I'm building an LED array to install inside my motorcycle's stock tail light housing using 1W 10mm LEDs I sourced off of ebay. As to the stock tail/turn configuration, I've got turns in stock pods outboard of the rear fender and am going to leave them that way. The stock tail light housing had one large center lens and two smaller side lenses that were used as turn signals in other countries. However, in this country the side lenses were too close to the center lens to meet the minimum FMVSS distance so for this market the maker left the bulbs and sockets completely out of the housing.
I've already added sockets and bulbs to the side lens housings to use as brake lights only, with 1156 type bulbs. Thus, in tail light mode only the large center lens is lit, and in brake mode the center lens is brighter and the two side lenses light up, accentuating the brake light effect.
My planned final configuration is to have 40 of the 1W LEDs in the center housing as brake and tail light, and 15 in each of the side lenses set up as brake only, for a total of 70 LEDs. I'm going to use simulated 1156 and 1157 bulb bases to connect the LED array boards to their respective sockets so that from the bike's wiring point of view everything's stock.
The simplest setup would be to run two separate LED arrays, one for the tail light and one for the brake light. The LEDs are rated max
2.3V with 2.1 typical, I'll run them at 2V so I would use 5 LEDs per string and use a resistor on each string to limit string current to
250mA (max continuous rating is 300mA). Using an online calculator and the parameters of 11.5V-14.5V input, I will need a 22 Ohm resistor rated for 1W per string, and it will dissipate 920mW as heat, over double what the LED part of the string will. Since I intend to run 70 LEDs that means I'll have 14 strings with 14 resistors, and in total the LED array at full brightness will dissipate 18.6W, of which 12.9W will be as waste heat from the resistors. More that two-thirds of the total power used will be used to heat resistors, and I find this to be particularly crude and wasteful, specially because one of the reasons I'm using LEDs is to reduce alternator loading in order to free up wattage to run my electric heated riding gear.
I want to find a more efficient way to drive these 70 LEDs since the actual power dissipated by the LEDs is only 5.7 watts. Theoretically, some design that was 75% efficient would have a total power consumption of 7.6W, a far better number than the 18.6W used by a simple resistor-limited array. For reference the stock 1157 bulb is rated 26.8/8.3W high/low filaments respectively, for a total of 35.1W.
The LED calculator I used:
formatting link
I've been looking for driver circuit designs to accomplish this but am having no luck. I've found automotive rated chips at Maxim, but apparently don't understand them well enough to figure out how to use them for my application. I see LED arrays on cars all over the place, so I figure this has to be fairly established technology by now.
You should think of LEDs as rated by current, not voltage. The only reason you are concerned with voltage is to insure the driver has enough voltage compliance. [Insure that the current source will work at the maximum expected voltage.]
Given all the variables, it is a good idea to not use a resistor, which of course is why you did the post. The next step up is an analog current source. Search the net for a TL431 based circuit that uses an external transistor to handle the high current. This is only slightly better than a resistor limiter in terms of efficiency, but it more reliable.
A switch mode power supply (SMPS) will be the most efficient, but also the most complicated. It is easy to hack a boost converter to drive LEDs. Generally for a SMPS, the output voltage is divided down by a resistive divider in a manner so that the bottom leg of the resistor is at the bandgap reference voltage when the top of the resistor divider is at the desired voltage. For current regulation, the hack is just to use one resistor from the feedback pin to ground, and calculate it such that at a bandgap voltage it has the desired current. In your case, say the reference was 1.25V. 1.25/.25 = 5 ohms at 312mW. Crappy, but not as bad as a simple resistor limiter. There are parts that use a lower voltage in feedback circuit to get a higher efficiency.
Now I am accustomed to using boost circuits in LED driving since the general goal is not to have a lot of batteries in series. But for your motorcycle, you need the output voltage lower than the input. That might be a harder circuit to find. The goal isn't all that much different than a hysteretic converter used in battery charging. The goal is a switcher that regulates current rather than voltage, and in this case, continuous conduction.
This TI paper covers some common charger circuits that maybe you could hack.
a) Your math above is wrong. If the LEDs drop 10 volts out of 14v and resistors drop the rest, the efficiency is 10/14ths (71%). Power wasted is 4/14ths, not
2/3rds. b) There's no way the 10mm epoxy packages I just saw on eBay can handle
1W--that's hype, just FYI. They'll get way too hot and die if you try. Heatsinking will be essential. c) Why so many LEDs? They're far more efficient making red light than incandescent bulbs--you can get equal brightness with far less power. d) DOT has tight specs on this stuff. Your proposed setup is likely enough brighter than the bulbs you're replacing that it's illegal. e) The heat's gonna kill you if you're not extremely careful, especially on the running lights. That's your biggest challenge. SMPS helps. f) Most LEDs have directional packages. Don't forget to spread the light around so you get the same visibility.
sistors drop the rest, the efficiency is 10/14ths (71%). Power wasted is 4/=
14ths, not 2/3rds.
