New DIY dynamo light project

[Brucey]

i/ know it sounds too good to be true but when you see it you will realise I am not joking.

[jb]

Without any disrespect for the enthusiasm of DIY builders of electrical circuits this thread reminds me of my valve amp kit building days when the magazine's were full of home projects to better the circuits designed by professional companies. Some times they achieved a small gain but mostly at the expense of either reliability or safety or both - not helped by the fact that the quality of the sound was mostly subjective.
Still, it's an interesting thread to follow & I'm sure 'edocastor' is getting a lot out of it, as am I.

[cycle tramp]

i/ know it sounds too good to be true but when you see it you will realise I am not joking.

The pledge of 50 quid still stands... you could be sitting on a goldmine and cause bottle dynamos to under go a complete re-birth...

..After all the pneumatic tyre, the telephone and even the radio were simply the products of people following an idea through to reality...

[Brucey]

.....Don't keep us in the dark.......

see what he did there?

[rjb]

.....Don't keep us in the dark.......

see what he did there?

Pun intended.

[Cyclothesist]

It's certainly generating some interest! Can you be induced to tell us more Brucey?

[SA_SA_SA]

It's certainly generating some interest! Can you be induced to tell us more Brucey?

I have a suspicion of what it might very very possibly be
But I wonder if Brucey was wondering if it was patentable?

[gat]

To @edocaster ,

Your description of how that circuit operates is correct. There are a some additional points:

the LM393 has an open collector output (NPN), so the 10k resistor will pull the MOSFET gate and source to the same voltage. A seasoned switch-mode circuit designer would likely be horrified by the time constant: it will take about a microsecond to turn the thing off. But even with a bottle that happens only at a few kHz and it is only 6V, so the switching loss is low.

The switching happens on every half-cycle of dynamo output. That seems important for performance as it keeps the load voltage low while current is building in the early part of a half-cycle.

The smoothing capacitor works with the placement of the sense resistor to skew the switching points late in half cycle. At 30kph the load is still 12v well after the next half cycle has begun, helping to kill what is now out-of-phase current. I should admit that I did not intentionally design that feature; it is more of a happy accident.

On the lack of comparator hysteresis: it seems fine in simulation, perhaps because the sense current is either rising or falling strongly at the switch point. In the real world a small positive feedback capacitor or RC may be useful.

To @jb,

I think your analogy may not hold, because there is not so much tradition of designing control circuits for a permanent magnet alternator at such varying speeds. I have seen only one attempt to find out what is inside a modern LED headlamp for dynamos, and the circuit was fairly simplistic. Also, the Forumlader USB charger, that I think came from discussions like this, is reported to significantly out-perform competitors.

[gat]

Since @Brucey is being coy about his invention, we are invited to speculate. My entries:

It is a variable-strength spring to push the wheel onto the tyre. Mostly a low pressure, so quiet. When it rains, something (electronic?) detects slip and applies more pressure. It resets when you pull the bottle off the tyre.

It is a purely mechanical extended bottle cap with spokes that engage the wheel spokes. As it speeds up a cenrtifugal governor withdraws some spokes so it runs at roughly constants speed.

Any others?

I never had any problems with bottles after accepting the whine, but I have not used one much in the rain. I suspect the drag is less than a flapping jacket causes.

[jb]

I would have thought the lightspin dynamo with a decent friction wheel would be the ultimate....

[edocaster]

@gat

Ah, I hadn't figured the switching happened every half cycle. Just one of those things with rectified dynamo output and whether or not it's smoothed.

The mention of forumslader is interesting too. I made a very simple linear regulator for USB back in the day, but remember reading about forumslader, but it's pretty hard to follow in English, and it seems later versions are significantly more complicated. Maybe a future project!

[Cyclothesist]

@edocaster
I just wanted to day how much I enjoyed this thread as a long time tinkerer in electronics. Thanks for sharing and inspiring. I've a couple of old bottle dynamos waiting to combine your circuit inspiration with Brucey's free energy mechanical add on.

[Brucey]

Any others?

nice try, but
no cigar....

[SA_SA_SA]

Any others?

nice try, but
no cigar....

Is it patentable?

[Cyclothesist]

i/ know it sounds too good to be true but when you see it you will realise I am not joking.

Is it?:
You dismantle the bottle dynamo, remove the magnets and epoxy them to the rear disk brake rotor arms (inside the braking track and clear of the calipers). Then you flatten out the copper coils with a few judicious taps of your aero spoke hammer. The coils are then attached to the chainstays and seatstays level with the rotor magnets using more epoxy and a couple of shaped spoke formers. The coils are connected to your lighting circuit using connectors made from some old brass brake bushings and cable end caps (a 2 minute job for anyone with an angle grinder, bench vice and laser cutter). Jobs a good 'un!