Standlight circuits (Will this circuit work?)

syc

Newly Enlightened
Joined
Jun 10, 2008
Messages
155
Hi,
I've been trying to soak up all the material on this forum, and reading from Martin's site with all it's excellent circuits for dynamo lights. I'd like to try building a fairly simple dual Cree headlight, but with hopefully with a standlight.

I haven't looked at circuit diagrams since I was a kid building crystal radios, so please forgive me if my questions are too obvious. Reading the forum messages, it sounds like you need to add a supercap in parallel to one of the led's for the standlight to work. But I've also read that you need to put a resistor in there as well, and on another site there was a diagram with a zener diode as a shunt to protect the goldcap from too much voltage.

What is the simplest way to get this working? Is it as simple as this?

2963780296_eb9a2efbfc_o.jpg


Or does there need to be more between the Goldcap and the LED? What are the drawbacks to this design? I'd like to get a minute or two of light from the standlight when stopped and I would usually ride this bike at around 10mph for commuting. I'm planning on driving this first light from a bottle dynamo if that makes a difference.

Thanks!
 
Last edited:

MrAl

Flashlight Enthusiast
Joined
Sep 9, 2001
Messages
3,144
Location
New Jersey
Re: Will this circuit work?

Hi there,

I dont see any current limit resistor in there.
Also, if you are going to use this circuit, a good idea
is to use a reverse protection diode across the top Cree.
Many LEDs are not designed to take too much reverse
voltage, and actually zero reverse voltage is the best.
A reverse diode will bring this down to 0.7 volts and
that is usually close enough.

So that's two things to add then:
A current limit resistor
A reverse diode across the top LED

What we dont know is how much current your generator
puts out while pedaling at full speed.
 

syc

Newly Enlightened
Joined
Jun 10, 2008
Messages
155
Re: Will this circuit work?

I think this message was mistakenly moved to this forum - the circuit is for a dynamo powered light, with a bridge rectifier. There's not batteries involved and the current of the dynamos are limited to around .5 amps.
 

MrAl

Flashlight Enthusiast
Joined
Sep 9, 2001
Messages
3,144
Location
New Jersey
Re: Will this circuit work?

Hi again,

Ok, then a reverse protection diode should do it.
 

Mr Happy

Flashlight Enthusiast
Joined
Nov 21, 2007
Messages
5,390
Location
Southern California
Re: Will this circuit work?

I was going to say you might need a zener diode too, but as long as the Crees are always in circuit and never switched out, they themselves will act as a shunt to protect the capacitor. The maximum voltage across the capacitor should not exceed the highest forward voltage of a Cree, which should always be less than 5 V.

Be thoughtful about the maximum current from the dynamo however. If you should be cycling fast or going down a hill you might be able to generate more dynamo current than you expect.
 

syc

Newly Enlightened
Joined
Jun 10, 2008
Messages
155
Re: Will this circuit work?

Thanks - I am trying to understand the reason for the reverse protection diode. Is it to prevent the goldcap from discharging the wrong way through the top led? So in that case, does it go in series right below the top led, but before the junction with the 2nd led and the goldcap. Or do you mean in series before the top led and after the smoothing cap?

Or, when you say across, do you mean it is parallel to the first led?

Once I get an idea what you mean, I'll upload another sketch and you can tell me if I'm understanding you properly.

Thanks!
 

Mr Happy

Flashlight Enthusiast
Joined
Nov 21, 2007
Messages
5,390
Location
Southern California
Re: Will this circuit work?

I think the reverse protection diode is indeed to prevent the Goldcap from discharging the wrong way through the top LED. However, when I look at your circuit I don't see any discharge path for this to happen, so it seems to me you would be safe as things are (except for the possibility of excess current blowing up the Crees if you happen to pedal too fast).

I will wait for MrAl to comment further as he made the original suggestion.
 

syc

Newly Enlightened
Joined
Jun 10, 2008
Messages
155
Re: Will this circuit work?

I was going to say you might need a zener diode too, but as long as the Crees are always in circuit and never switched out, they themselves will act as a shunt to protect the capacitor. The maximum voltage across the capacitor should not exceed the highest forward voltage of a Cree, which should always be less than 5 V.

From what I've read, the Crees are in the neighborhood of 3.2V - 3.7V

Be thoughtful about the maximum current from the dynamo however. If you should be cycling fast or going down a hill you might be able to generate more dynamo current than you expect.

