LED wiring question... I think I'm on the right track...

CapSmashy · Post by **CapSmashy** » Fri Jun 20, 2008 5:42 pm

http://www.lck-led.com/product_info.php/products_id/376

I'm eyeballing some led strips like pictured int he link above and per the description:

Runs on 12-Volt Systems like Solar Cell Charged Batteries, 12 Volt Battery or 120 Volt-to-12 Volt Power Pack Converter

For a 12 volt battery system, would a set up like this work for my power source to run one (or more) strips like this?

http://www.pinecomputer.com/12vdcbapobox.html

ibdave · Post by **ibdave** » Fri Jun 20, 2008 5:50 pm

That's cool stuff.. I hope a pack like that works...

CapSmashy · Post by **CapSmashy** » Fri Jun 20, 2008 5:51 pm

I do too. I have a beater tail coat that is screaming for some illumination.

ibdave · Post by **ibdave** » Fri Jun 20, 2008 5:57 pm

Cap, I just got these the other day.. I use a AA battery pack to power them up.. VERY bright.. http://vibelights.com/bili.html

CapSmashy · Post by **CapSmashy** » Fri Jun 20, 2008 6:09 pm

Pretty sweet.

I am also planning on making some of these guys.

http://www.instructables.com/id/LED-Throwies/

MikeVDS · Post by **MikeVDS** » Fri Jun 20, 2008 9:49 pm

Yep, that should work great for you. You should be able to run many of those off that pack all week. If you're attaching more than one to a single pack, remember that have to be wired parallel, not in series, or you'll have a voltage drop (the LEDs will be dim or not light up). Series would mean you attach one strand to the end of the other to form a chain. That's a no-no. You have to attach each strand directly to the pack and not to the end of another strand.

Sorry if I'm telling you basic stuff, but it seemed like a basic question this might help with.

Here is a link with some basic diagrams of series vs parallel

http://www.bcae1.com/srsparll.htm

phil · Post by **phil** » Fri Jun 20, 2008 11:43 pm

Hi, Cap,

The ad for the lights doesn't say what amperage they draw. The box that holds 10 AAs says it puts out 12 VDC - AAs are about 1.2VDC each plus or minus, depending on whether they're rechargeable. So 10 x 1.2 gives you your 12V.

I'd read the label on the batteries. If they're alkaline, they may be 1.5 volts each, which would give you 15VDC into the LEDs, and that may melt something.

Without knowing what the LED string draws, you can't figure out how long your batteries will last.

LeChatNoir · Post by **LeChatNoir** » Sat Jun 21, 2008 12:19 am

Cap,

Phil makes a good point. As long as the battery voltage adds up to 12vdc or close that pack will work just fine. And it should run one string of those LEDâ€™s for quite a while. Iâ€™ve got a string very much like this one ordered to test for additional lighting on The Contraption. Hopefully theyâ€™ll be here soon so I can play.

Captain Goddammit · Post by **Captain Goddammit** » Sat Jun 21, 2008 1:35 am

LEDs aren't that sensitive to voltage, 15 volts won't fry them. Their current draw is extremely low, ten AAs should run them a long time.

Dork · Post by **Dork** » Sat Jun 21, 2008 3:02 am

You can save a few bucks by picking up some AA battery holders at Radio Shack and putting it all in a ziplock bag. I'd use 8 AA's if Alkaline, 10 if NiMH or NiCd. Using a couple of the flat 4-battery holders makes for easy hiding under clothing.

CapSmashy · Post by **CapSmashy** » Sat Jun 21, 2008 7:06 am

MikeVDS wrote:Yep, that should work great for you. You should be able to run many of those off that pack all week. If you're attaching more than one to a single pack, remember that have to be wired parallel, not in series, or you'll have a voltage drop (the LEDs will be dim or not light up). Series would mean you attach one strand to the end of the other to form a chain. That's a no-no. You have to attach each strand directly to the pack and not to the end of another strand.

Sorry if I'm telling you basic stuff, but it seemed like a basic question this might help with.

