Standlight circuits (Will this circuit work?)

[Steve K]

Re: Will this circuit work?

......
This is the circuit i built: Link in Flikr

I was referring to the one equivalent to the D1 in your first circuit, i just tried to connect R4/C2 to the + side of the rectifying diodes, i guess maybe some flow from the supercap can go to the shutdown without it.

In my circuit, R4 and C2 form a low pass filter. This filter smooths the AC voltage into something much closer to a constant voltage. This keeps the standlight turned off when the AC voltage drops to zero (which happens twice during the AC waveform, or about 52 times for each revolution of my SON dynamo). D1 allows current to flow into C2 and be stored there. If D1 is removed, then C2 will discharge through the LED, and the standlight would turn on.

yes i will only use one led Cree Led... i will make some test including more or maybe a couple of 5mm

looking at your circuit... am I correct in thinking that this is not going to be a combined headlight and standlight?
It looks like your intent is to make just a stand-alone standlight.
If you plan to wire this in parallel with a regular headlight, you should know that this is probably not going to work. If it is wired in parallel with a regular headlight, this standlight will turn on at a lower voltage than the headlight, and will try to absorb all of the current from the dynamo.

When I made a front standlight that was separate from the headlight, I used the front standlight shown in the schematic and photo in my previous post. This was wired in series with the headlight, and used the voltage drop across a bunch of paralleled yellow LEDs to charge a nicad battery. The yellow LEDs only conducted current in one direction, so I used red LEDs (as a taillight) to conduct current in the opposite direction. A rectifier diode could replace the red LEDs if a taillight isn't needed.

 

[pbrena]

Re: Will this circuit work?

Thanks Steve for your previos and current clarifications

It looks like your intent is to make just a stand-alone standlight.

Right !

is not going to be a combined headlight and standlight?

well, it was meant to be an independent front light that would eventually be in series with a currently undetermined rear light.

looking at your circuit... am I correct in thinking that this... If you plan to wire this in parallel...
with a regular headlight, you should know that this is probably not going to work. If it is wired in parallel with a regular headlight, this standlight will turn on at a lower voltage than the headlight, and will try to absorb all of the current from the dynamo.

When I made a front standlight that was separate from the headlight, I used the front standlight shown in the schematic and photo in my previous post. This was wired in series with the headlight, and used the voltage drop across a bunch of paralleled yellow LEDs to charge a nicad battery. The yellow LEDs only conducted current in one direction, so I used red LEDs (as a taillight) to conduct current in the opposite direction. A rectifier diode could replace the red LEDs if a taillight isn't needed.

I see, thanks a lot for all of these... I hadn't realize i could integrate both and just tried to tackle the front first, following Alex's idea for the front light that will stay lit for some minutes if the bike stopped. If you bear with me, a short description of my needs is called for.

In my Bike fleet and lights stock, I have a good number of vintage front and rear lights that I love (such as Cibie, Radios, Soubitez, Luxor, etc.) and I will be replacing the old incandescent circuitry in them with LED electronics, please take a look at these pics to have a an idea of the lights and the preliminary results:

FrontOff, FrontOn, FrontNoFlash, FrontStandMode, PosibleRearExample

So...

* I had found my Alex's circuit tests promising and a big leap comparing with bulbs, and i was planing to build it and eventually use it in series with a rear circuit like a rear Toplight for example.
* I find combining the front and rear circuits compelling, I just hesitate because:

** as you might remember, I would prefer the caps circuits (similar to the commercial options) in order to avoid the switch that i m certain i will forget to use.
** In your circuit rear circuit the Toplight provides the rear stand light right???
** combinations of 10 reds, 10 yelow and 4 Front might not fit in the planned light bodies...
** In the case of the pictures above for example, I would use only a single Cree, the rear might have about 6 red leds like yours (or maybe little more if using some surface mounting like those red Osram I ve read about.

