Simple Standlight - Passive Components

kg6gfq

Newly Enlightened
Joined
Nov 9, 2015
Messages
14
After trudging through much of the old standlight thread, with its various lines of development and numerous missing images, it was a great relief to find the more recent Dynamo lights for Dummies, even if I was out of my depth after the first few posts.

Given those resources, I put together the following on a breadboard, which has been hanging off my rear rack, plugged into the taillight output of an old B&M IQ Fly running off a Sturmey-Archer X-FDD for a couple of days:
simple_standlight.png


It works reasonably well, I was able to put it together with parts I had on hand, and I [think that I] mostly understand what it's doing.

The LEDs are all generic "superbright" 5mm red LEDs.

Now that it's working, I'd like to upgrade the LED(s) across the standlight to something brighter. After browsing www.ledsupply.com, I'm thinking a Cree XP-E2 with maybe a Carclo 10140 frosted wide-angle lens, but I would love to get suggestions from folks with more experience selecting and buying LEDs.

I realize this will require changing the upper LED (D3), and I'll probably also get a higher voltage Zener (D1) and change D2, the clamping diode on the capacitor, accordingly. I figure D3, the non-standlight LED, will probably end up being 2 red 5mm LEDs angled out for side visibility.

Anyway, I have three questions:
1) Does this circuit look viable, or is it just biding its time, waiting to blow up on me?
2) Is there another LED/optic that would be brighter or give a better beam pattern?
3) While I'm asking dynamo-related questions, is there any LED/optic combo available for a headlight that will yield a beam pattern similar to what I've become accustomed to with B&M lights?

Thanks!

-Darin
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
The circuit seems workable... it looks like it is a tweaked version of what came out of the standlights thread.

The only questions that come to mind are:

1. what is the zener for? D3 and D4//D5 should be clamping the voltage.

2. Is D5 wired in parallel with D4 for optical reasons or electrical reasons? If the LEDs are rated for 20mA, then I would understand. Most taillights are designed to draw 60mA, so it would make sense to use LEDs rated for at least 60mA.

Not sure what to tell you about optics. It's really a matter of what you want. I've gone for a rather tight beam, but my riding is primarily on straight, rural roads. Savvas posted a link (in his thread) to a neat optic that spreads the light out in azimuth (left to right), but keeps the light very concentrated vertically.

I don't know of an optic that would provide a proper headlight beam that conforms to the German standards like the B&M lights do. This seems to be beyond the ability of a lens, and seems to require a suitably designed reflector. My goal is to find a dead B&M headlight and put my own electronics into it. :)
 

Savvas

Enlightened
Joined
Jun 11, 2010
Messages
222
Darin,

I would echo Steve's query about the zener (D1). Can't quite see why you need it. I'd just use whatever red leds you can get. I understand that XP-Es may be more efficient but the cheap red leds on stars from DX also work OK too.

As Steve indicated, Ledil make a 'beacon' optic which Cutter sell quite cheaply. I think 1 What (or maybe someone else) on the old 'designing a tail light' thread may have used it - pictures now gone unfortunately. I am building a battery powered light to try it out, but there's no reason not to experiment with a dyno light either. Ledil and others also make conical TIR optics that have directional diffusers that spread the light laterally. Have a look at the Cutter web site or maybe one of the other suppliers.

I don't think you're going to blow anything up! ;-)

Savvas
 

kg6gfq

Newly Enlightened
Joined
Nov 9, 2015
Messages
14
Thanks for the comments!

The zener came from this circuit, which I was working with previously. It's good to know that I don't need it.

Steve - D4 and D5 are in parallel for optical reasons. A single 5mm LED seemed like an insufficient standlight. As for the front, I might try dismantling an old IQ Fly and fitting a newer LED into it.

Those LEDiL flare optics look like they could be pretty great. I'm going to order some from Mouser to go with a couple of red Cree XP-E2s from LEDsupply. I'll report back when I have a chance to try them out.

-Darin
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
Thanks for the comments!