Here is a link with some basic diagrams of series vs parallel

http://www.bcae1.com/srsparll.htm

No that's perfect for me.

I am much more of an open the package and plug it in/add batteries and tun it on kind of guy when it comes to small stuff.

Big scale electrical, like rewiring a house, I have no issues with. Its this small stuff dealing with DC and low power draws is what screws me up.

CapSmashy · Post by **CapSmashy** » Sat Jun 21, 2008 7:10 am

phil wrote:Hi, Cap,

The ad for the lights doesn't say what amperage they draw. The box that holds 10 AAs says it puts out 12 VDC - AAs are about 1.2VDC each plus or minus, depending on whether they're rechargeable. So 10 x 1.2 gives you your 12V.

I'd read the label on the batteries. If they're alkaline, they may be 1.5 volts each, which would give you 15VDC into the LEDs, and that may melt something.

Without knowing what the LED string draws, you can't figure out how long your batteries will last.

I actually had wondered about that after I posted the link to that box, did some hunting and uncovered basically what Captain GD posted.

CapSmashy · Post by **CapSmashy** » Sat Jun 21, 2008 7:13 am

Dork wrote:You can save a few bucks by picking up some AA battery holders at Radio Shack and putting it all in a ziplock bag. I'd use 8 AA's if Alkaline, 10 if NiMH or NiCd. Using a couple of the flat 4-battery holders makes for easy hiding under clothing.

I'm heading to Fry's today to look first since Radio Shack tends to be quite proud of their products, but yeah, I was planning on acquiring most of the parts locally. I'll probably have to order the led strips unless I get lucky.

Fry's should have everything I need for the LED Throwies though.

phil · Post by **phil** » Sat Jun 21, 2008 11:21 am

I never argue with Captain Goddamit.

[youtube][/youtube]

Never connect an LED directly to a battery or power supply!
It will be destroyed almost instantly because too much current will pass through and burn it out.

http://www.kpsec.freeuk.com/components/led.htm

Although LEDs may be indestructible, the wiring isn't. I've melted the innards of more than one device when I used a wallwart that passed to much power to the device - the smell of solder melting in the morning. It's like the smell of victory.

Your mileage will vary, and I'll be happy to be wrong. Long and happy lighting on the playa!

MikeVDS · Post by **MikeVDS** » Sat Jun 21, 2008 1:34 pm

Most power sources vary in voltage a bit anyway. Plus or minus a little bit will not hurt most common electronics. I wouldn't be worried about 15v one bit for the wiring or anything.

gyre · Post by **gyre** » Sat Jun 21, 2008 2:12 pm

Keep in mind that most LEDs are 1.5 to 3 volts, with some luxeon running at up to 8 volts.
So there is a power loss with almost all 12 volt LED products.
Some truck lighting systems use an electronic regulated power supply to be more efficient but most are resistor types.
There are simple ones with luxeons and better ones available in small form.

It may be more efficient for you to use direct wired LEDs matched to the power supply.
1.5 volt units can be wired to batteries in parallel.

MikeVDS · Post by **MikeVDS** » Sat Jun 21, 2008 10:34 pm

I'd have to assume those LEDs are wired in series, which is what makes them a 12v system. The extra LEDs are the "resistors" in the circuit. No need for any resistors when you have multiple LEDs already.

If you just want one battery, parallel is the way to go. Much lighter and it's easy to change out that battery, if you even need to.

gyre · Post by **gyre** » Sun Jun 22, 2008 1:28 am

I have seen some strings wired in series.
It's a really bad practice as the trigger voltage is almost never the same and you are feeding the current through all of the LEDs in series.
Some products do this though.

MikeVDS · Post by **MikeVDS** » Sun Jun 22, 2008 11:06 am

LEDs are pretty forgiving and their current is so low that unless you have some strange specialty LEDs that are very sensitive, I don't see any problem with that. Or if you're going to wire many thousands of them in series you may have to worry.

gyre · Post by **gyre** » Sun Jun 22, 2008 11:29 am

I can't believe they get away with it at all.
As low as the current is, it is more than they were designed for.
I have tried this myself and I always had at least one LED with s different resistance from the rest that would burn out too quickly or refuse to light.
They do make these work, but when one goes, you lose the series unless it is still conducting.