So, What are your thoughts??? Could i use a supercap instead somehow?
How could I vary the number of leds to accommodate a particular configuration? Are my only combination options 10 leds vs a single diode??? and may I use a single Cree instead of your 4 front leds???
the toplight in your schematic is the circuit posted in a previos post ? or is an actual commercial rear light? (i even think i saw a proposed improvement from you)

Thanks!

 

[Steve K]

Re: Will this circuit work?

Thanks Steve for your previos and current clarifications

(Steve's question was "is it a stand-alone standlight?)
Right !

(Steve's question: "looking at your circuit... am I correct in thinking that this is not going to be a combined headlight and standlight?")
well, it was meant to be an independent front light that would eventually be in series with a currently undetermined rear light.

I see, thanks a lot for all of these... I hadn't realize i could integrate both and just tried to tackle the front first, following Alex's idea for the front light that will stay lit for some minutes if the bike stopped. If you bear with me, a short description of my needs is called for.

In my Bike fleet and lights stock, I have a good number of vintage front and rear lights that I love (such as Cibie, Radios, Soubitez, Luxor, etc.) and I will be replacing the old incandescent circuitry in them with LED electronics, please take a look at these pics to have a an idea of the lights and the preliminary results:

FrontOff, FrontOn, FrontNoFlash, FrontStandMode, PosibleRearExample

So...

* I had found my Alex's circuit tests promising and a big leap comparing with bulbs, and i was planing to build it and eventually use it in series with a rear circuit like a rear Toplight for example.

The LEDs are a huge improvement over the older incandescent bulbs, but they do have different electrical characteristics (and thermal characteristics, and optical characteristics...) that require attention from the designer.

When you say "Toplight", I'm assuming that you are referring to a B&M Toplight? They make a few models, and I used to have one many years ago. It was designed to be wired in parallel with the headlight, and I think that's pretty standard.

* I find combining the front and rear circuits compelling, I just hesitate because:

** as you might remember, I would prefer the caps circuits (similar to the commercial options) in order to avoid the switch that i m certain i will forget to use.

The smaller supercaps don't provide an especially bright standlight, nor does it last long, but it is certainly compact, charges up quickly, and doesn't need a switch.

** In your circuit rear circuit the Toplight provides the rear stand light right???

I've made a lot of lights, and don't use a Toplight. My taillights are made with 10 red LEDs in parallel, and the standlight operates from the voltage drop across those 10 LEDs. This design will only be practical when using a nicad (or NiMH) battery to store the power. A supercap charged up to only 1.3V (or so) just won't provide much of a run time for the standlight.
These taillights are wired in series with the headlight.

** combinations of 10 reds, 10 yelow and 4 Front might not fit in the planned light bodies...
** In the case of the pictures above for example, I would use only a single Cree, the rear might have about 6 red leds like yours (or maybe little more if using some surface mounting like those red Osram I ve read about.

So, What are your thoughts??? Could i use a supercap instead somehow?

maybe? If you only want one LED in the headlight, then a new circuit would be required, and I haven't given any thought to that sort of circuit.

How could I vary the number of leds to accommodate a particular configuration? Are my only combination options 10 leds vs a single diode??? and may I use a single Cree instead of your 4 front leds???
the toplight in your schematic is the circuit posted in a previos post ? or is an actual commercial rear light? (i even think i saw a proposed improvement from you)

there are lots of options... but as far as circuits that I've designed and built, I haven't designed a headlight with a single white LED.

The subject of retrofitting LEDs into vintage lights comes up now and then, and I find it to be an interesting topic. There are a lot of issues involved, as I noted earlier. The main issues are optical and thermal, both of which are more of a mechanical matter than an electronic matter.

The optical issue is the problem of taking optics/reflectors that were designed for an incandescent bulb that radiates light over a (nearly) complete sphere, and then trying to use it with a LED that radiates over less than half a sphere.

The thermal issue is pretty basic; incandescent bulbs work because a piece of wire is incredibly hot. LEDs work because electrons jump across a bandgap and produce photons. LEDs do not like to be hot, and actually generate a fair amount of heat. It is very important to provide a way for the LED to get rid of the heat. This usually involves using aluminum or copper to conduct the heat from the LED to the air.

