seanspotatobiz

I just finished a project to convert an old carbide lamp for my Hercules Commuter (some old bike). It's illuminated by two Cree XP-G power LEDs and powered by a bottle dynamo. The dynamo also charges five NiCad cells which maintain power when stopped. Not shown is a relatively boring rear lamp (not carbide) which is also powered from the same dynamo and battery). I can't believe it works!!

Click for additional images:

Digital_Cowboy

Looks nice, how long did the conversion take?

seanspotatobiz

Thanks! I guess it took about five hours, spread over a month or so (ordering parts I hadn't thought about and waiting for them caused most of the delay). The circuit is really simple. There are more complicated circuits that might use only a single NiCad cell but mine uses five.

silmarillion

Hey nice mod Sean.

I'm a caver, and that would be a nice novelty lamp to put on my MB when I go to caving events.

I'm not to sure about which brand lamp you used. Is that an old Justright or a Greir Bros? I have a pretty nice collection of carbide lamps. I went and had a look through them, but I don't have one of those...

I still prefer using a Petzel laser if I'm caving in a dry cave.

seanspotatobiz

Thanks, buddy. The lamp was manufactured by Joseph Lucas. You can see some similar lamps here. Joseph Lucas' company was based in Birmingham, England, so maybe these particular designs are scarcer in the US.

Ixaviers

Hi, I don't have knowledge about electronic. Could you please publish a step by step tutorial of how to convert the carbide to Dynamo lamp? I own one and I would like to make the convertion. Thanks & Regards

2_i 

The circuit is on the wacko side, but if this setup works for you in practice, there may be no reason to change it.

alo

I suggest, buy an appropriate LED light with rechargeable batteries, and put the parts in your light housing. If it is connected the same electrically, but just in a different location, it should work the same.

I suggest, don't go with a bottle dynamo. They create too much resistance when riding.

alo

There might be reasons why you have done some of what you have done, that I don't know about. Following are some ideas as to what I would do. If you have good reasons for doing what you have done, that is ok.

Do you really have your front and rear light in series? Is your taillight as bright as your headlight? I normally have a much brighter headlight, than taillight. Having them in series would probably reduce the brightness of the headlight. If it was me, and I left everything else the same, I would put both leds in the headlight, and use a solar tail light.

I would use lithium ion batteries. You could set it up so the lights would run for a long time without the dynamo turning.

It appears to me that C1, C2, and R1, are to prevent too much power going to the leds and burning them out. But this also limits how fast the batteries charge.

It appears that the lights are always on, to prevent the batteries from overcharging, which would destroy them. It may still be possible with this set up to overcharge the batteries when going fast, or even burn out the leds.

I would get rid of C1, C2, and R1, so you can get more power from the dynamo.

I would run lights in parallel. But you need some circuitry to limit the current.

Then with lithium ion batteries, you probably need a regulator to stop then charging when they are fully charged. Even with your existing batteries you should have a regulator. Using C1, C2, and R1, and having the lights always on, makes do, but could fail.

What you have done may be simpler. What I have suggested would enable you charge the batteries, then you could run the lights for a long time while the dynamo is not turning, but is a bit more involved.

I believe with your set up, if you always ride fast, the lights will be brighter, and if you always ride slow, the lights will be dimmer. If you slow down for a short time, you probably wont notice it, as it takes time for the batteries to discharge. With what I have suggested, you could run the lights at optimal brightness.

alo

Here is are examples of current limiter circuits, you can use with led lights.

https://www.homemade-circuits.com/un...rrent-limiter/

Tourist in MSN

Nine year old thread.

gilesa

The age of the thread is a lot less than the age of the parts! Even if OP is long gone, the topic still seems relevant, at least to those who like both old parts and seeing ahead.

My first reaction to this was: should it be done at all? Carbide bicycle lamps are now rare (I think I never saw one outside a museum) and this feels a little close to vandalism to me. On the other hand, it did put the thing back into active use.

Second, the OP missed out the most interesting part: how were those LEDs mounted and connected? Those surface mount parts are designed for mass production and hard to use in a one-off, at least in my experience.

Third: the electronics. I disagree with most of the comments above. There is no need for current limiting, as the LEDs are capable of absorbing the full output of the dynamo. Internal inductance makes dynamo current self-limiting, unless capacitance is added to cancel it. C1 and C2 do just that, providing a boost to output at some fairly low speed. C3 and R1 seem unnecessary. Why put a capacitor in parallel with batteries? R1 is there to bias C1/C2 but it will not help much as they will likely charge through their inherent reverse diodes much faster. A diode from the other dynamo terminal to the C1/2 junction would work better, but I am not confident that is worthwhile. Also, why not Schottky diodes?

The weakness of the circuit is that the actual output, moving and stationary, depends on balancing the poorly-specified and temperature-dependent voltages of batteries and LEDs. But the simplicity is nice. My guess is that it worked well in practice.

steelbikeguy

the circuit is a hodge-podge of ideas that are far from optimal, but still produce some light, which is probably all that the OP really wanted.

The arrangement of just wiring 5 series nicads (NiMH?) cells across two LEDs in series is also far from optimal.
On the plus side, you'll never overcharge the cells, at least as long as they are all fairly equally charged. Fully charged cells would be about 1.35V, Since the battery is connected in parallel with the two LEDs wired in series, the voltage across the LEDs will determine the voltage that the battery charges to. It's not clear, but I'm assuming that one LED is white and the other is red. This will add up to less than 6V total, meaning that the cells will be charged up to less than 1.2V, which is probably a bit less than half capacity.

On the down side... once the dynamo stops, the battery will start sourcing current to the LEDs. After a while, the cells will eventually discharge. One cell will inevitably discharge first, and then the other cells will start forcing current through the discharged cell, reverse charging it. A better design would disconnect the battery before there is a chance to fully discharge a cell.
There's also the issue that very little capacity of the battery is used, since the voltage will start dropping as the cells discharge, causing the LEDs to be dimmer and dimmer.

Steve in Peoria

unterhausen

I have a car headlamp and it's missing the reflector. I guess this is pretty common. Beautiful piece of metal though.

gilesa

I was assuming two white LEDs. With a red the batteries would barely charge.