I think you will always use more power than is necessary.
You only need enough to just trigger full light.

I am really intrigued about the idea of using side emitting fiber optics.
You could use bright, very efficient LEDs and run the fiber where you choose.
It would require very few to have a lot of light with the brighter LEDs now.

MikeVDS · Post by **MikeVDS** » Sun Jun 22, 2008 12:37 pm

I think you will always use more power than is necessary.

Not true. If you know the voltages of the LEDs, you can match the voltage and not have a problem. I can't say certainly, but I'd bet quite a bit of money that LEDs are designed to carry much more current than they draw individually. 1st, you want a factor of safety and to compensate for manufacturing flaws, second, I doubt more people run LEDs on circuits that are at their exact voltage. Most systems are 12v, 24v etc. I don't know if I've ever seen any industrial applications that were 1.5-3V. They alter the circuit so the LEDs are at the proper voltage, but they are still carrying the higher amperage.

I have tried this myself and I always had at least one LED with s different resistance from the rest that would burn out too quickly or refuse to light.

That shouldn't have anything to do with the resistance in the LED, just a bad LED, incorrectly designed circuit (Are you putting mismatched voltages in parallel? That would do it.), or a LED with a different lifespan. Yeah, the series will fail when one burns out, but you can replace that one. They will last 1000's of hours when done properly.

gyre · Post by **gyre** » Sun Jun 22, 2008 12:49 pm

I just mean manufacturing tolerances create variances between nominally identical LEDs.
It seems you would end up needing more power to fire the string than if each was drawing just enough.
And it's just a bad approach.

i suspect they are specing units that are more tolerable of variations and can handle the current.
That would make them functional without adding much cost.
I have noticed brightness variations in the strings.

Anyway, that's how I would solve the issue if I was making them.

MikeVDS · Post by **MikeVDS** » Sun Jun 22, 2008 5:31 pm

It seems you would end up needing more power to fire the string than if each was drawing just enough.

Nope. I don't see why that would be the case. Just because a LED is carrying higher amperage doesn't mean it'll draw any more.

And it's just a bad approach.

I see no reason why it's bad as long as you're not trying to run more amps than a circuit can carry. Many old street lights were designed in series with a shunt so they all wouldn't go out when the bulb burned out. They were on variable voltage transformers so the voltage of the circuit would drop when a bulb burnt out so that all bulbs would remain 5000 V. It's a quite practical design method for many purposes and see no reason why you shouldn't use it for LEDs. A long string of parallel LEDs is going to use more wire, which all has resistance and will probably end up using more power than series lights. Parallel is more expensive to wire. The only benefit to parallel is that when you lose one light you don't lose them all. If you don't want to mess with things and want them to just work, it might be worth putting them in parallel, other than that I see no advantage. If I'm missing something that you can explain, let me know.

gyre · Post by **gyre** » Mon Jun 23, 2008 12:07 am

Your points about parallel wiring are true and that plus the added resistors in separate circuits is why I was trying it in the first place.
In reality, they have variations in resistance, so some will have higher or lower resistance with consequences for the circuit.

The bigger problem is that LEDs are not strictly resistors as a filament is.
I used to know all the formulae on the ins and outs on this, but I have forgotten them.

But hell, if they work and you're satisfied with the reliability, it doesn't really matter.

I would rather keep it simple.
Anyone have a source for a mounting arrangement for the end of a fiber optics piece?
This is the article that got me thinking about this approach.
He makes his own.
That may work fine.
http://www.allyn.com/Bike_lighting.pdf

MikeVDS · Post by **MikeVDS** » Mon Jun 23, 2008 6:21 am

Your points about parallel wiring are true and that plus the added resistors in separate circuits is why I was trying it in the first place.

