Hub dynamo to drive P7 headlight?

ihab

Newly Enlightened
Joined
Oct 21, 2009
Messages
9
Hi folks,

[ Newbie, first posting. If I miss something, please let me know. ]

I have a Shimano DH-3N72 dynamo on my bike at the moment. It's driving a B&M IQ Fly Senso Plus headlight and B&M DToplight XS Plus taillight.

Everything works great except that I'm jealous of friends who use P7 700-900lm flashlights from DealExtreme. :) Hence I'm wondering if I can build a dynamo headlight that will compete.

For the bike riding during which I need the brightest light, I am going 15-20mph. At that speed, I imagine the dynamo should be able to put out quite a bit of power. From data on this site, it seems that I can get 6W at 15mph:

http://www.pilom.com/BicycleElectronics/HubDynamo.htm

From other postings here, it seems the dynamo produces 0.5A with voltage varying according to speed. From the manufacturer specs, the P7 seems to want up to 2.8A. Should I build a step-down transformer on the AC side, prior to rectifier input, to get a higher current?

What is the minimal and most efficient circuitry to protect the LED?

All in all, is this whole project worth doing?

Thanks and regards,

Ihab
 

znomit

Enlightened
Joined
Aug 1, 2007
Messages
979
Location
New Zealand
P7 is no good off the dyno as all 4 leds are in parallel. You want a cree MC-E wired in series(equivalent to 4led circuit on pilom.com).
Personally for a road light the MC-E/P7 isn't very good as it requires large optics to get a decent beam. You are likely better off with 2-3 XP-E or G.
 

ihab

Newly Enlightened
Joined
Oct 21, 2009
Messages
9
P7 is no good off the dyno as all 4 leds are in parallel.

I see, tx for the tip. That would explain the high current requirement.

You want a cree MC-E wired in series(equivalent to 4led circuit on pilom.com).

The data sheet shows (only) about 130lm at 500mA with all 4 LEDs wired in series. Unless I'm missing something, that does not make it worth replacing the (~ 100lm I think) B&M IQ Fly.

Personally for a road light the MC-E/P7 isn't very good as it requires large optics to get a decent beam. You are likely better off with 2-3 XP-E or G.

Again, from the data sheet, at 500mA, an XP-E has ~ 3.3V drop and produces ~ 140lm. With 2-3 of these, now you're talking.

At 500mA, an XP-G has ~ the same voltage drop and produces 140% of 139lm (R5 bin) = 194lm. With 2-3 of these, now you're really talking.

I checked out your builds and will think some more and come back with questions. Thanks for the help!

Ihab
 

yellow

Flashlight Enthusiast
Joined
Oct 31, 2002
Messages
4,629
Location
Baden.at
You must have misinterpreted the datasheet value,
an easy-and-rugged point of view is, that the MC-E is the same than four XR-E.
So when running series wired XR-E or MC-E all the data is (roughly) the same.
The individual lumen of the XR-E add inside the MC-E, making for the same 500+ lumen

downside: with You hub dynamo there might be some flickering at very low speeds,
four led in series is a tiny bit too much.
But should work. Give it a try.
 

Edocaster

Newly Enlightened
Joined
Mar 27, 2014
Messages
51
Should I build a step-down transformer on the AC side, prior to rectifier input, to get a higher current?

Resurrecting an old thread here...

So, would a step-down transformer work to raise the current going into one or two LEDs, as opposed to running a long string of LEDs to extract more power out of a dynamo?

Something with, say, a 2:1 ratio to get 1A (as opposed to the usual 500mA) into one or two LEDs? If so, would it be as simple as using the transformer in something like this: http://www.amazon.co.uk/dp/B002SIVUJM/

...or will this not work at all due to something in the nature of using a bike dynamo as a source?
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
the transformer in question is rated for 45W at 110v at the output, so the output windings are rated for 0.41A. This is less than the desired 1 amp.

Another consideration is that this transformer is designed for 60Hz. Some/many hub dynamos have 26 poles, so the wheel would have to make two revolutions per second to get close to 60Hz. Below this speed, the transformer won't be as effective.
 

inetdog

Enlightened
Joined
Mar 4, 2013
Messages
442
Not only would the transformer not produce an effective output, it would also be a much greater load on the wheel rotation, even with no load on the output.
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
Not only would the transformer not produce an effective output, it would also be a much greater load on the wheel rotation, even with no load on the output.

can you explain this a bit? The power dissipation in a transformer with an open secondary is fairly small due to the high impedance at the secondary winding being reflected back to the primary winding.
 

inetdog

Enlightened
Joined
Mar 4, 2013
Messages
442
can you explain this a bit? The power dissipation in a transformer with an open secondary is fairly small due to the high impedance at the secondary winding being reflected back to the primary winding.
Yes, BUT...
When the frequency drops low enough the transformer core saturates or has a reduced impedance and you are left with the primary current limited by the primary resistance alone.
Your explanation is correct only when the frequency is held constant at a value within the design range of the transformer.
As a limiting case think of DC on the transformer input. Efficiency =0.
 