This page has some tests of similar dynamos:
http://www.pilom.com/BicycleElectronics/Dynamo.htm

I don't think I'll be traveling over 40kph/25mph on this particular bike (even on a downhill roll), so based on his description, I may get very little over 3W until I put in a voltage doubler.
 

Mr Happy

Flashlight Enthusiast
Joined
Nov 21, 2007
Messages
5,390
Location
Southern California
Re: Will this circuit work?

Incidentally a minute or two of light from a Cree and a 1 F capacitor is a lot to hope for. I think you are likely to see only a few seconds with this setup.

Note that the original circuit you linked to was using small LEDs and current limiting resistors rather than a big fat Cree.
 

Mr Happy

Flashlight Enthusiast
Joined
Nov 21, 2007
Messages
5,390
Location
Southern California
Re: Will this circuit work?

This page has some tests of similar dynamos:
http://www.pilom.com/BicycleElectronics/Dynamo.htm

I don't think I'll be traveling over 40kph/25mph on this particular bike (even on a downhill roll), so based on his description, I may get very little over 3W until I put in a voltage doubler.
Those are interesting tests and make fascinating reading.

Be aware however that he was testing with a high voltage and low current load. When the dynamo power was 6 W, the current at the dynamo terminals was about 500 mA and the dynamo voltage was therefore about 12 V.

When you have the LEDs sinking the dynamo current the voltage will be unable to rise above the Vf of the LEDs and so the current could easily become higher than 500 mA. Whether this does or does not happen to a harmful extent will depend on the characteristics of the particular dynamo.

Also note that a voltage doubler in the circuit will not increase the output power but reduce it. It is a voltage doubler, not a power doubler. (A power doubler would be a wonderful thing but is sadly outside the currently understood laws of physics.)

Another thing to mention is that the Vf of an LED is not a single value, but varies with the current. As you increase the current through the LED the Vf increases until you exceed some maximum current when the LED is destroyed. For many high power LEDs the maximum Vf at LED destruction will not be much higher than 4 V.
 

syc

Newly Enlightened
Joined
Jun 10, 2008
Messages
155
Re: Will this circuit work?

Incidentally a minute or two of light from a Cree and a 1 F capacitor is a lot to hope for. I think you are likely to see only a few seconds with this setup.

Note that the original circuit you linked to was using small LEDs and current limiting resistors rather than a big fat Cree.

Thanks for the feedback. It sounds like I should either discard the standlight idea, or else just hook up a very small led to the supercap. If I did that, would I need a resistor like they have in that page with the rear light design?

With respect to the overloading the LED's, it sounds like this passage from that page answers the question:
Question: Why does a standard 3 W light system with the Dymotec6 not burn out the bulbs at 50 km/h ? This is because at such high speed, the matching of the load is off (current is too high) so that it can't drain maximum power.
Question: Where goes the lost energy when the load doesn't drain the maximum possible power ? It's not lost. The dynamo spins with less effort. Try and short-circuit the dynamo at full speed, the current of the driving motor will drop a lot.
 

Mr Happy

Flashlight Enthusiast
Joined
Nov 21, 2007
Messages
5,390
Location
Southern California
Re: Will this circuit work?

With respect to the overloading the LED's, it sounds like this passage from that page answers the question:
Question: Why does a standard 3 W light system with the Dymotec6 not burn out the bulbs at 50 km/h ? This is because at such high speed, the matching of the load is off (current is too high) so that it can't drain maximum power.
Question: Where goes the lost energy when the load doesn't drain the maximum possible power ? It's not lost. The dynamo spins with less effort. Try and short-circuit the dynamo at full speed, the current of the driving motor will drop a lot.
Yes I read that, and you might be OK. But remember he is talking about incandescent bulbs and not LEDs in that statement. Bulbs and LEDs have different voltage/current characteristics. Simplified a bit, it can be said that bulbs burn out if you exceed the maximum voltage, but LEDs burn out if you exceed the maximum current.

Take special note where he says "the matching of the load is off (current is too high)". This could be a sign for caution, since high current is exactly what destroys LEDs. To be really safe, it might be advisable to include some kind of current limiting in the LED circuit.
 

MrAl

Flashlight Enthusiast
Joined
Sep 9, 2001
Messages
3,144
Location
New Jersey
Re: Will this circuit work?

Hi again,

The circuit has at least two basic modes of operation.
One is when the generator is supplying power, the other
is when the cap is supplying power.
When the generator is supplying power, no problem.