The zener came from this circuit, which I was working with previously. It's good to know that I don't need it.

-Darin

that circuit is a simple current regulator, and the zener is protecting the supercap.
In your proposed circuit, the LEDs are clamping the voltage across the supercap, more or less.

going back to the circuit that you linked to... I'm a little surprised to see that the LED current is limited to approx. 13mA. That'll keep it running for a while as a standlight, but it'll make a terrible main taillight. I hope this was to be used in addition to a regular taillight.
 

Savvas

Enlightened
Joined
Jun 11, 2010
Messages
222
Darin,

I think I actually saved all of the circuit diagrams from the Great Standlight Thread somewhere, as well as a lot of the discussion which seemed most relevant. My aim (apart from trying to get my head around the simple electronics involved) was to put together a 'stand light manual' for my own benefit and continued reference. I'm probably just a compulsive 'gatherer'!

IMHO, simpler is always best. For instance, I note that the really excellent Philips Saferide Lumiring tail light only uses 1 or 2 standard low power LEDs, probably in a circuit very much like the penultimate Standlight Thread designs. This light is very, very effective. Bright 'n big! The trick is obviously to make maximum use of the limited light available. Philips and several other very similar lights use clever plastic lenses that create a relatively large 'lit area' & far more than is evident with just 1 or 2 isolated 5mm LEDs.

Few (if any) of these commercial rear dynamo lights use focussed TIR or reflector lenses, often cited as essential if aiming to get the attention of drivers at any distance. I assume that most of the European dynamo light makers would say that this is what reflecting materials are for! I've actually had people remark that my Philips Lumiring light as well as the B&M Topline (which uses a 'light bar' rather than a light-conducting ring) are both plenty bright enough from a distance. Both use relatively simple 'supercap' circuits, again not very different from those that evolved in the Standlight Thread. For example, here's the schematic for the old B&M Toplight (see post 139). I'm guessing that most of the components (the resistors and diodes) are probably there to balance impedance with the front light to prevent it drawing too much juice.

I'd be interested to know how bright your tail light is, especially without the series resistor. If we can assume that the dynamo generates about 500ma and our lighting system runs at 6v and 'soaks up' (for want of a better word) all of this 500ma, then it's running at 3 watts (I think).

I've read somewhere that one should aim to produce about 0.6 W with the rear light, leaving about 2.4w for the front. With the dynamo nominally producing 6 volts, then it seems to me that up to 3 standard red LEDs (vf = 2.2V, 20ma) in series could be used in up to 5 parallel strings (15 LEDs in total) to run the tail light at the required 100ma (100ma x 6 volts = 600mW). Am I thinking straight on this?

I've never tried this but it seems a worthwhile venture! I'm building a little jig using an old sowing machine motor to try and do some serious experimenting!

Sam.
 
Last edited:

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
.....
I've read somewhere that one should aim to produce about 0.6 W with the rear light, leaving about 2.4w for the front. With the dynamo nominally producing 6 volts, then it seems to me that up to 3 standard red LEDs (vf = 2.2V, 20ma) in series could be used in up to 5 parallel strings (15 LEDs in total) to run the tail light at the required 100ma (100ma x 6 volts = 600mW). Am I thinking straight on this?

I've never tried this but it seems a worthwhile venture! I'm building a little jig using an old sowing machine motor to try and do some serious experimenting!

Sam.


hi Sam,
The standard, from back in the incandescent days, was to have a 2.4W bulb in front and a 0.6W bulb in the back. Whether or not this is still appropriate depends on the headlight design. It's not a bad place to start.

I'm not sure about getting 15 LEDs powered from 6v. Again, it was simpler back in the incandescent days. :)
You'll need/want some sort of current regulation or limiting, so that will likely eat into some of the available voltage. I think you'll do well to get two red LEDs wired in series. Running 5 parallel strings seems reasonable, though. I've got ten 5mm LEDs for my taillight, and it is as bright as I could imagine needing.
Running two larger red LEDs in series would be fine too, and give you the option of using TIR optics. Perhaps use the "beacon" optic on one LED and a fairly tight conical beam optic on the second LED?
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
Gee I miss the Great Standlight Thread.... ;-)

Sam

Is there a bike electronics group somewhere else that is active? Bike Current still exists in Google Groups, I think, but it's been many years since there was activity. I think the MTBR DIY lights forum is pretty slow nowadays too.