W--that's hype, just FYI. They'll get way too hot and die if you try. Hea= tsinking will be essential.
candescent bulbs--you can get equal brightness with far less power.
gh brighter than the bulbs you're replacing that it's illegal.
on the running lights. That's your biggest challenge. SMPS helps.
around so you get the same visibility.
The listed specs on the LEDs are:
1.9-2.3V Forward voltage
300mA Continuous current max
450mA Pulsed current max
I assume pulsed at some duty cycle. So 450mA pulsed x 2.3V max =3D 1.035W c= alculated, which is where I assume the seller got the specs. I'm running 25=
0mA continuous, not pulsed, and I chose 2V rather than 2.3, specifically to= avoid running these near their maximums. As to supply voltage, the bike's = measured alternator output varies from 12.5-14.2V, but spends most of its t= ime in the 13-13.8V range. I chose 14.5 as the upper limit as a worst-case = scenario. So, 250mA x 2V =3D 0.50W, about 72% of their rated continuous max= imum. I built a protoboard array using a current-controlled bench power sup= ply, with 6 strings at 6 LEDs per, adjusted voltage to 12V and current to 1= .5A and let it run for a couple of hours. The LEDs did get warm to the touc= h, perhaps 15-20F above ambient. I mounted my stock lens over the array and= brightness was perceived to be about 50% brighter, though I have no way to= measure actual luminosity.
The reason for many smaller LEDs rather than just a few large CREE elements= is to get the lens evenly lit. The CREE leds that I played around with fir= st just produced a few terribly bright spots on the lens. The tests I did w= ith the array produced a lens that was evenly lit across the full face with= no dark spots or dark corners/edges. Along with even lighting, the diffusi= on of the lens also gave me great off-axis brightness, I had to move past 4=
5 degrees off axis before noting a decrease in brightness.
And yes, my intent is to get brighter brake lights as well as freeing up al= ternator output for my heated gear. I've had several near misses in the las= t few years where I've almost been rear-ended, the one driver I was able to= talk to said that my rear lighting didn't really stand out from the car br= akelights around me, and this was after I added the brake-only incandescent= bulbs in the outer pods. Subsequently I've installed LED 1156 and 1157 bul= b replacements installed which are as bright as the incandescent bulbs (tha= t 1157 cost me $42!)=20
I've already discussed this with my local inspection station and they said = brighter tail and especially brake lights were no problem, and in fact he t= hought it was a good idea as long as I didn't modify or cut the existing DO= T certified housing assembly.=20
Ideally I'd like some boost arrangement that allowed me to run larger strin= gs, maybe 30V or 40V strings, mainly to limit driver circuit counts. I'll c= all it the "device", the circuit that has power going in and the voltage/cu= rrent regulated power out to the LED strings. The center tail light housing= would have two "devices", one to drive 15 LEDs for the tail light and one = to drive the 25 LEDs for brake lighting. Each of the two side pods would h= ave its own circuit board with one device to drive the 15 LED array for tha= t pod. Each assembly will be fully enclosed within the stock lens housing, = using a BA15 base to plug into the stock bulb socket. I've already designed= the board layout and perimeter profile to mount into the housing such that= they're held in place by the lens. I've got access to a laser cutter so cu= tting the convoluted edge shape to accomplish this will be trivial.
The only thing left is the "glue", the stuff inside the "device", and thoug= h I took basic DC and TTL classes in school, that was over two decades ago.
Down the road probably the only change I'll make is to use PWM via a Tiny13= or Arduino product to handle the light output changes for brake and tail.
Plan for current, not voltage. You only care about having enough voltage to drive the current, but the current is the value the LED cares about - the voltage is what it is when the current is right. So, look at current sources.
The 11.5 V low is only going to happen when the bike is not running and the battery is down, or the charging system is fubar (or at least fuinr.)
Dirt simple but not the best way (better than a resistor, though) is to use an LM317 (20 cents in wimpy TO-92, 44-50 cents to get a TO-220) as a current source. Basically acts as a variable resistor of exactly the right value that gives you the desired current, so long as enough voltage is available. Simple, cheap, robust, only needs a resistor, though it will appreciate a capacitor or two.