I really do like your headlight.. is it a Radios model? Very stylish!
I've got an old Jos dynamo block that I bought back in the 1970's....

There is very little space in the light for much circuitry. I've thought about a way to mount two tiny Cree XP-E LEDs in the space of the incandescent bulb and still get the heat out through the aluminum reflector.

I've also got a Union headlight that came with an inexpensive dynamo set, and the bulb mounts from the rear. There is no internal space at all, so any circuitry or heatsinking would have to be external. This would also be a challenge!

Can you share photos of how your light is assembled? i.e. how is the bulb mounted? how do you access it? How are the wires make connection with it?
My assumption is that owners of these vintage lights would like to get the performance of an LED light, but not change the light itself.
Based on my Jos light, I think it will be hard enough to get LEDs mounted in the light. Any extra circuitry would have to be outside. If the circuitry is mounted in a shiny aluminum housing and attached out of the way, I don't think anyone would complain too much.

Well, that's a lot!
In summary, I don't know of a good way to do it, but perhaps I've overlooked something??

 

[pbrena]

Re: Will this circuit work?

arent Jos lights lovely? i also have one like yours and one of the Soubitez "golf ball"

I really do like your headlight.. is it a Radios model?

Thanks!, this one is a Cibie, curiously enough the led light bouncing back from the incandescent parabolic optics gets deflected by the non-so-transparent front acrylic and projects some brights straps into the road, not useless at all in the urban environment this particular beater bike is intended for.

Can you share photos of how your light is assembled? i.e. how is the bulb mounted? how do you access it? How are the wires make connection with it?

By all means !


first thou... this is my initial attempt and a kind of "proof of concept" (just to test the bike around with some heat sinking) it is rather basic and crude (later on i will try the ideas exposed bellow). i just hammered, bent and filed a small (very soft) alloy rod(a hight tension cable leftover the electric company left nearby and i picked up), so that its ends touched the aerodynamic body and the center touches the Cree.


it was surprisingly easy! and i just spend about an hour, the original setup had a (non-coiled) steel spring attached to the body acting as cathode touching the rear tip of the bulb and a small round metal socket as anode, where the bulb was screwed, that piece then slides into the optic .


Well, with a file it was easy to modify the socket to receive the Cree from behind, and i just snugged it inside filling a small gap with small piece of alloy).
I made the center of the rod somewhat flat trying to imitate the dimensions of the bulb/socket assembly
With some trial and error, i bent the rod in a "C" shape, in a way that its center pressed a bit the cathode/spring and the ends the touched aerodynamic body.
after a few tries i succeeded to enclose them all together and a bit pressing allowed me to close it and screw the holder all the way, all felt solid, after a year or so mounted in the bike and about 2000mi ridden that included some falls, it has proved to be bomb proof stable and with some previous (not so successful standlight) circuits i've even manage to get it on for a some hours.

The subject of retrofitting ... vintage lights comes up now and then... There are a lot of issues involved...are optical and thermal... so any circuitry or heat sinking would have to be external...

Right on!, i 've also read a lot about it (lots of fun!) and have picked up ideas here and there, for example:


* keeping most stuff external makes lots of sense, it will simplify everything, keep weight outside and improve light body stability (less torque). Modern lights manufacturers have to offer a product as a whole, we instead have the freedom to use previous bike resources and space and even integrate them if we can, for example the frame and fenders as electric wiring? or even its metal mass as heat sink?
* i saw pics of a circuit fitted within the frame tubes, very cool ! (link lost), my shot.
* Other option could be keeping circuits within a Gilles Berthoud bag for example, lots of space!
* as for cooling, what about the idea behind CPU's Fanless Case Enclosures?, that it is quite simple actually, it is a sandwich (Pic, pic)
of alloy plates, thermal compound, copper (or alloy?) wires and some screws connecting the CPU (led) and a massive alloy PC case (light body, metal fenders, frame and/or stem/handlebars)... i 've been always surprised how cool (even uncomfortable) to the touch a riding bike is, even in sunny days.