I still have a few bike light projects in mind, so I should be posting each of them. One is a "peak power tracker", or max power point tracker, idea for hub dynamos. I need to brush up on my C programming and learn about the little Atmel processors first. ..maybe run a few spice simulations too... but I think it will be fun. It's probably too specialized to draw in a big crowd like the standlight thread did.
 

kg6gfq

Newly Enlightened
Joined
Nov 9, 2015
Messages
14
Just ordered LEDs, optics, and other components for a taillight and a headlight with extras left over. I'm looking forward to further experiments!

I think I actually saved all of the circuit diagrams from the Great Standlight Thread somewhere, as well as a lot of the discussion which seemed most relevant. My aim (apart from trying to get my head around the simple electronics involved) was to put together a 'stand light manual' for my own benefit and continued reference.
Wow - I'd love to see that! Forum threads aren't necessarily the easiest format to follow, and the early part of the Great Standlight Thread is especially tricky with all the missing schematics.

I was thinking about writing up everything I've learned about dynamo lights and posting it to my own site, as a reference for others. I'm definitely not the best person for the job, but I still think it could be useful.

I'd be interested to know how bright your tail light is, especially without the series resistor. If we can assume that the dynamo generates about 500ma and our lighting system runs at 6v and 'soaks up' (for want of a better word) all of this 500ma, then it's running at 3 watts (I think).
I doubt it's running at 3W. It's hooked up to taillight output of a B&M headlight, which I believe just puts it in parallel. My (rudimentary) understanding of parallel circuits leads me to believe that they should both be operating in the vicinity of 6V, with the current through each inversely proportional to their resistance (or rather impedance, since it's AC), so I doubt the taillight is getting a full 500mA. Maybe I'll tape a multimeter to my handlebars and take some measurements when we get a break in the rain here.

As for apparent brightness... I haven't had much opportunity to look at it on the road, since it's always behind me. That said, when I glance over my shoulder it seems to throw a fair bit of light on the road and nearby objects, maybe comparable to a Planet Bike Superflash in steady-mode? I don't know what it looks like in the dark from behind, though, because the only time I've had a clear view of it was on the bench in a well-lit workshop, hooked up to a DC power supply.

One is a "peak power tracker", or max power point tracker, idea for hub dynamos.
Sounds interesting! Probably better than the multimeter-on-handlebars approach. (-:

After revisiting some of the posts on p.5 of the old thread, particularly Steve's comments in post #142 on the B&M standlight circuit, I begin to understand the advantages of running lights in series - you don't have to worry about relative impedance of the two lights, and each light can have a different voltage drop, or even a variable voltage drop.

When designing lights in parallel, it seems you have to use either resistors or active current regulators to get the proper balance of current between front & rear, and both are driven at the same voltage.

Thanks again to both of you for sharing your expertise!

-Darin

P.S.: I noticed the Vishay 15F/5.6V supercap for <$10 on Mouser. Would something like that be suitable for an extra-long standlight?
 

Savvas

Enlightened
Joined
Jun 11, 2010
Messages
222
Hi Darin,

When I suggested we assume that your lighting system was taking the whole 500ma from your generator I meant both front and rear lights together. Hence my reference to the use of the extra components in the Toplight circuit as probably balancing the two separate loads.

The Eu designs probably stick to parallel loads for good reason:
- redundancy: lose one light and you still have the other assuming over current protection etc)
- compatibility: across most of the lighting manufacturers (including those still producing incandescent headlights)
- economy: no extra capacitors etc.
And the German standards of course!