Last edited:

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
low wheel speed is a limited case, and probably not one where additional load is an issue.

to be a bit philosophical.... in the limiting case of DC, can the efficiency be zero if the output is zero?? ;)

I think there is agreement that a transformer is a suboptimal way of matching a single LED to a typical bike dynamo, but it could probably work within a certain range of speeds, especially if the transformer was designed for the application.
 

alpg88

Flashlight Enthusiast
Joined
Apr 19, 2005
Messages
5,283
not even going into detaiails, and technical issues, you better off with batteries. prbly wont find many dynamos capable of putting out 10w, and those that do will make it noticably harder, to say the least, to pedal your bike.
 

inetdog

Enlightened
Joined
Mar 4, 2013
Messages
442
low wheel speed is a limited case, and probably not one where additional load is an issue.

to be a bit philosophical.... in the limiting case of DC, can the efficiency be zero if the output is zero?? ;)

I think there is agreement that a transformer is a suboptimal way of matching a single LED to a typical bike dynamo, but it could probably work within a certain range of speeds, especially if the transformer was designed for the application.
Since efficiency is output over input, then with non zero input and zero output you definitely have an efficiency of zero.
It would be undefined if the input were also zero.
If you take the limit as speed goes to zero, the output decreases faster than the input, so the efficiency approaches zero too.
 

inetdog

Enlightened
Joined
Mar 4, 2013
Messages
442
not even going into detaiails, and technical issues, you better off with batteries. prbly wont find many dynamos capable of putting out 10w, and those that do will make it noticably harder, to say the least, to pedal your bike.
Batteries charged by a dynamo as well as recharged at home would be an option too.
The last thing you want is your taillight going out when you stop.
But the run time on a good battery pack may be long enough that the dynamo is not needed.
 

alpg88

Flashlight Enthusiast
Joined
Apr 19, 2005
Messages
5,283
Batteries charged by a dynamo as well as recharged at home would be an option too.
.
in theory yes, but how many such devices have you seen in real world? i have not seen even 1, there is prbly a very good reason why.

a small 1 inc. bulb light, with dynamo that i had, from good old days, made it feel like i'm pedaling uphill. a pack of 4 18650 will power a lot stronger light than any dynamo driven light can put out, with plenty of runtime. if you talking about few 5mm leds front and back, than dynamo is ok, anything more will make you tired a lot sooner.
 

Edocaster

Newly Enlightened
Joined
Mar 27, 2014
Messages
51
not even going into detaiails, and technical issues, you better off with batteries. prbly wont find many dynamos capable of putting out 10w, and those that do will make it noticably harder, to say the least, to pedal your bike.

I've already got over 7.5W - http://www.candlepowerforums.com/vb/showthread.php?385350-5x-Cree-XM-L-dynamo-light-(in-progress) - but that's using a string of LEDs. Drag-wise, it's not bad at all. Plus, I would only really be using high mode on the flat or downhill. It's actually brighter than my Magicshine clone.

Low mode (voltage doubler) has half the drag (not that I think I'd confidently know which mode I was in blindfolded), and still plenty of light (output around 3.5W).
 

find_bruce

Newly Enlightened
Joined
May 5, 2011
Messages
84
in theory yes, but how many such devices have you seen in real world? i have not seen even 1, there is prbly a very good reason why.

a small 1 inc. bulb light, with dynamo that i had, from good old days, made it feel like i'm pedaling uphill. a pack of 4 18650 will power a lot stronger light than any dynamo driven light can put out, with plenty of runtime. if you talking about few 5mm leds front and back, than dynamo is ok, anything more will make you tired a lot sooner.
alpg88 you need to get out more - there are a number of devices for charging batteries from a bike dynamo, B&M ewerk for example.

What you say is correct for the sidewall dynamos I remember from way back when - obvious difficulty pedaling, to produce a feeble light.

It is not correct for a modern hub dynamo where the drag, whilst it exists, is imperceptible. It is so minimal that I didn't bother with an off switch for my lights, they are on whenever I am riding. Front & rear lights stay on for at least 5 minutes after stopping using a couple of capacitors. Other ways to skin that cat such as using a nicad batttery

I've already got over 7.5W - http://www.candlepowerforums.com/vb/showthread.php?385350-5x-Cree-XM-L-dynamo-light-(in-progress) - but that's using a string of LEDs. Drag-wise, it's not bad at all. Plus, I would only really be using high mode on the flat or downhill. It's actually brighter than my Magicshine clone.

Low mode (voltage doubler) has half the drag (not that I think I'd confidently know which mode I was in blindfolded), and still plenty of light (output around 3.5W).
I have gone a slightly different route to Edocaster with similar results. Plenty of light, the front appears to be brighter than my magicshine, the rear is waaay brighter than most commercially available lights and I never have to remember to recharge batteries.

BTW Edocaster I am no expert but think you may be mistaken about the drag in voltage doubler mode - there is no free lunch and double the available voltage, half the current = same power.
 

Edocaster

Newly Enlightened
Joined
Mar 27, 2014
Messages
51
BTW Edocaster I am no expert but think you may be mistaken about the drag in voltage doubler mode - there is no free lunch and double the available voltage, half the current = same power.

That would be a shame if it really did have the same drag. But do remember that I mean half the drag 'at speed' (i.e. when the output maxes out in whatever mode the light is in). The voltage in both modes is clamped/shunted by the LED string - to about 14.3V in full-wave rectifier mode, and 13.7V in voltage doubler mode (the difference due to the LED Vf being lower at half current). So the voltage doubler doesn't reach double the voltage of full-wave rectifier mode - it just gets to the same voltage (slightly lower, in fact) at a lower speed, and with half the current.
 

find_bruce

Newly Enlightened
Joined
May 5, 2011
Messages
84
Ah, it seems we were talking at slightly cross purposes & I didn't think about the effect of speed. If I had, I would have referred back to Martin's graph
DualModeComp.gif


To the extent that the drag is produced by the load on the dynamo (as opposed to friction etc) voltage doubler mode would create more drag at low speeds & half the drag at higher speeds.

Not that it matters much - 1/2 not very much is still not very much
 
Last edited:
Top