Standlight and Martin's Circuit 12 help please!

piesoup

Newly Enlightened
Joined
Oct 31, 2008
Messages
114
Location
East Anglia
Hello

I have read a lot of helpful info on here but am getting stuck in one department.
I usually make battery powered lights using taskled drivers, but want to make a system for a SP dynamo

I like Martin's circuit 12, with the voltage doubler and the rectifier and how its automatic. But I would also like to have a few minutes standlight for when I stop and want to set up camp etc.
Can I get a hand in adding this capability please?

Here is circuit 12
http://www.pilom.com/BicycleElectronics/DynamoCircuits.htm#AutoMode

Thanks!
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
I haven't used Martin's circuits, and have only briefly read his explanations. To add standlight capability, the quickest thing to do would be to increase C5 to something much larger. My only concern is that this might affect the tuning that lets the circuit extract more power from the dynamo at low speeds. You could certainly try a larger cap and see what happens.... the worst thing that will happen is that you get less power at low speeds, and you can always revert back to the recommended value if you don't like the large cap.

I've played around with making my own dynamo lights with standlights. Most used a AA nicad battery for the standlight, and it could run for 30 minutes or so. The latest uses a 100F, 2.5v supercap, and it runs for 10 minutes. The light is better than the average standlight, but it's not a nice bright flashlight/torch.
Here's the info on it, in case it is of interest...
https://www.flickr.com/photos/kurtsj00/sets/72157649424813991/
 

Edocaster

Newly Enlightened
Joined
Mar 27, 2014
Messages
51
If you look at http://www.candlepowerforums.com/vb...rcuit-work-)&p=2706860&viewfull=1#post2706860 the right-hand side of that circuit gives a nice, easy 'be seen' standlight with simple components. On the Pilom circuit 12 diagram you linked to, you can apply this across LD1 and LD2, or LD2 and LD3.

When I tried this (for a manual switching circuit), I used 2x1.5F capacitors in parallel and lowered the current-limiting resistor to 68 ohms. It's still only a 'be seen' light though - there's a sweet spot for efficiency with this simple circuit, with little to gain by trying to increase standlight current too much.

Note: Because the voltage across the LEDs will differ in voltage doubler and full-wave rectifier mode, you will need to put/pick sufficient diodes to clamp the voltage across the goldcaps at a 'worst case scenario' (i.e. either read off the forward voltage in the LED datasheet for the highest current you think you can get in full-wave rectifier mode, or measure it, and then look at the maximum voltage of the goldcap, which is probably 5.5V).

Let us know how it goes!
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
I wasn't even thinking about that circuit and thread... there's a lot of good material there.
 

znomit

Enlightened
Joined
Aug 1, 2007
Messages
979
Location
New Zealand
Circuit 12 is quite involved. You only really need something like that for MTB riding where you might have to slow down/stop from speed suddenly for something technical. It was good for that(used it for a few years until a swim killed the hub bearings). The manual switching circuit isn't a bother usually on the road unless you're using a large number of LEDs.

What about a hybrid dyno/drive setup. Charge an 18560 to run as a boost or stand light?
 

piesoup

Newly Enlightened
Joined
Oct 31, 2008
Messages
114
Location
East Anglia
Hi everyone, thank you for the detailed replies!
I didn't mention, I will be using it mainly on a mtn bike. Now I know it'll work better on that bike, I can make another circuit for my road bike at alter date.

I like that approach edocaster, nice and simple. Probably my ignorance (I'm a mech eng!) but when the supercap is charging, would this not take power from the LEDs. I know it won't be for long, depending on speed, but will it be noticeable?

Steve, I can certainly try a larger cap for C5, what's the worst that could happen!? Your circuit is nice and simple, I'll have a go a it, thanks.


Zomnit, just what I was thinking! I like it as I understand all the goings on. Inside the light is the Dynamo circuit, driving 4 LEDs for example. Then on a switch, is a taskled driver, powered from the battery, driving either one of the four LEDs, or an additional one.