As soon as you go to series circuits (like Martins and like most of those on the Standlight Thread) you encounter distinct advantages as well as disadvantages. I reckon a market based on series circuits would see much less compatibility and much more confusion simply because of the likely proliferation of circuit variables from one light to the next. Commercial manufacturers like the certainty associated with 'fitting in' with a general standard.

The advantages of series circuits - as near as I can figure - mostly involve being able to run both white and red (front and rear) power leds together in the one system, obtaining higher efficiencies to derive maximum light and to utilise highly refined TIR and reflector lenses. As far as I can see, the disadvantages of the series approach are really only evident if we are trying to market our dynamo lights in agreement with the current Eu standards!

With series circuits you can actually start extracting more power out of the dynamo system than is implied by the nominal 3w rating by using resonance circuits and other bits of cleverness. And because the dynamo behaves (I think) like a 'current limiting device' you can keep adding LEDs in series to raise the voltage load - if you can pedal sufficiently fast! I'ce actually made a 6-LED series light using XR-Es - but being a pootler I couldn't pedal fast enough to really realise it's advantages.

I think most people use either the Elna of Japanese 1F or 1.5F, 5.5v super caps (they're what most of the Eu manufacturers seem to use). They're cheap, small, readily available and do a good job. It seems you can also use the 2.5V rated 1.5F 'coin' caps if you are careful. The 15F super cap you mention looks great, especially with regard to size. However it also seems a bit expensive and its capacity may be unnecessary. The 1F stand lights work quite well. However wouldn't it be interesting to try 15F! Although there's probably some disadvantage associated with internal resistance or maybe charge time etc.

You've got me all enthused again - THANKS!

Savvas.

ps; I'll try to put what I saved from the S/L thread together into a pdf sometime.
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
After revisiting some of the posts on p.5 of the old thread, particularly Steve's comments in post #142 on the B&M standlight circuit, I begin to understand the advantages of running lights in series - you don't have to worry about relative impedance of the two lights, and each light can have a different voltage drop, or even a variable voltage drop.

My advice was valid when LED lights were still trying to "look" like a 12 ohm incandescent bulb. I'm not sure that they haven't changed... you'd have to measure the relationship between their input voltage and input current to be sure. Or at least confirm that they pass 0.5A under normal conditions. I don't own a commercially built dynamo light, so I can't check it myself.

If you design your own headlight and taillight, then you can set things up in a way that optimizes the performance for both.
 

kg6gfq

Newly Enlightened
Joined
Nov 9, 2015
Messages
14
Well, I'm back from visiting family and the components have arrived. The red Cree XP-E2 with a LEDIL "Flare" optic looks promising on the bench. I am, however, having some circuit design issues. The circuit currently looks like this:
current.png


I'm testing with a DC power supply, since it's handy, using 6V and up to 0.5A. The rectifier seems to eat between 0.5V and 1V with only the voltmeter for a load. With the rest of the circuit hooked up, the voltage across the + and - pins of the rectifier is 4.5V - I'm not clear on why this is the case. According to the power supply, it's drawing about 0.024A. That leaves about 3.5V after R1, which isn't enough for D2 and D3.

I've tried decreasing R1's resistance by adding resistors in parallel and even bypassing it entirely, but it looks like D2 tends to get slightly more voltage than D3 (e.g. 2.1V and 2.0V respectively). Could this be because D2 (a generic "superbright" 5mm red LED) is probably rated for less current than D3? Will any of this change when it's run from the dynohub?

I'll continue to research and experiment, but right now it's a little beyond my comprehension...

Thanks!

-Darin
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
dropping 1.5V across two silicon diodes is not that odd. It seems a bit high when only drawing 24mA, though. How big or small is the bridge rectifier?

Are the LEDs lighting up? If so, which one(s) and how much?

as a rule of thumb, it helps to measure all of the voltages in the circuit. This helps clarify which is/are wrong, and therefore where to start looking for stuff that's not right. I would recommend measuring them in reference to pin 4 of the bridge rectifier; i.e. that would be considered ground for the sake of looking at the voltages.
As the voltages are measured, write them on the schematic, and when completed, post this to the thread.
 