When the cap is supplying power the top LED is mildly
reverse biased. This is because the diodes at the generator
can conduct a little in reverse, which puts a pseudo ground
at the anode of the top LED. With the ground at the anode
and some positive voltage at the cathode, the top LED
becomes reverse biased. Now the reverse biasing *might*
not hurt anything if the LED can take a reverse bias equal
to it's own forward voltage, but it may not want to take this
voltage for very long because many LEDs are not meant to
be run at all in reverse bias, especially on a regular basis,
even if it's a voltage equal to it's forward drop which for
some LEDs could be as low as 3v.
Put the reverse protection diode across the top LED, and
this potential problem goes away. The LED will see a max
of 0.7v with a standard si diode or 0.5 with a Schottky.
The way i see it, it's better to be safe then blow out the LED.

Now there is the counter point where the reverse conduction
of the diodes at the generator is lower than the reverse
conduction of the reverse biased top LED, meaning that the
reverse voltage that actually gets across the LED is less
than the total reverse voltage. This is true, but then that
is a static measurement too. The capacitances come into play
also which means the dynamics of the current and voltage
have to be examined. Doing this it gets a little more difficult
to analyze, but in general the capacitance of the rectifier
diodes will be higher than the capacitance of the LED, meaning
the dynamic current (transient response) at least puts
more reverse voltage across the LED than in the purely static
case, which again means that a reverse diode is certainly a
good idea.
 
Last edited:

Bandgap

Enlightened
Joined
Aug 14, 2002
Messages
284
Location
London
Re: Will this circuit work?

The original circuit is fine and will work as it is, providing the super cap is rated at more than 3.6V.

With due respect to Mr Al, who does some great stuff, Cree XR-Es are rated at 5V reverse bias so no amount of reverse leakage in the bridge rectifier is going to hurt the LEDs in this circuit.

There is no need for a zener anywhere.

Ditch the 1,000uF completely, it is a waste of volume, although will reduce flickering a bit if operating from a hub dynamo.
I use a hub without that capacitor, it is fine above 10k/h 6mph

The standlight may dim really quickly depending on the supercap.
Not all supercaps are equal.
They vary dramatically in internal resistance depending on their design
Low impedance types will charge quickly, but start bright and very quickly go dim.
High impedance types will charge slowly - a minute or so - start dim and get dimmer but glow a long time.
You are not going to get a very long standby with such a simple circuit.
Small modifications could make it run dim, but predictably and longer.
Also, do not go over the rated supercap voltage as it is pretty close - under 0.5V - from the damage voltage.

Why isn't this thread in the bicycle light section?

Good luck

Steve
 
Last edited:

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
Re: Will this circuit work?

Just a quick vote to agree with Steve, a.k.a. Bandgap. We've both been working with hub dynamos for close to a decade or more, and they work well with the current high power LEDs. Current is limited to 500mA, so no regulation is needed. As for a protection diode.... probably not needed, but couldn't hurt. Adding a 1N4148 is cheap and doesn't take up much space.

I hear a lot about people using supercaps wired across the LED as a standlight, but my limited experience wasn't very good. The supercap can only charge up to the LED's Vf, so any discharge at all will bring its voltage below the nominal Vf. As the voltage drops, the LED current will decrease, so a reduced amount of light will be available for some period of time.

I'd recommend trying a supercap and seeing if it is good enough. Certainly, it's better than nothing. And if it's not good enough, come back here and we'll discuss ways to improve things. :)

My preference at this time is to charge a single AA nicad cell, and use a small Zetex boost converter to drive the LED at a low current. The primary advantage is that it can run quite a while from the AA nicad. The disadvantage is the complexity and issues related to rechargeable batteries.

I'd recommend that folks calculate the energy stored by the cap or battery, calculate how much energy can be extracted, and how much energy is needed (power times the desired run time). This can give you a good estimate on whether a design is viable or not.

and as always, don't be afraid to try new ideas, and have fun!

regards,
Steve K.
 

MrAl

Flashlight Enthusiast
Joined
Sep 9, 2001
Messages
3,144
Location
New Jersey
Re: Will this circuit work?

Hello there,


Yes, in this case the LED has a reverse rating, but i think i
would want to be careful anyway. If i was going to connect
two LEDs to a generator, i would want reverse protection
diodes across both of them. With the caps in the circuit
however just the top one would suffice.

Now that i take another more close look at this circuit i see
that there could be a better way to run the lower LED.
Charge the cap directly, with maybe a 4.7v zener across it,
then run the lower LED in series but also with the collector
of a PNP (through 5 ohms) where the PNP turns on when the
generator stops turning. This is just one example, where
the lower LED would run longer due to better utilization
of the energy storage capability of the super cap.