I can easily rig the switches so the taskled cannot turn on if the dynamo is powering the light. And ideally, I'd like a short, simple stand light for mountain biking. Don't want to find I'm hesitating on a techy bit, only to have no light.
I hope the diagram below makes sense!

Thanks again for the help, I've made more headway from this one post than days of trawling the net!
16115676330_783008f412_c.jpg
 

Edocaster

Newly Enlightened
Joined
Mar 27, 2014
Messages
51
When the supercap is charging it will take current from away from the light, but the exact amount will depend on the supercap.

Most people using this setup use 'memory backup' coin caps (as they are small, relatively cheap, and one of the more available caps at 5.5V - and it's what almost all the commercial dynamo lights use). These actually have a fairly high internal resistance I think (they tend to only quote equivalent series resistance on the spec sheets though - and I don't know how this applies exactly - but expect to see figures of 10 to 30 ohms).

The benefit of this is that the charge rates will be limited, as there will always be an alternative path for the rest of the 500-600mA of the dynamo to go through - typically via the current-limiting resistor (and, eventually when the voltage is high enough, via the main LEDs path).

Of the two lights I have built using this standlight circuit, one of the two LEDs spanning the standlight will light up immediately, and the other will light up at a higher speed, until the capacitor is near fully charged.

BTW - your diagram looks very impressive and very ambitious! It looks like you know exactly what you are talking about though, so I'm sure you can get this to work. The bit in the red box is a major project by itself, which many people have had trouble cracking reliably. The pink cuboid looks like a Traco Power switching regulator. Plenty to read about getting switching regulators to behave with bike dynamos here: http://pages.citebite.com/d2f2p1r7t4dcn and http://singletrackworld.com/forum/topic/tell-me-about-dynamo-lights (DoctorRad on the second link helped me on another forum to get the overvoltage protection sorted with a slightly more involved solution than the one written about on STW - more details if you're interested.)

For what it's worth, I would probably separate the two functions (light with standlight - and USB charger with cache battery) entirely, although I can't quite put my finger on why.
 

piesoup

Newly Enlightened
Joined
Oct 31, 2008
Messages
114
Location
East Anglia
Kerry at light makes/assembles a solution for this. The trick is to use a cache battery that allows itself to be used at the same time as its being charged.
http://www.klite.com.au/#!love-your-pet/c17i2

This is exactly what I would like. But at the mo, using the sparkfun / adafruit lipo charger, it will only charge the battery at 500mA, if I'm going fast enough to generate 500mA! For a light driven from a single Li-in, I'd want to run the led at at least 1A. So eventually, I'd run out. I can use this for bike packing, as I cant see me riding for all that long at night, so the battery could charge the next day. But for 24hr racing, I'd want a fool proof system where the battery wouldn't run out. Do you think I'm asking for the moon on a stick? I cant understand why it hasn't been home built before.

Edocaster, Thanks for that link, very interesting. Particularly the protection circuits. I haven't yet settled on a 5V reg but I'll have a look at their spec sheets and choose the best one. Inside the 'red box' is just off the shelf components, from adafruit and sparkfun. They work in the 500mA to 1A region, which I think is fine as you aren't going to get any more out of the dyno.

So far..
Protection is sorted with an AC zener with a suitable voltage to protect the regulator. The vener is coupled to a thermoswitch with opens the circuit when it warms up, protecting it further.
Then to the 5v switching regulator.
Then to this https://learn.adafruit.com/adafruit-powerboost-500-plus-charger/overview . That will provide power to my gps and cache battery. Charge at 500mA, can discharge at 2.4A MAX I think.
Cache battery can charge any USB device once stopped too. (via the powerboost)

All that can be turn off, so the dynamo powers the light.

My main problem area is still the standlight. I'm notdead set on using Martin's circuit 12, I just want one what will work well off road AND have a standlight.
Those coin caps look good, nice and small. Which circuits did you use Edocaster? The one at post 70 or post 89? I'd prefer the one with the caps TBH. I understand that the higher resistance will limit the charge rate. But bearing that in mind, if you have a stop start section where you are relying on the stand light, it could potentially run out. And charging will be at the expense of lumens once you get going again.
I'm still reading through the big thread you linked to on CPF.