Savvas

Enlightened
Joined
Jun 11, 2010
Messages
222
I agree with Steve - if you have a workbench where you can take measurements, do a thorough survey of voltages across your circuit. I would not expect to find 4.5v at the back of the rectifier - are you sure you are supplying 6v? If you can, I'd try building this using dyno output from the start. Maybe put your bike in a stand close to where you are working. Bear in mind that the dynamo is a current source and will supply whatever voltage is required to pass the current demanded by your circuit (at least that's my rudimentary understanding of how it works).

Regarding your current circuit, I suggest you get rid of R1 - IMHO it's quite unnecessary and a drag on your circuit. The voltage drop across all of the LEDs is an important number as is the voltage across the super cap (needs to be as close as you can get to 5.5v). If you are continuing to run your IQ Fly, I'd also maybe stick to standard T (5mm) red LEDs at the rear. You could use 3 in series to get close to 6v voltage drop and just use the cap on the last one, adjusting the v/drop with D1. If you have decent LEDs this will make a pretty bright tail light! You could look at the high-brightness or superflux LEDs from LED-DNA - I've found that they work very well.

I was riding behind a friend last week. He was running a Shimano hub and a B&M Luxos combined with the B&M Topline, which uses 2 standard 'T' LEDs behind a fresnel optic. I was very impressed with the visibility and intense brightness of just these 2 little LEDs. For a dynamo tail light you obviously don't need a 1W or 3W powerLED at the rear - although it's obviously good to experiment if you have the current available!

Savvas.
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
.......
Regarding your current circuit, I suggest you get rid of R1 - IMHO it's quite unnecessary and a drag on your circuit. .....

Savvas.

If this was a headlight, I would agree. IIRC, this is a taillight, so R1 is needed to limit the current.

BTW, that reminds me that the current was only 24mA. With the circuit wired as shown, it should have been at least 80mA. Something is not right.
 

kg6gfq

Newly Enlightened
Joined
Nov 9, 2015
Messages
14
Ok... I've done a variety of tests. First, the circuit as shown before, with a 6V DC power supply feeding the rectifier. I gave the capacitor a few minutes to charge before measuring anything. According to the power supply, the circuit was carrying a current of about 0.025A.

current_withpoints.png


After experimenting for a while, I made several improvements. I switched from the RS101 back to a rectifier made with 1N5818s and replaced D1 with an SB330 that has a much lower voltage drop. I also adjusted the values of R1 and R2, and I suspect I can fine-tune those values to drive the LEDs with the correct voltages by doing some testing with potentiometers. What I currently have is:

simple_standlight_update2016.12.14.png


Again, voltages are measured with a 6V DC power supply in place of the dynamo. I forgot to note the current from the power supply, but based on the voltage drop across R1 it appears to be 0.070A.

I also taped a multimeter to my handlebars and took some readings while riding. Unfortunately, I don't have a cyclecomputer so I can't correlate these with speed.

Voltage @:
A: 5V - 7.15V
B: 3.6V - 3.95V
D: 1.9V - 1.99V

I also measured the voltage across R1 as 2-3V, so the current ranged from 0.085A to 0.128A.

I discovered Horowitz & Hill's The Art of Electronics, and have been reading through the first chapter with the goal of understanding more clearly how this circuit behaves - especially how to determine its apparent resistance at a given voltage or current. So far I gather it's a matter of finding the dynamic resistance spec. for the various diodes and applying Ohm's law.

Anyway, that's the current state of the project. Advice/corrections would be welcome! Hopefully I'll have time to work out the correct resistor values this week so I can move it to a protoboard and figure out a housing.

