I'm just greedy].

What I was thinking was for most of the commute running a single-triple xpe through a Carlco 20mm medium and then when needed, flick on the other triple through a narrow lens that has been shimmed up a few degrees to light up a bit further down the road [or blind oncoming traffic driving down the cycle path]. Switching would be via 2 switches; single-double and high-off-low.

The issue is I was going to stuff these into a little Hammond enclosure (either 1455 or 1457) and keep it as small as possible. So I want to keep the electronics to a minimum. Problem is my knowledge in this area is limited (but growing), and looking at Martin's diagrams I could probably build them, but modifying them is a different story. At this stage I could build 2 separate circuits, but I'm sure that is not the optimum solution and wouldn't fit in the enclosure I'm looking at. And if I just have a switch that turns on 3 or 6 LEDs, I think there will need to be some tinkering with the caps [and it is at this point my brain explodes].

So I'm asking the brains trust: is there a more elegant solution than building 2 separate circuits to switch between? Is my idea workable or an I dreaming?

 

[yellow]

dreaming!

The idea is good, but effort is too high

Just do the triple-setup.
have all of them be offset (from main axis) at a few degrees, decide how You want Your triangle beam set
(I have the horizontal part up and the single led lighting down at short distance.)


For the offroad/power uses, mount an additional light to the bars.
F.e. my triple XR-E (P3) w. 8 deg. optics @ ~550 mA from the dynamo seems to be switched OFF, when I turn the single XR-E (P4) w. 19 mm SS reflector Flupic @ ~850-900 mA ON
(thats the reason why I always wonder when someone plans to use Quad-emitters at less than 1A and still argues it will be noticeably brighter than a single at the same current)

or get a quad-emitter light for even more output

 

[znomit]

:welcome:

A triple is likely fine for what you want.
Beam brightness, width, throw are all personal choices so its hard to say from here what you will like.
The 20mm XP-E from carlco is actually a reasonably wide beam at around 16deg FWHM so does ok even through twisty stuff, though it doesn't throw a mile down the road like a tight 8deg beam. Its a nice light to ride with. Yes a single 8 deg at 1A will have a brighter spot that a quad 16deg at 500mA, but you wont see the rabbit at the side of the road thats about to run out in front of you.

If you are going to go with two circuits I'd do a double and a quad, both just using smoothing caps. Quad for most light, double for general and low speed use. I would just use smoothing caps on full wave circuits because the boost caps take up a lot of room.

If your hign speed sections are straight, an aspheric on an MCE may be the way to go, very well focused, no spill though. This guy does well, beamshots over at MTBR.

 

[Humph]

Thanks for the replies guys. After a bit more reading I think I will stick with the triple, at least to start with and add to it as I see fit.

I was thinking of using the XPEs but need a bit of clarification. I thought that the dynamo had a limit of 500mA but the voltage changes with speed. But have seen figures floating around where the current varies. The thing is that the XPEs have a limit of 700mA and don't want to pop one or two going down a fast hill.

Having seen the beamshots for Troutie's helmet light I think the 10mm square optics would be OK.

 

[unterhausen]

the current will vary with voltage. There isn't much voltage change across the leds, so that determines the voltage more than the dyno. My understanding is that the dyno is more of a current source, the voltage is determined by the load.

 

[jbassett]

<SNIP> I thought that the dynamo had a limit of 500mA but the voltage changes with speed. But have seen figures floating around where the current varies. The thing is that the XPEs have a limit of 700mA and don't want to pop one or two going down a fast hill.

I recently did some testing with a dynohub. The voltage varied hugely and we managed to reach about 20V when the load was fairly high resistance.

The maximum current we got was 0.53A at moderate speed and wouldn't increse regardless of the load applied and speed of the wheel.

The current seems to be limited by the dynohub design. It has something to do with the dynohub becoming saturated due to the magnetic flux available and is related to the amount of ferrous material around the windings.

If your dynohub is rated as 0.5A it is unlikely to get significantly above that.

 

[Humph]

If your dynohub is rated as 0.5A it is unlikely to get significantly above that.

Thanks,
I found a similar answer elsewhere on the forum just after I posted but couldn't edit it because it was in pergatory and had to be blessed by the mods for its release.

 

[Humph]

Just another couple of questions. Why do people use schottky diodes instead of an all in one bridge rectifier as it has a Vf of ~1.1V. Are the Vf savings worth it or is there another reason?

Also, with capacitors, I can understand using one to smooth out the Amps so you don't get low speed flicker, but is there a real world advantage in using the boost caps per Martin's circuits for multi-LEDs?

 

[jbassett]

Just another couple of questions. Why do people use schottky diodes instead of an all in one bridge rectifier as it has a Vf of ~1.1V. Are the Vf savings worth it or is there another reason?

Also, with capacitors, I can understand using one to smooth out the Amps so you don't get low speed flicker, but is there a real world advantage in using the boost caps per Martin's circuits for multi-LEDs?

The reduced Vf of Schottky diodes are definitely worth the effort at low speeds as the hub voltage is low and every little saving helps.

Using a boost circuit is a matter of personal choice. If you have a boost circuit fitted you will get approx. double the voltage which will get a longer string of LEDs glowing at lower speeds, but you'll only have half the current available so the maximum brightness is much reduced. I think the best option is a boost circuit that can be switched on and off (such as #8 - #10) is the best solution. These offer the advantages of higher voltage to get the LEDs glowing at slow speed when boosted, but allow the ful current to be available when non-boosted at higher speeds (i.e. when the hub voltages is high enough anyway). Circuit #12 is a great idea, but a lot more complex to build and a fair bit more expensive.