Thanks again for your help guys
 

find_bruce

Newly Enlightened
Joined
May 5, 2011
Messages
84
My understanding is that people are not using li-ion batteries for as cache batteries because the chemistry is not suited to small repeated charging - eg Steve uses Nicads for his standlights.

I use a similar circuit based on post 70 - have never been able to detect any drop off in lumens while charging the supercaps - the main drop off I notice is my lack of climbing ability. If you find your standlight is running out, either (1) use a bigger capacitor, I found some 5 farad ones on fleabay, sadly slightly to large to fit in the steerer tube or (2) put 2 or more capacitors in parallel - eg use 2 x 1.5F to give you 3F
 

Edocaster

Newly Enlightened
Joined
Mar 27, 2014
Messages
51
Those coin caps look good, nice and small. Which circuits did you use Edocaster? The one at post 70 or post 89? I'd prefer the one with the caps TBH. I understand that the higher resistance will limit the charge rate. But bearing that in mind, if you have a stop start section where you are relying on the stand light, it could potentially run out. And charging will be at the expense of lumens once you get going again.
I'm still reading through the big thread you linked to on CPF.

I used the circuit at post 70. To be honest, I don't think it will be enough for you, as you've mentioned you want up to an amp. The basic standlight circuit might get 15 to 50mA only. So lumens barely into two figures. Still the basic standlight is good as a baseline - i.e. you can incorporate it as a secondary system, which ensures you get a few minutes of be-seen light. There's nothing to stop you from making a non-coin cap main standlight (using a 'big' supercap with low resistance and high current handling) which might only last 15 seconds or so - enough to get over a technical obstacle. But charging then becomes a little more involved.

If you go down the li-ion route, there's an argument to not hybridise Martin's system (which is essentially powering the light directly from the rectified dynamo current, with no management of current) with a voltage regulated USB/li-ion system, but instead to stick the light on the output of the USB/li-ion system. You could use a circuit like the 'Forumslader' system: http://www.forumslader.de/Schaltplan.212.0.html - which is essentially like the Martin switching circuit, but feeding voltage regulators. The downside is that there are losses from the multiple regulators and battery inputs/outputs, and you risk either emptying the cache battery if the load presented by the light is too high, or overvolting the system if the load is not enough.
 

znomit

Enlightened
Joined
Aug 1, 2007
Messages
979
Location
New Zealand
The supercap is definitely for a be seen light only. I used them a few times but a better solution is a standalone blinky or helmet light.
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
My understanding is that people are not using li-ion batteries for as cache batteries because the chemistry is not suited to small repeated charging - eg Steve uses Nicads for his standlights.

to add some context to why I chose nicads... part of it is that I don't have experience with li-ions.
My bike commuting happens through all weather and temperatures. As such, the battery needs to operate happily even when the temperature is around 0F (-18C). Nicads will have a higher resistance, but they still work. My recollection is that li-ion doesn't like getting charged when it is that cold.
I should also mention that I use voltage regulated charge on the nicads in my standlights. Setting the charge voltage slightly low allows for more gentle charging of the battery, leading to a longer life. With a AA nicad, I still got a very long run time, and I've seen battery life of around 10 years (in one light, at least).

My latest design uses a 100F supercap, just because I got it for a good price and it seemed like an interesting change. The run time is "only" 10 minutes, but it's not a problem. One interesting "lesson learned" is that a 2.5V supercap can still forward bias a schottky diode and a Cree white LED, even if it is only with a few mA. I never had that issue with a 1.25V nicad! :)
 

piesoup

Newly Enlightened
Joined
Oct 31, 2008
Messages
114
Location
East Anglia
Gee, didnt realise my post was that long!