While I have a voltmeter on my handlebars I'll also test the B&M headlight alone.
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
Ok... I've done a variety of tests. First, the circuit as shown before, with a 6V DC power supply feeding the rectifier. I gave the capacitor a few minutes to charge before measuring anything. According to the power supply, the circuit was carrying a current of about 0.025A.

current_withpoints.png

The numbers for this circuit seem to work out, mostly. Based on the voltage across R1, there is 21.3mA going into the circuit.
Based on the numbers for R2, there is 19.8mA through R2.
This means that there is only 1.5mA (at most) going through D2. This doesn't seem right, I think, because the voltage across D2 should be less than the rated Vf. Is there a high resistance connection in series with D2? A bad solder connection, maybe?

D3 is quite a bit below its Vf too, but it's a large LED and it's only getting 21.3mA (or less), so this isn't much of a surprise.

D1 is marked as being a 1N4733, which should be a 5.1v Zener diode. However, it isn't being used as a zener diode. Why not?

simple_standlight_update2016.12.14.png


Again, voltages are measured with a 6V DC power supply in place of the dynamo. I forgot to note the current from the power supply, but based on the voltage drop across R1 it appears to be 0.070A.

for this circuit, the voltage across R1 shows that there is 70.2mA going through it, which sounds close to what is desired.
The voltage across R2 shows that it has 65mA through it.
This leaves only 5mA (at most) going through D3, even though the voltage across D3 is equal to Vf. Again, this doesn't make sense. There must be some resistance in series with D3.
D4 is getting 65mA to 70mA, and the Vf is roughly what is measured. Seems to be correct.


I discovered Horowitz & Hill's The Art of Electronics, and have been reading through the first chapter with the goal of understanding more clearly how this circuit behaves - especially how to determine its apparent resistance at a given voltage or current. So far I gather it's a matter of finding the dynamic resistance spec. for the various diodes and applying Ohm's law.

Anyway, that's the current state of the project. Advice/corrections would be welcome! Hopefully I'll have time to work out the correct resistor values this week so I can move it to a protoboard and figure out a housing.

While I have a voltmeter on my handlebars I'll also test the B&M headlight alone.

The Art of Electronics is a good book. I'm not sure how much you have to know before you can get benefit from it. Being able to do basic circuit analysis is very helpful.

As far as figuring out how to do circuit analysis with a diode in the circuit, the rule of thumb is to use the value for Vf. If you know that the current will be below the current that Vf is measured at, drop the estimated forward voltage, using the published voltage-current curve for the diode. As one of my teaching assistants told us in college.. "Approximate to graduate". High precision isn't usually required. Just getting into the ballpark is usually good enough.
 

kg6gfq

Newly Enlightened
Joined
Nov 9, 2015
Messages
14
That zener in the first circuit was just what I had sitting around at that moment with a Vf near what I thought I needed - turns out I didn't need it to be that high.

I just noticed that in editing that second schematic I forgot to indicate that D4 is a Cree XP-E2, not the generic 5mm with a measured Vf of 1.8. (Careless editing is also why there's no D1.) The XP-E2 datasheet shows a Vf of about 2.0 at 100mA, so 1.87 @ about 70mA is probably not that far off.

I'm not sure what's up with D3 (D2 in the first circuit). Maybe it's an issue with my breadboard or a quirk of that LED. I'll test it, and also see if I can approximate a voltage-current curve for that LED while I'm tinkering.

Thanks for going through the process of analysing these - now that I've seen it done, I have a much clearer sense of how to do it myself.
 

Savvas

Enlightened
Joined
Jun 11, 2010
Messages
222
Apologies - I forgot that impedance needed to be balanced with the front light. Although I would have thought this might be better done by first establishing the overall voltage drop and current capacity of the LED circuit. Maybe once it's all built up, R1 could be replaced with a pot for some on-road (or on the repair stand) impedance balancing? Maybe that's what the OP is discussing in post 17.

Kg - is D3 a powerLED or just a standard 5mm 'T' (20ma) type? If it as well as D4 are powerLEDs, are you noticing much loss of power to the headlight, particularly at lower 'speed-equivalent' voltages? I reckon this level of testing really needs to be done with a generator source to really appreciate what's happening. I'm following this experiment with interest!

Savvas
 
Top