The optical issue...

... has been very interesting too!, have some ideas (not that i really need them for this particular Cibie as explained above):


One option is buy some led optics and with the help of a dremel (Im fan too) cut the incandecent optic and merge them with epoxi.


Also...how about this home made sugru negative copy out of a frend's Luxos IQ2 optics i manage to take?


out of it a number of acrylic positives copies could be made with this diy vacuum forming (video) the photos don really show, but it is remarkable detailed, and even can be used to position the LED (pic).


I have seen some promising materials for reflectiveness, sprays and "mirror" stickers

I've also got a Union headlight ... There is no internal space at all...

same here... in the case of a much seek after Jos minimalistic design ... I would like to come up with an artisan design for a heatsing/mounting hardware could be a bike rack, or even a self made handlebar stem

My assumption is that owners of these vintage lights would like to get the performance of an LED light, but not change the light itself.

My particular (maybe heretical) approach: i love the looks, but i would love them more if they actually worked, i use modern lights and appreciate their functionality but i don't really like them.


So, I can not buy what i would really like but i came to think i can make them myself and have fun in the process. a lot can be done except the (very light weight) alloy aerodynamic body that would require multi-ton heavy machinery, nobody is doing it nowadays and CNC machined copycats just don't get the spirit of them.

I'm assuming that you are referring to a B&M Toplight?

Right!

It was designed to be wired in parallel with the headlight, and I think that's pretty standard.... and don't use a Toplight...

I see, my mistake. i got mixed up when i read lumotec in your 04-Feb-03 schematic and upon looking for it stumbled with this page

My taillights ... standlight operates from the voltage drop across those 10 LEDs.. will only be practical when using a nicad (or NiMH) battery to store the power. A supercap charged up to only 1.3V (or so) just won't provide much of a run time for the standlight. These taillights are wired in series with the headlight.

i see, maybe it is time to drop the Alex's circuit. what about your 2 LED circuit from 3 oct 14, do both LEDs stay on while standing??? maybe i may use one of them as a rear (red) LED?

a standligtht of about 4-5 minutes bright enough to bee seen in a dark night would be more than enough for me

there are lots of options... but as far as circuits that I've designed and built, I haven't designed a headlight with a single white LED.

As you can tell my electronics skills are close to none (i work in IT) and as it is a huge topic info in the web is very dispersed. Maybe you could recommend some bibliography that help me prepare for this particular challenge?

 

[Steve K]

Re: Will this circuit work?

The conversion of the Cibie light is a bit crude (not unlike some of my work), but glad to hear that it is working out! Having the LED face forward loses the benefit of the reflector, but... if the Cibie is like my Jos, then the reflector is a dull bit of aluminum. If it fits, the best option might be to use a 10mm optic with the LED.

There were a lot of comments and questions, so let me see if I can address them...

* keeping most stuff external makes lots of sense, it will simplify everything, keep weight outside and improve light body stability (less torque). Modern lights manufacturers have to offer a product as a whole, we instead have the freedom to use previous bike resources and space and even integrate them if we can, for example the frame and fenders as electric wiring? or even its metal mass as heat sink?

The frame has historically been used as a return path for current. Unfortunately, this has a variety of issues. One is the need to keep mechanical connections electrically conductive, even when they weren't designed to be. You also have to be sure that the correct part of the light is grounded. Not a big deal with if you are careful, but if the dynamo has one terminal grounded, and the LED in the headlight is grounded, there is a chance that one/some of the diodes in the bridge rectifier will end up shorted and not work.

Using the fenders as a heat sink might be good, but may also require maintaining electrical isolation to avoid unintended electrical connections. There are insulating materials commonly used for this exact purpose on heatsinks (mica washers and such).

* i saw pics of a circuit fitted within the frame tubes, very cool ! (link lost), my shot.

some people are fond of putting circuits in the steerer tube. Other options might be putting the stuff in a polished aluminum tube and mounting it some place where it'll blend in with the other stuff.