Bruce, from the post on MTBR about the protection
I was chatting with Jens from forumslader, he advised an AC diode, 1.5KE56CA to be placed before the rectifier. Then a thermal switch with about 60 degree C. There are loads of thermal switches available.

I have gone off Martin's circuit 12. I will have to spend over £50 for the components as you cant buy one of each! I'd have 100s of diodes left over!

Edocaster, when you say not to hydridise the dyno circuit and the USB circuit, do you mean like in my diagram above? As it stands, they are switched between, only one on at a time. Or do you mean constantly connected. I did think about hooking the light up to the USB output, but then I'll be limited on the power surely? Having a read of the forumslader circuit again, no harm in trying it if I can get the bits.

Znomit, I think that's the way I'll go for a reliable 'standlight'. I already have a bar mounted switch for my battery light. It wont be used much so could charge in the day and be used for standlight duties when the dyno isnt producing enough. I've got used to turning it on and off for techy sections in 24hr races to supplement my main light!


Steve, Yeah, Ive tried to charge li ions at 3am, it didnt happen. That race was finished with me in my sleeping bag! Good to know about what a 100F cap can do, hope it doesnt happen to me!
 

Edocaster

Newly Enlightened
Joined
Mar 27, 2014
Messages
51
Edocaster, when you say not to hydridise the dyno circuit and the USB circuit, do you mean like in my diagram above? As it stands, they are switched between, only one on at a time. Or do you mean constantly connected. I did think about hooking the light up to the USB output, but then I'll be limited on the power surely? Having a read of the forumslader circuit again, no harm in trying it if I can get the bits.

I might have read your diagram wrong, but if you switch between either the main light or the USB/li-ion system, and the standlight runs off the USB/li-ion system, then the standlight relies on something outside of the light. I.e. you would need to switch over to the USB/li-ion system to charge the standlight.

You'd also need to find a way of getting the li-ion to power the standlight only when the dynamo stops powering the light directly, which seems like a lot of hassle trying to 'inject' the standlight in from one system to the other. You might as well have the li-ion powering a separate light.

If you were to run something off the regulated USB output instead of direct from the rectified dynamo output, the power is dependent on the design. The forumslader guys apparently claim over 1A at USB voltage from their circuit (which I believe is dependent on having the first regulator set at over 12V to get more power from the dynamo). You might be able to skip the USB stage too, if your light is happy with whatever voltage the forumslader is set to at the first stage.
 

piesoup

Newly Enlightened
Joined
Oct 31, 2008
Messages
114
Location
East Anglia
Yes, you're right. There would need to be some funky switching involved for everything to be in the right place! Though about putting an arduino in there to take care of it all. Again, too much hassle just to get a light to turn on. I need this to be as simple as can be. If I had time, I could program the arduino to sense the speed, select the voltage doubler or the bridge rectifier, light the standlight if I slow down too much, manage the charging of the caps etc. But that will require a thread longer than this one!

But I'm interested in using the 12v regulator. Load up the dynamo at all times to eek out as much current as possible. Going to look into it more, did wonder why there was a 12v reg in his circuit.
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
....

But I'm interested in using the 12v regulator. Load up the dynamo at all times to eek out as much current as possible. Going to look into it more, did wonder why there was a 12v reg in his circuit.

I may have missed some of the conversation... as far as getting as much current as possible, I'll assume that you are really desiring to extract as much power as possible from the dynamo. To a large degree, this is done by going as fast as possible. :)

For those times when you are going a bit slower, there is the challenge of matching the load to the dynamo. This is what Martin did with his series capacitance, although the shortcoming is that it is effective over a narrow range of speeds.
I built one dynamo light that tried to maximize power by switching between a load that was two LEDs in series, and a load that was 4 LEDs in series (for speeds above a certain threshold). The speed detection was done by a dual one-shot IC.... although a little microcontroller would be a more modern method. Pictures and schematics and scope shots here....
https://www.flickr.com/photos/kurtsj00/sets/72157621965148305/

When you complicate matters by combining battery charge management, load management, and peak power tracking (a.k.a. maximum power point), then it gets trickier. Many years ago, I did a design of this sort and it was fun. It was also a lot of work.
IIRC, Linear Technology makes a chip that does a bit of this. Actually, a search for "maximum power point" on the linear.com site returns a bunch of ICs. The LTC4000 is probably the one I was thinking of... it handles input voltages of 3v to 60v. At high speed, I've gotten at least 100v out of my hub dynamo, so you'd have to understand and characterize the parts you are working with.