However, if you go poke about at mouser or newark or digi-key you'll find the cute little things called LED drivers, made for the job. The Maxim ones are kinda spendy, as they go.
Looking at Mouser, the Diodes Inc. AP8800 is a down-convert (buck) switcher that works on 8-28V that can drive up to 350mA at $1.19 for one, or $1.05 if you are buying 25. You'll need a few other parts to make it work - look at the datasheet. They have another one (8801) that will drive 500mA, but you don't need that.
Infineon has something with the catchy name ILD4035 E6327 @ 1.30 for 1 or $1.10 for 10 that works from 4.5 to 42V
Then again, you could use something like a SuperTex 9930 that would allow you to boost voltage and drive more LEDs in a string. Fewer strings, fewer drive circuits needed. Does need quite a few external components, but the data sheet does include notes for automotive use, so it's headed in the right direction. $1.25 for 1
MPS Mp4012 is yet another buck/boost option, at $1.75 for one.
The key to understanding the use of these things is to read the data sheets, which most of the places selling them have handy links to. In the case of the LM317, the datasheet generally has "how to use it as a current source" (or you can google LM317 current source, and find many things, some of which will be better than drunk-posting-college-student-quality) and then you have to apply the fact that we've told you to drive the LEDs with a current source to figure out that you hook the LEDs up to it - though some of your search results will show that, too.
For far less $ on parts you can also explore the blocking oscillator/joule thief flavor of circuit, where a cheap transistor and a roll-your own inductor (and a resistor, possibly a diode or two...) will do the trick, but most of those published are not intended for 12V drive or 250-300mA diodes, so you would need to experiment a bit to find values that would work. May not be your cup of tea - or it might be.
--
Cats, coffee, chocolate...vices to live by
Please don't feed the trolls. Killfile and ignore them so they will go away.
On Tuesday, March 6, 2012 3:49:43 PM UTC-5, Lasse Langwadt Christensen wrot= e:
m
e
-
t
.)
a
o
a
l
e
2v-5-pack-66270
On Tuesday, March 6, 2012 3:49:43 PM UTC-5, Lasse Langwadt Christensen wrot= e:
m
e
-
t
.)
a
o
a
l
e
2v-5-pack-66270
That's a darn good solution. It's a 300mA constant-current source with 10v compliance. So, each driver could drive two strings of [4 LEDs plus one ballast resistor] each, say, 3.3 ohms, 1/8W. (150mA per string)
That gives 2v x 150mA =3D 300mW dissipation per LED. Heatsink the cup lead, and the LED might even last a while. Efficiency will be ~75%.
Each driver drives 8 LEDs.
Another choice would be strings of 5 series LEDs, amounting to roughly 10V, plus a limiting resistor per string. That's almost as efficient as the SMPS above, it's dirt-simple, and rugged as heck. Drawback: output varies with battery voltage.
A boost regulator could drive strings of 10 LEDs or more in series. The effciency's good, but the reliability's much worse.
resistors drop the rest, the efficiency is 10/14ths (71%). Power wasted is =
4/14ths, not 2/3rds.
1W--that's hype, just FYI. They'll get way too hot and die if you try. H= eatsinking will be essential.
incandescent bulbs--you can get equal brightness with far less power.
ough brighter than the bulbs you're replacing that it's illegal.
y on the running lights. That's your biggest challenge. SMPS helps.
ht around so you get the same visibility.
calculated, which is where I assume the seller got the specs.
There's no way the package I saw on eBay can handle 1W, no way. If you carefully heat-sink the cup lead, I might believe 300mW. Standard T-1 3/4 5mm LEDs are rated to 70mW. The 10mm LED just simply isn't 12x better.
specifically to avoid running these near their maximums.
You have no choice as to voltage. You choose the current, the LED characteristics set the voltage.
From your description I thought that the LEDs had a rated Vf of 2v. If that's not true, some of my calculations might need adjustment.
2.5-14.2V, but spends most of its time in the 13-13.8V range. I chose 14.5 = as the upper limit as a worst-case scenario. So, 250mA x 2V =3D 0.50W, abou= t 72% of their rated continuous maximum.