* as for cooling, what about the idea behind CPU's Fanless Case Enclosures?, that it is quite simple actually, it is a sandwich (Pic, pic)
of alloy plates, thermal compound, copper (or alloy?) wires and some screws connecting the CPU (led) and a massive alloy PC case (light body, metal fenders, frame and/or stem/handlebars)... i 've been always surprised how cool (even uncomfortable) to the touch a riding bike is, even in sunny days.

The copper "wires" are actually heat pipes. They use an internal fluid that circulates internally, etc. Very high tech stuff. For bike lights, the goal is to get the heat from the LED to the air, and usually the best solution is to just send the heat to the light's body. Sending it to a fender or mounting bracket might be a good solution too.

One option is buy some led optics and with the help of a dremel (Im fan too) cut the incandecent optic and merge them with epoxi.

Also...how about this home made sugru negative copy out of a frend's Luxos IQ2 optics i manage to take?

out of it a number of acrylic positives copies could be made with this diy vacuum forming (video) the photos don really show, but it is remarkable detailed, and even can be used to position the LED (pic).

I have seen some promising materials for reflectiveness, sprays and "mirror" stickers

I'm in favor of not destroying any part of the light, so pulling out the old reflector and rigging up a mount for modern optics is a good idea. I did that with a crappy Vistalight and a Cateye Mity, but it might be tougher with vintage lights. My little Jos isn't designed to have the reflector removed. It's really dependent on the design of the light itself. Creativity is essential!

same here... in the case of a much seek after Jos minimalistic design ... I would like to come up with an artisan design for a heatsing/mounting hardware could be a bike rack, or even a self made handlebar stem

that's a cool little light! That's probably a place where you might want to make an aluminum mounting bracket that would conduct the LED's heat to a fender. You might even be able to make a finned bracket that would be able to get rid of the heat all by itself.

My particular (maybe heretical) approach: i love the looks, but i would love them more if they actually worked, i use modern lights and appreciate their functionality but i don't really like them.

I don't think it's just you. Heck, even B&M makes a modern light styled after a shiny old chromed light. Some people just buy a modern light and transplant everything into a vintage light, which seems like a challenge, assuming that the reflector and optic have to be installed. It's sure not going to work with some of the smaller vintage lights, though.

You've made a good effort so far. There are ways to improve it, such as fitting a small optic to the LED, or using a buck converter to 1 amp to the LED.

As far as adding a standlight to what you've got (i.e. a bridge rectifier and LED), that's harder.

Going back to my schematic for front and rear standlight used with a Lumotec headlight... yeah, the Lumotec was a reference to the incandescent headlight.

i see, maybe it is time to drop the Alex's circuit. what about your 2 LED circuit from 3 oct 14, do both LEDs stay on while standing??? maybe i may use one of them as a rear (red) LED?

a standligtht of about 4-5 minutes bright enough to bee seen in a dark night would be more than enough for me

In the dual LED headlight with standlight (i.e. what you are referring to as Alex's circuit), the upper LED allows generating a "high" voltage that will allow a fair amount of energy to be stored in the capacitor. The lower LED is the only one powered by the boost circuit. As such, both are essential parts of the circuit.

So what are the options?

Well, you could replace the upper LED with a power zener diode that produces the same 3V drop as the LED. .. . or use 4 plain silicon rectifier diodes (1N4001 or such) wired in series. It does waste some power, but would get you a standlight, and would also produce the 6V at the headlight terminals that would allow a regular Toplight to be used with it.

Another option would be to use an upper LED, but mount it in a small cylinder with optics somewhere stealthy. This would provide a second headlight but not take away from the aesthetics of the Cibie.

As you can tell my electronics skills are close to none (i work in IT) and as it is a huge topic info in the web is very dispersed. Maybe you could recommend some bibliography that help me prepare for this particular challenge?

boy, it's been 30 years since I got my EE degree. I'm not really sure where people go to pick up electronics as a hobby. It's hard enough to be fully skilled with a 4 year degree!