You do realize that putting in resistors is the exact same thing as putting in extra LEDs to match a voltage. For example, if you have a 12V circuit and you put 6 2V LEDs in series to match the voltage or you take one of the same 2V LEDs and put a resistor in series with it to match the 12V you are drawing the exact same current and the LED is carrying the same amperage.

In reality, they have variations in resistance, so some will have higher or lower resistance with consequences for the circuit.

In reality they are different, but not enough for most multimeters to even detect or to make a difference in your circuit. For example, let's say your 2V LED actually likes 1.97V. Regardless of if you stick it in series or not, you're not going to use very expensive equipment to perfectly match it to 1.97V, you're just going to hook it up to a 2V source. And actually, in series, your variance could help with that problem since you'll probably end up with just as many that like higher voltage as like lower voltage than the nominal value, so the series will likely better match your circuit. Either way, in reality it probably changes nothing and if it does, statistically it'd hurt parallel circuits more.

The bigger problem is that LEDs are not strictly resistors as a filament is.

That probably is the bottleneck. I'm not sure if filament is the right word for LEDs, but regardless I'm guessing they can carry at least a large fraction of an amp without trouble. If you find documentation on any LEDs that has these numbers I'd like to see them. I'd be floored if their carrying load was close to what they draw. I'd have to say it's many time higher, as it is with almost all electrical devices.

But hell, if they work and you're satisfied with the reliability, it doesn't really matter.

There shouldn't be any drop in reliability, so yeah, I'd be satisfied with that.

gyre · Post by **gyre** » Mon Jun 23, 2008 7:23 am

As I recall, the efficiency is the cause of the problem.
With such low resistance involved, the smallest variance from one to the other causes relatively large changes.
This resulted in one often overheating or failing to trigger.
Perhaps manufacturing tolerances are better now and the diodes are more flexible in voltage handling too.
The larger the current, the less this is an issue.

This issue often crops up in high efficiency circuits in alarm systems.
Even the ultra high efficiency relays I use draw so little power that the wiring becomes a second resistor in the circuit and minor changes in the wire are an issue.
I think some circuits use an added resistor to alleviate this by increasing the current, as much as to inhibit hotwiring.

ricochet · Post by **ricochet** » Tue Jun 24, 2008 8:52 pm

gyre wrote: I would rather keep it simple.
Anyone have a source for a mounting arrangement for the end of a fiber optics piece?
This is the article that got me thinking about this approach.
He makes his own.
That may work fine.
http://www.allyn.com/Bike_lighting.pdf

Oh my. I believe you may have just changed my life.

ricochet · Post by **ricochet** » Tue Jun 24, 2008 9:05 pm

For what it's worth, I use this LED series/parallel array wizard to figure out how to design the circuit.

See the LED basics link for a section on Why do I need a resistor with an LED?. Makes sense to me.... Everything I've read says, yeah, you could do a series circuit, relying on the LEDs to limit current, but it's asking for trouble.

And all that said, really: blowing up electronic bits is part of the learning curve. Hack around with it and if it works, great. And try whatever you're comfortable trying.

MikeVDS · Post by **MikeVDS** » Tue Jun 24, 2008 9:57 pm

I see what they mean about "asking for trouble" but that's basically only if you're designing the circuit wrong. If you assume a LED has a constant resistance with varying voltage, you're doing it wrong to begin with. If you run a LED at a voltage that it's not designed to run continually at, then you're doing it wrong, regardless of how you wire it. Maybe I'm just a nerd, but the math, design and optimization is the fun part for me. I know it's over some peoples heads, so they just like to keep the math simple and not have to worry about "ohhh, varying 'resistance'" but it's not complex, just V=IR basic stuff. Pick LEDs that will work for you and do the math correctly.

galaxybeing · Post by **galaxybeing** » Sat Jul 05, 2008 6:16 pm

I got these last year and they worked great. Solar powered string lights - 60 leds and they lasted all night.

http://cgi.ebay.com/SOLAR-37-ft-60-LED- ... dZViewItem

LED wiring question... I think I'm on the right track...

LED wiring question... I think I'm on the right track...

solar string lights