That'll burn it up. The continuous ratings often assume cooling.
with 6 strings at 6 LEDs per, adjusted voltage to 12V and current to 1.5A a= nd let it run for a couple of hours. The LEDs did get warm to the touch, pe= rhaps 15-20F above ambient. I mounted my stock lens over the array and brig= htness was perceived to be about 50% brighter, though I have no way to meas= ure actual luminosity.
ts is to get the lens evenly lit.
That's a good idea.
y bright spots on the lens. The tests I did with the array produced a lens = that was evenly lit across the full face with no dark spots or dark corners= /edges. Along with even lighting, the diffusion of the lens also gave me gr= eat off-axis brightness, I had to move past 45 degrees off axis before noti= ng a decrease in brightness.
alternator output for my heated gear. I've had several near misses in the l= ast few years where I've almost been rear-ended, the one driver I was able = to talk to said that my rear lighting didn't really stand out from the car = brakelights around me, and this was after I added the brake-only incandesce= nt bulbs in the outer pods. Subsequently I've installed LED 1156 and 1157 b= ulb replacements installed which are as bright as the incandescent bulbs (t= hat 1157 cost me $42!)=20
d brighter tail and especially brake lights were no problem, and in fact he= thought it was a good idea as long as I didn't modify or cut the existing = DOT certified housing assembly.=20
ings, maybe 30V or 40V strings, mainly to limit driver circuit counts.=20
You'll want at least a few redundant strings. Otherwise, one LED failed open, or a bad boost converter, and you're dark.
ltage/current regulated power out to the LED strings. The center tail light= housing would have two "devices", one to drive 15 LEDs for the tail light = and one to drive the 25 LEDs for brake lighting. Each of the two side pods = would have its own circuit board with one device to drive the 15 LED array= for that pod. Each assembly will be fully enclosed within the stock lens h= ousing, using a BA15 base to plug into the stock bulb socket. I've already = designed the board layout and perimeter profile to mount into the housing s= uch that they're held in place by the lens. I've got access to a laser cutt= er so cutting the convoluted edge shape to accomplish this will be trivial.
ugh I took basic DC and TTL classes in school, that was over two decades ag= o.
13 or Arduino product to handle the light output changes for brake and tail= .
That could easily be done in analog--just sum the braking signal into the converter reference.
--=20 Cheers, James Arthur
l
l
d
d
e
.
m
resistors drop the rest, the efficiency is 10/14ths (71%). Power wasted is =
4/14ths, not 2/3rds.
1W--that's hype, just FYI. They'll get way too hot and die if you try. H= eatsinking will be essential.
incandescent bulbs--you can get equal brightness with far less power.
ough brighter than the bulbs you're replacing that it's illegal.
y on the running lights. That's your biggest challenge. SMPS helps.
ht around so you get the same visibility.
calculated, which is where I assume the seller got the specs. I'm running =
250mA continuous, not pulsed, and I chose 2V rather than 2.3, specifically = to avoid running these near their maximums. As to supply voltage, the bike'= s measured alternator output varies from 12.5-14.2V, but spends most of its= time in the 13-13.8V range. I chose 14.5 as the upper limit as a worst-cas= e scenario. So, 250mA x 2V =3D 0.50W, about 72% of their rated continuous m= aximum. I built a protoboard array using a current-controlled bench power s= upply, with 6 strings at 6 LEDs per, adjusted voltage to 12V and current to= 1.5A and let it run for a couple of hours. The LEDs did get warm to the to= uch, perhaps 15-20F above ambient. I mounted my stock lens over the array a= nd brightness was perceived to be about 50% brighter, though I have no way = to measure actual luminosity.
ts is to get the lens evenly lit. The CREE leds that I played around with f= irst just produced a few terribly bright spots on the lens. The tests I did= with the array produced a lens that was evenly lit across the full face wi= th no dark spots or dark corners/edges. Along with even lighting, the diffu= sion of the lens also gave me great off-axis brightness, I had to move past= 45 degrees off axis before noting a decrease in brightness.
alternator output for my heated gear. I've had several near misses in the l= ast few years where I've almost been rear-ended, the one driver I was able = to talk to said that my rear lighting didn't really stand out from the car = brakelights around me, and this was after I added the brake-only incandesce= nt bulbs in the outer pods. Subsequently I've installed LED 1156 and 1157 b= ulb replacements installed which are as bright as the incandescent bulbs (t= hat 1157 cost me $42!)=20
d brighter tail and especially brake lights were no problem, and in fact he= thought it was a good idea as long as I didn't modify or cut the existing = DOT certified housing assembly.=20
ings, maybe 30V or 40V strings, mainly to limit driver circuit counts. I'll= call it the "device", the circuit that has power going in and the voltage/= current regulated power out to the LED strings. The center tail light housi= ng would have two "devices", one to drive 15 LEDs for the tail light and on= e to drive the 25 LEDs for brake lighting. Each of the two side pods would = have its own circuit board with one device to drive the 15 LED array for t= hat pod. Each assembly will be fully enclosed within the stock lens housing= , using a BA15 base to plug into the stock bulb socket. I've already design= ed the board layout and perimeter profile to mount into the housing such th= at they're held in place by the lens. I've got access to a laser cutter so = cutting the convoluted edge shape to accomplish this will be trivial.