I will say that Dave Jones has a neat youtube channel called the EEVblog. In his archived videos, he has some "fundamentals Fridays", where he covers a lot of the basics. Dave has a very energetic style, so at least you won't fall asleep.

 

[pbrena]

Re: Will this circuit work?

So what are the options?

Well, you could replace the upper LED with a power zener diode that produces the same 3V drop as the LED. .. . or use 4 plain silicon rectifier diodes (1N4001 or such) wired in series...would get you a standlight, and would also produce the 6V at the headlight terminals that would allow a regular Toplight to be used with it.

Another option would be to use an upper LED...

Thank a lot Steve.

Now i do all the testing with a rear light circuit running in parallel (the improvement you suggested in your previous post #142) to ensure an integral solution.

For the time being... i will build what i got for this particular bike with the Cibie front and Luxor rear so as to use the electronic parts i bought. Later in this post i will detail my testing and general findings so far in case someone might find them useful.

Nevertheless, for other bikes and lamp bodies i will seek a solution for some problems i got building a circuit closer to the one you shared, but before ordering the parts... please your thoughts between the 2 options bellow and solve some doubts i still have.

Option1

I find your circuit Rev00e/30oct2014 the most interesting...

a) I m assuming i can use this with a rear light, right??
b) about your remark "...use an upper LED, but mount it in a small cylinder..." Certainly, a second head light would be very useful (for example in another lamp body and position), but... could i also use a Red (or orange/yellow) LED, let say a high power Cree, without major change in the circuit? (it could be a security light like this):

c) I could use the rear light in post #140 or better yet (if i understand right your comment in post #142) use a second Rev00e/30oct2014 circuit but this time with red high power Crees as taillight, right?
d) about your comment "- use better, more efficient LEDs." in quoted post 142, do you have better leds in mind? some Red Orsam Leds I bought in the past are listed now as "discontinued"
f) do you remember the stand light times you got with your 100F 2.7V Cap?
g) May I replace the 100F 2.7V Cap with the 10F 2.7V one i already got without further changes?? i find the stand light times i got from the 10F actually quite good.



Option2


I wonder if its a viable/better option to modify your circuit Rev00e/30oct2014 to get a single circuit to drive both the rear and frond light and still get stand light for both front and rear (remember I will stick to Caps)???

this would mean to drive two standlight Leds, one high power Cree (or alike) white and one red with a single zetex.
i read the zetex datasheet and it seems that a suitable current/voltage could be selected with a proper resistor/inductor/diode combination (after solving some linear equations), right???




DETAILS/FINDINGS OF MY LAST TESTs


a) I was not able to make the upper, not standing Cree consistently turn on, only using a Red Osram the 2 of them were on, using a white 5mm Niche it seemed to dangerously change its color, I cant figure out why, reading Alex Wetmore post (and his blog where the schematic dont even depict a second LED) it states it is "optional" i guess that mean it worked for him... if i do omit it , the stand light times seems to reduce up to an almost useless point. Perhaps is my bottle generator (i currently have no SON wheel build) i measured .6 Amps AC out of it.
b) I suppose that if the shutdown function is used the circuit becomes more efficient, right???, although it doesn't seem to affect the stand light time
c) at first I found the shutdown function confusing, this is what I could make after reading the Data Sheet:


Shutdown (on) = LED (off)
Shutdown (off) = LED (on) = "Normal" operation.


it is controlled by the voltage in the shutdown pin (PIN) according to the following table:


0 <= PIN <= 0.1 --> Shutdown (on)
0.1 < PIN < 0.7 --> Undocumented, my experience: .3 or .6 => LED (on)
0.7 <= PIN --> LED (on)
If PIN = Null (open circuit) --> LED (on)


d) I was able to get 4-6 minutes of stand light (with the 1 Ohm resistor in the sense pin) out of the 2.3V 4.7F Cap and about double out of the 2.7V 10F

 

[Steve K]

Re: Will this circuit work?