Another idea of course would be to run the lower LED with
a converter circuit, which would utilize more of the cap
energy storage too.
 

syc

Newly Enlightened
Joined
Jun 10, 2008
Messages
155
Re: Will this circuit work?

Thanks everyone for your replies!

I'd recommend trying a supercap and seeing if it is good enough. Certainly, it's better than nothing. And if it's not good enough, come back here and we'll discuss ways to improve things. :)

That's pretty much what I'll do - though I'm tempted to track down a breadboard from one of my friends and prototype it to see if the bottle dynamo can power both LEDs at a reasonable speed.

I'll probably build this circuit:
2968822390_a5b637df67_o.jpg


Would adding a resistor between the goldcap and the Cree limit the rate at which the goldcap is discharged? I only really hope for the standlight to be a "be seen" light, so limiting the amount of current to around the minimum to get some light out of the led is fine.
 
Last edited:

FrontRanger

Newly Enlightened
Joined
Oct 22, 2007
Messages
166
Location
Colorado, USA
Re: Will this circuit work?

Reading the forum messages, it sounds like you need to add a supercap in parallel to one of the led's for the standlight to work. But I've also read that you need to put a resistor in there as well, and on another site there was a diagram with a zener diode as a shunt to protect the goldcap from too much voltage.
...
What is the simplest way to get this working?

Not to rain on your parade, syc, but have you considered just mounting a flashlight on your bars? To make a repair in the dark, you need a light source that isn't dynamo-powered. And if you're carrying one of those around anyway, why not use it as a standlight? With this reasoning in mind, I ditched the standlight idea and augment my dynamo headlamp with an L2D.
 

isomax

Newly Enlightened
Joined
Apr 17, 2008
Messages
2
Re: Will this circuit work?

Hi, sorry for my very bad english, i am spanish. I have made something similar to this. I have not the circuit at hand at this time. But i will made some concerns to your circuit.

1. The goldcap you have suggested is not sufficient. This goldcaps are designed for ultralow current applications (memories, etc...). If you plan to use a goldcap you must use the bigger ones, maybe 10F and 2,5v. Obviously it is not possible to power a white led with one (the 2,5v...) and it is neccesary to put 2 or more in series (the resulting capacity is reduced). In my project, i used a yellow powerled (2,2 Vf vs 3,5 of the white) and the 5mm ones in an dynamo-converted battery tail light in series-parallel (the half 5mm reds in parallel, the other half in parallel and in series with other ones, and this group in series with the yellow power led), powered by a 3 goldcaps "battery" of 10F each one, connected in series. This gave me a 6,4 Vf, pretty close to the 7,5 V of the battery maximum V and far from overcharge it. There is a limiting resistor "before" the goldcaps to prevent no lighting until the goldcaps are charged, and another one "after" the goldcaps to ensure some backup time.

2. I putted the white powerleds in series after being rectified by a diode bridge, like you, but there is a more efficient way but more expensive: putting 2 in series and another 2 in series in reversal with other ones, and not using a rectifier. In this image, you have a similar (and much earlier circuit with 5mm leds) circuit wich can give you ideas, the final circuit does not have the backup whites. If finally you put the white poweleds after a rectifier, don't use a pre-made one, with silicon diodes, make one with germanium (or shotky barrier) diodes, with have a lower Vf.

3. I have not used a zener. In earlier stages of the development i considered it, but after a few testing i refused. It is not neccesary and introduces some important losses. The current of the dynamo is similar to self-regulated and with a SSC's P4, capable of driving one amp (wich the dynamo is not capable of generating) it's not possible of burning the leds. With a good design it is not possible to have a voltage greater than the 7,5 V of wich the 3 series goldcaps is capable.

Sorry of no providing you the final circuit but i am sure you can make a similar one with my explainings.

pict0224pt3.jpg
 

StevelKnievel

Newly Enlightened
Joined
Oct 13, 2007
Messages
114
Re: Will this circuit work?

I have built lots of dynamo lights using the circuit you have drawn! It works very well exactly as you have it! The standlight will provide sufficient light to be seen for around 5 minutes, more if you use a 1.5F 5.5V goldcap. There is no resistor needed in between the Cree and the goldcap. If you want an even better standlight use two gold caps in series in place of the smoothing capacitor. Since you are using a bottle dynamo flickering at low speed will still be minimal. This arrangement will keep both LEDs lit when you stop for even better visibility. If you do this you should use some voltage balancing resistors (10K ohms or so) across each goldcap to make sure the caps discharge fully overnight, this will prevent the caps from charging unevenly.
 
Top