ugh I took basic DC and TTL classes in school, that was over two decades ag= o.
13 or Arduino product to handle the light output changes for brake and tail= .
It's an PWM LED dimmer for 12v and 8 amps. It may or may not be usable for this situation, but it seems someone on here is always looking to build a PWM. This one in $9.99 in a box with in/out connector. The website is loaded with LED lights. I got a coupon at the Orlando Hamfest. They had 100s of LED lights on display. If you buy anything, use coupon code hamcation12.
10% discount until 3-15-12.
That brings up a question. If a PWM circuit used a 555, could pin 5 be used to keep the output for stable when the input varies from 11.5 volts to 14.8 volts, as in an automobile. Mikek
snipped-for-privacy@fierocentral.com wrote in news:19185217.503.1331059396386.JavaMail.geo-discussion-forums@ynhs12:
a single 1157 is probably only around 12 watts,so if you go with 40 1W LEDs,you're using nearly THREE TIMES the power.
you'd need a complicated BOOST driver circuit,to raise the ~13V of your system to 30-40V. one more thing to fail. Plus,if a single LED fails,the whole string goes out.the longer the string of LEDs the more light you lose.
LEDs are not driven by voltage,but by CURRENT. if you put current thru them,they light up at ~2.1 volts(red LED),reaching rated brightness at the specified current. exceed the CURRENT,and the LED burns up. Once the LED lights up,the Vdrop varies by 10's of millivolts,like a poor zener diode,over a wide range of current.(current changes the brightness,not voltage.)
your LED string has to be below the 13.8 volts of the typical automotive battery charging voltage and still allow voltage drop of the current regulator-driver circuit. You really want to be below the 12 volts of the battery,and leave the rest for headroom.
you can figure the "wattage" of the LED by multiplying the Vdrop by the rated current.
I'd go with 5-LED strings,with a resistor for each string, and a current limiting voltage regulator IC,like an LM317 programmable V regulator in current-limit mode;one resistor and a couple of chip caps and you're in business. maybe a protection diode for reverse current spikes.
if the taillight also does double duty as a brake light,then you need to have a 2 current setup,like a dual filament brake lamp,one regular brightness,one high brightness for braking.
Note that if your boost converter is feeding the string of LEDs and the regulation is set up for current rather than voltage, then if the load is disconnected, the boost converter will keep pumping until something breaks down, potentially destroying the boost circuit. Hopefully this is obvious. Basically if there is no current flowing in the sense resistor, then the switcher just keeps on pumping. Thus you need to set up an overvoltage protection scheme unless you are 100% positive the load will never be disconnected.
If you are working on brake lights for a motorcycle..consider these things...
A SINGLE light is so easy to loose in the jungle of lights and also it is hard to see in perspective. As you get closer and closer, nothing much changes except that it gets a little brighter. I think you want two or more lights as widely spaced as possible. As someone approaches you, the object the lights outline appears to get larger in perspective. I think ONE light is very dangerous because you loose perspective. I think this is true of the front headlight as well. I added two more driving lights spaced apart on the front of my bike, not just so I have more light, but so that I appear to others as an object, not as a point.
I set up the rear turn signal lights which are bright yellow and spaced apart to flash when I put the brake on, unless the turn signal is on. So if I'm stopped and not making a turn, the brake light is on and both rear turn signals flash. If I'm stopped and making a turn, the brake light is on and one turn signal flashes. Something is always flashing. When I stop at a light, I will watch the car behind me in the mirror and keep the brake on so the lights are flashing until I can see that the car is slowing down and will stop. When it's stopped, I let off the brake so the lights stop flashing in an annoying way.
As far as each individual light is concerned, I think for the same total amount of light, an intense point source is better at catching your eye compared to an evenly dispersed light source. My reasoning is, that when you are far away, they both look like points, and as you approach the point source will get very bright, while the dispersed source will not get as bright.