Option1

I find your circuit Rev00e/30oct2014 the most interesting...

for those playing along at home, let me include a link to the schematic.....
https://c2.staticflickr.com/8/7464/15674268650_e97ca7b575_o_d.png

a) I m assuming i can use this with a rear light, right??

you should be able to. It's got the same nominal voltage at the terminals as the standard 6V, 2.4W headlight.

b) about your remark "...use an upper LED, but mount it in a small cylinder..." Certainly, a second head light would be very useful (for example in another lamp body and position), but... could i also use a Red (or orange/yellow) LED, let say a high power Cree, without major change in the circuit? (it could be a security light like this):

The important thing is that the device that replaces the upper LED have a voltage drop of 3V. Careful selection of silicon and schottky diodes placed in series with the desired color LED should get you in the ballpark of 3V, which is probably good enough.
BTW, that's a neat little light. What's the story on it?

c) I could use the rear light in post #140 or better yet (if i understand right your comment in post #142) use a second Rev00e/30oct2014 circuit but this time with red high power Crees as taillight, right?

The schematic referenced in post 140 has resistors that limit the current (probably to 100mA).
By contrast, my headlight design doesn't have anything to limit the current (it relies on the dynamo's internal impedance to limit current). As such, it's not suitable for use as a taillight.

d) about your comment "- use better, more efficient LEDs." in quoted post 142, do you have better leds in mind? some Red Orsam Leds I bought in the past are listed now as "discontinued"

I usually just search Mouser or Digi-key.
Cree has some 5mm red LEDs that look pretty promising. The Digi-key part numbers are C503B-RAN-CZ0C0AA1-ND and C503B-RAN-CA0C0AA2CT-ND.
Mouser has one of them (maybe both?)... p/n 941-C503BRANCA0C0AA2
There's a Optek red 5mm LED that looks good too... p/n OVLGS0C8B9.

f) do you remember the stand light times you got with your 100F 2.7V Cap?

I think it's about 10 minutes. Plenty long for most intersections.

g) May I replace the 100F 2.7V Cap with the 10F 2.7V one i already got without further changes?? i find the stand light times i got from the 10F actually quite good.

That should work. You can also change the R6 resistor value to reduce the brightness, which will extend the run time.

Option2

I wonder if its a viable/better option to modify your circuit Rev00e/30oct2014 to get a single circuit to drive both the rear and frond light and still get stand light for both front and rear (remember I will stick to Caps)???

this would mean to drive two standlight Leds, one high power Cree (or alike) white and one red with a single zetex.
i read the zetex datasheet and it seems that a suitable current/voltage could be selected with a proper resistor/inductor/diode combination (after solving some linear equations), right???

The simplest approach would be to just put a small resistor in series with a large red LED (or group of small LEDs in parallel), and then connect that in parallel with the lower white LED.

a) I was not able to make the upper, not standing Cree consistently turn on, only using a Red Osram the 2 of them were on, using a white 5mm Niche it seemed to dangerously change its color, I cant figure out why, reading Alex Wetmore post (and his blog where the schematic dont even depict a second LED) it states it is "optional" i guess that mean it worked for him... if i do omit it , the stand light times seems to reduce up to an almost useless point. Perhaps is my bottle generator (i currently have no SON wheel build) i measured .6 Amps AC out of it.

boy.. I really need to see a schematic of whatever circuit you are building. There's not much I do without it.

Regarding Alex's circuit, he has an error in the schematic. The right side of the inductor should be connected to Q1's collector and D2's anode instead of to Q1's base. I think that was mentioned and discussed later in the thread... but I'm not going to go look for where it (probably) happened.

I don't understand how Alex got much run time from a 10F cap charged up to 2.1V. There's just not a lot of energy stored under that condition.
wait... it looks like he's using a 1 ohm sense resistor for the boost converter, where as I use a 0.1 ohm sense resistor. That reduces the brightness of the standlight quite a bit, so it would run much longer.

b) I suppose that if the shutdown function is used the circuit becomes more efficient, right???, although it doesn't seem to affect the stand light time

the shutdown input controls whether the boost converter is shut down. It doesn't affect the efficiency of the converter, but will avoid draining energy out of the capacitor (or battery) while the dynamo is running.
In my case, since I use batteries in some of my standlights, the shutdown function allows me to turn off the standlight while the bike is parked in order to avoid completely discharging the battery.