Anyway... just tossing those ideas out there. :)
 

Edocaster

Newly Enlightened
Joined
Mar 27, 2014
Messages
51
But I'm interested in using the 12v regulator. Load up the dynamo at all times to eek out as much current as possible. Going to look into it more, did wonder why there was a 12v reg in his circuit.

Looking at the forumslader link above, it looks like there are two circuit diagrams, one with a normal forward-biased diode on the ground pin of the 12V regulator, the other with a zener. I think the idea is to regulate to match either 3 li-ion cells in series, or 4.

With the normal diode, the output should be around 12.6V, which would equate to about 4.2V per cell. I'm not sure how risky or crude such a charging method is (I'm not sure if balancing is covered), but at least the charging rate shouldn't exceed 0.5A.

If you ignore the USB part of the circuit for now, and put on the output ('EX1'), and hence parallel to where the li-ion cells will go ('AKKU'), a constant current LED driver, you should get a light which behaves as if it's connected to a normal battery. As long as the driver can take the voltage with some safety margin, you should be able to use any driver you want, at any current setting the batteries will take (although they may drain faster than they can charge if you pick too high a current).

Anyway, that's my guess. If you build it, do let us know!
 

piesoup

Newly Enlightened
Joined
Oct 31, 2008
Messages
114
Location
East Anglia
I may have missed some of the conversation... as far as getting as much current as possible, I'll assume that you are really desiring to extract as much power as possible from the dynamo. To a large degree, this is done by going as fast as possible. :)

For those times when you are going a bit slower, there is the challenge of matching the load to the dynamo. This is what Martin did with his series capacitance, although the shortcoming is that it is effective over a narrow range of speeds.
I built one dynamo light that tried to maximize power by switching between a load that was two LEDs in series, and a load that was 4 LEDs in series (for speeds above a certain threshold). The speed detection was done by a dual one-shot IC.... although a little microcontroller would be a more modern method. Pictures and schematics and scope shots here....
https://www.flickr.com/photos/kurtsj00/sets/72157621965148305/

Anyway... just tossing those ideas out there. :)

Yeah, going fast is the easy bit! I did think about using a microcontroller, using a voltage divider to sense the voltage and therefore the speed. Could use a FET to turn on the extrab LEDs couldn't I? With a IC out could get real geeky and have it adjustable too!


This is great, thanks Steve. I won't be building this immediately, just want to do as much research as possible first. And I should be studying for my aircraft engineer degree, not tinkering in my workshop!
 

piesoup

Newly Enlightened
Joined
Oct 31, 2008
Messages
114
Location
East Anglia
If you ignore the USB part of the circuit for now, and put on the output ('EX1'), and hence parallel to where the li-ion cells will go ('AKKU'), a constant current LED driver, you should get a light which behaves as if it's connected to a normal battery. As long as the driver can take the voltage with some safety margin, you should be able to use any driver you want, at any current setting the batteries will take (although they may drain faster than they can charge if you pick too high a current).

Anyway, that's my guess. If you build it, do let us know!

Yeah, I was a bit concerned about the Li Ion charging. I'll need to have another look at it.
This is almost exactly what I'm after. Using the batteries as a buffer.
I've come to realise, super caps or batteries, I'm still going to have to manage the standlight either manually or with a microprocessor. Nothing an arduino can't handle.
So like Steve mentioned above, I could use a driver to light up the powered by the batteries, with the dynamo charging the batteries.
It will be a little while before I get round to building this. Supposed to be studying, and I have another project involving servos, an arduino and the compression settings on my bike!
 
Top