To clarify, I'm talking about having multiple point sources so as someone approaches, each one gets very bright, but together they spread apart from each other so as to appear as an object.
I think this subject of visibility of lights is key to staying safe on a motorcycle.
resistors drop the rest, the efficiency is 10/14ths (71%). Power wasted = is 4/14ths, not 2/3rds.
handle 1W--that's hype, just FYI. They'll get way too hot and die if you= try. Heatsinking will be essential.
incandescent bulbs--you can get equal brightness with far less power.
enough brighter than the bulbs you're replacing that it's illegal.
especially on the running lights. That's your biggest challenge. SMPS = helps.
light around so you get the same visibility.
1.035W calculated, which is where I assume the seller got the specs. I'm = running 250mA continuous, not pulsed, and I chose 2V rather than 2.3, = specifically to avoid running these near their maximums. As to supply = voltage, the bike's measured alternator output varies from 12.5-14.2V, = but spends most of its time in the 13-13.8V range. I chose 14.5 as the = upper limit as a worst-case scenario. So, 250mA x 2V =3D 0.50W, about 72%= of their rated continuous maximum. I built a protoboard array using a = current-controlled bench power supply, with 6 strings at 6 LEDs per, = adjusted voltage to 12V and current to 1.5A and let it run for a couple = of hours. The LEDs did get warm to the touch, perhaps 15-20F above = ambient. I mounted my stock lens over the array and brightness was = perceived to be about 50% brighter, though I have no way to measure = actual luminosity.
elements is to get the lens evenly lit. The CREE leds that I played = around with first just produced a few terribly bright spots on the lens. = The tests I did with the array produced a lens that was evenly lit across= the full face with no dark spots or dark corners/edges. Along with even = lighting, the diffusion of the lens also gave me great off-axis = brightness, I had to move past 45 degrees off axis before noting a = decrease in brightness.
alternator output for my heated gear. I've had several near misses in = the last few years where I've almost been rear-ended, the one driver I = was able to talk to said that my rear lighting didn't really stand out = from the car brakelights around me, and this was after I added the = brake-only incandescent bulbs in the outer pods. Subsequently I've = installed LED 1156 and 1157 bulb replacements installed which are as = bright as the incandescent bulbs (that 1157 cost me $42!)=20
Part of the reason for that price is protection from the nasty electrical environment that is common to motor vehicles. You need protection from
100 V spikes of either direction that may last over 10 ms.
said brighter tail and especially brake lights were no problem, and in = fact he thought it was a good idea as long as I didn't modify or cut the = existing DOT certified housing assembly.=20
strings, maybe 30V or 40V strings, mainly to limit driver circuit counts.= I'll call it the "device", the circuit that has power going in and the = voltage/current regulated power out to the LED strings. The center tail = light housing would have two "devices", one to drive 15 LEDs for the tail= light and one to drive the 25 LEDs for brake lighting. Each of the two = side pods would have its own circuit board with one device to drive the =
15 LED array for that pod. Each assembly will be fully enclosed within = the stock lens housing, using a BA15 base to plug into the stock bulb = socket. I've already designed the board layout and perimeter profile to = mount into the housing such that they're held in place by the lens. I've = got access to a laser cutter so cutting the convoluted edge shape to = accomplish this will be trivial.
Longer strings has a bad drawback of having more single failure modes. I suggest being very thoughtful about more than 6 LED long strings.
though I took basic DC and TTL classes in school, that was over two = decades ago.
Tiny13 or Arduino product to handle the light output changes for brake = and tail.
My first reaction is to reconsider how many LEDs you may need. 70x 1W Leds at this power rating is going to be blinding for the people behind you- a bit like having fog lights on and then some.
I've no idea of the law on this, but to put things in perspective you can find LED replacements lights that consume a total of 4W
As others point out - you do not get to choose the voltage the LEDs run at as they are current devices, but things will be easier to design when you deal with fewer LEDs and use less power. There are plenty of high efficiency drivers for this job but you need to get the specs sorted out.
If this is a safety issue (not just a hobby project) you might want to find out where some cop bikes get their taillights. First, see if these are even legal on civilian bikes.
They have LED brake lights on my cities cop bikes that, when activated, produce a very rapid (and bright) flicker for a fraction of a second before they come on steady. Very visible.
--
Paul Hovnanian mailto:Paul@Hovnanian.com
------------------------------------------------------------------
The Three Laws of Thermodynamics:
1) You can't win.
2) You can't break even.
3) You can't quit the game.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.