 

[pbrena]

Re: Will this circuit work?

BTW, that's a neat little light. What's the story on it?

Neat indeed!, I got it from a French guy in ebay a long while ago, i remember the listing had no further information, it has no brand stamped anywhere and i have not seen any more since :shrug:

The simplest approach would be to just put a small resistor in series with a large red LED (or group of small LEDs in parallel), and then connect that in parallel with the lower white LED.

:thumbsup: Good !, I will go that way, Now...

* In your schematics you can read a reference to a LM4041 that i cant identify in the circuit, would you clarify?

* Would you comment on your lm431 selection vs lm317?

...I really need to see a schematic of whatever circuit you are building... Regarding Alex's circuit, he has an error in the schematic...

oh, sorry, here it is, it includes the shutdown sub-circuit borrowed from your posts and the correction that Alex Wetmore identified too, but didnt redraw (for those interested and for clarity's sake I will correct my previous posts (now that i manage to paste a picture)).

... he's using a 1 ohm sense resistor... That reduces the brightness of the stand light quite a bit, so it would run much longer.

That worked for me too, it gives a lower "be seen only but still useful" light similar to those in some previous commercial lights i got.

* besides the upper led concern in previous post... i managed to use the shutdown sub-circuit only if i connected the diode to an AC leg (before the bridge as in the drawing) if i instead connected it just after the bridge (positive) similar to your circuit the Led remained off (and the voltage between pin and ground close to cero) i cant figure out why.

 

[Steve K]

Re: Will this circuit work?

* In your schematics you can read a reference to a LM4041 that i cant identify in the circuit, would you clarify?

* Would you comment on your lm431 selection vs lm317?

the reference to the LM4041 should have been removed. Maybe I started out with the LM4041 and then realized I only had LM431's on hand? They both have their virtues. I am in the process of fixing a friend's Herrmans taillight and am replacing the zener diode with a LM431. The LM431 is capable of handling more current than the LM4041, which made it preferable for the taillight.

One of the fun things about trying something new, such as using a supercap instead of a nicad battery, is that you learn things that you didn't anticipate! For this project, I learned how much current the supercap will consume while charging. In my circuit, the current to charge the supercap is taken from the lower LED, and when intially charging, the LED gets very little current and is rather dim.

If I would have used a LM317, I wouldn't have been able to use the voltage from the lower LED to charge the supercap. The LM317 requires that the input voltage be at least 2V (or so?) greater than the output voltage, so it would have needed to be connected to the voltage directly from the bridge rectifier. This would mean that both LEDs would initially be dim, which wouldn't be useful. Granted, you could compensate by adding some resistance somewhere, but that's just inefficient. Another option would be to add a current limit circuit to the LM317, but that's introducing more voltage drop and complexity.

* besides the upper led concern in previous post... i managed to use the shutdown sub-circuit only if i connected the diode to an AC leg (before the bridge as in the drawing) if i instead connected it just after the bridge (positive) similar to your circuit the Led remained off (and the voltage between pin and ground close to cero) i cant figure out why.

I'm guessing that some current is flowing from other parts of the circuit and is keeping the transistor barely turned on. Possibly through the upper LED? I'm guessing that the leakage current on large LEDs is significant. A quick review of the Cree XP-E2 datasheet doesn't reveal even a mention of reverse leakage current.

Hope that helps!

 

[zerodish]

Re: Will this circuit work?

I had a tire running generator that would produce 40 volts and 4 amps at 40mph. I dumped this into 4AA batteries with a bridge rectifier with a 4.8 volt flashlight bulb in parallel. The bulb didn't pop the batteries made a good voltage regulator. The problem with that set up is I had to isolate the generator from the frame with rubber. The generator used the frame as a wire. I was never able to clamp the generator tight enough to get it to stay on the tire. It takes around 12 volts to damage one NiCd battery so there are limits to this circuit.