I’m making a Halloween costume for my son - a flying saucer suit. I’d like to light the suit up with LEDs. I’ve got a string of multicolored LED Christmas lights, and I’d like to use some of those lights and rewire them to run off of a battery pack.
The LED string has 70 lights total; two blocks of (35 LEDs in series). I assume that the two blocks are wired in parallel. There is no transformer at the plug, so I assume further that the total voltage drop across the 35 LEDs is 110 volts. Presumably the two blocks of LEDs are wired in opposite directions, so they flash alternately at about 120 Hz.
As I see it, I have a block of 35 LEDs with a total voltage drop of 110V, or an average of 3.14 V each (does that seem high?). There are red, orange, yellow, blue, and green lights, but the color may be coming from the colored plastic “bulbs” that cover the LEDs. The lights are arranged thusly: red, orange, yellow, (blue or green), repeat.
Sooo, here’s my plan: cut one block up into 8 sets of 4 LED’s in series, then wire them all in parallel and run them off a pack of 8 AA batteries. I’ll cut the LEDs so that each set of 4 is the same, colorswise. Should this work? I’m not looking for optimized brightness or anything; I just want to know whether I’ll burn out half of the LEDs because I’m not calculating the voltages properly, or something. Are my assumptions sound? I can throw in a different battery pack if that would work, but the kid is only 30 months, so no car batteries! It only has to stay lit for an hour or so.
If they are white LEDs with plastic covers over them, then about 3 volts seems right. White or blue LEDs will typically have a higher forward voltage than red or green LEDs.
I don’t like driving LEDs off of a voltage source, but yeah, it should work.
AA’s are typically rated at a couple of amp-hours (IIRC). Taking a wild guess that your LEDs are probably going to draw about 50 mA per set, I’d expect them to last for four or five hours. Even cutting that in half for a safety factor, you should get an hour or two out of them with no problem.
Actually, it might not work very well. The trouble with driving LED’s by voltage is that their marginal resistance is very low. As you crank the voltage up, nothing, nothing, nothing, and then all of a sudden there’s a huge current and you destroy them. Even if you could adjust it finely enough, the voltage threshold also changes with temperature.
Much better to drive them with a larger voltage that also goes through a resistor. Size the resistor and voltage so you get whatever drive current you want, usually 20 mA or less for visible-light LEDs. Assume the voltage on the LEDs themselves can change by a few percent and you still want the current to stay below 20 mA.
I don’t believe light emits diodes are being used in this case. I believe the question is about miniature christmas lights using a filament. It could be leds, but thats not the normal case for a string of 35 lights. Can brossa clarify this please? I’ll give this information now in case it’s what is being asked for.
The highest 120volts with 35 lights in series is a top voltage of 3.4 volts. The same lights go into 50 light sets, so the low end is 110V with 50 lights for a low voltage of 2.2 volts. Flasher bulbs in this type of light depend on epanding metal to open and close the connection for power.
These are definitely LEDs, not incandescent bulbs. They are molded into colored plastic ‘bulbs’ to make them look more traditional, I guess. I have no idea what the specs are for the individual LED’s; I can’t even tell if they are red, green, blue, etc., or all ‘white’ under the colored plastic. Looking at the strand of lights, I can’t see where they would have hidden a DC transformer or other circuits. FWIW, this is a light set from Holiday Creations, US patents 6461019 and 6830358. I looked at the patents, and it looks like they are driving the LEDs with AC and without a current limiter.
I’d be happy to try to run them all in series with an LM317, as Crafter_Man suggests, but as I said I don’t know anything about the ratings for the individual LEDs. I’ll assume that I want 20 mA to run through the whole series circuit. So how do I know what programming resistor to use when I have multiple LEDs of (potentially) different driving voltage? Let’s assume that I want to use 32 LEDs in series: 32(3.1V) = 99.2 V; do I really need 10 9v batteries to run it? Can I run groups in parallel to reduce my voltage requirement? That’s why I liked the idea of the series-parallel setup in the first place. Would I need a LM317 for each series portion of my series-parallel setup?
Clearly I know only enough electronics to be dangerous. So please walk me through it a bit more slowly… Let me try to rephrase my problem: imagine that I’ve handed you a box of LEDs that all look the same - but there may be as many as five different LED types inside the box. How would you hook up a circuit to use as many as possible, not knowing their individual specs, to run off of a reasonable number of batteries? I’d test the things, but all I have is a multimeter. What set of assumptions can I use that would work for most/all LEDs? They can run dimmer than optimal, but I’d like not to blow any up.
I have bought a couple strings of LEDs of this type at Target. They sell white, red, blue, yellow, violet, and mixed.
I also have ordered some battery powered ones from noveltylights.com but the site wasn’t very informative. I’m not even absolutely certain they’re LED and not filament. But they were only $4.50 per string.
So if I understand correctly, you want to use 8 AA batteries in series? That will give you around 12 V, which (along with an LM317) can easily power 4 LEDs in series. And you can probably get away with powering 5 in series with no problem. (If if works with 5 in series, then 5 in series would be more efficient than 4 in series.)
So I like your idea of making multiple “sets” of LEDs, with each set consisting of 4 or 5 LEDs in series, and the sets wired in parallel. The entire circuit is then connected to the 12 V battery pack and a series current regulator (using a LM317 and programming resistor).
If you can, try and make sure the sets are “matched” as best as possible, i.e. each set having the same mix of colors (e.g. 1 blue, 1 yellow, 1 red, 1 orange). That way, each set will have the same (or similar) I-V characteristics, and thus the same current will flow in each.
So if we can assume that the same amount of current will be flowing in each set, we only need to determine what the current will be. Let’s say we want 7 mA to flow in each set. If you have 8 sets, then the current regulator needs to be programmed to regulate 56 mA through the circuit.
I’m guessing that they are all white LEDs, because mixing and matching different color LEDs with different voltage drops would be difficult and much more expensive than just putting them all in series. A typical white LED is going to have a forward voltage of about 3.5 or 3.6 volts, so it matches up with what you’ve figured out so far. I’d figure that they are more likely to have a max current around 30 mA or higher, but they might be as low as 20 mA, who knows.
If you want to be safe about this, first turn on the lights and look at them long enough that you can tell how bright they are supposed to be. Then, cut off a section of 4 LEDs and run them in series with your 8 AA’s, just like you originally planned, only instead of just having the LEDs in series, also put a pot (a 5k or 10k should do nicely) in series with the LEDs. Important note - adjust the pot to a high resistance BEFORE connecting the batteries. Once you have the batteries connected, slowly turn down the resistance in the pot until the LEDs light up to the desired brightness. They will go from pretty much completely dark to very well lit on a fairly small change in the pot, so go slowly. Once you’ve got the pot set to the desired brightness, disconnect the batteries and measure the resistance of the pot with your multimeter. I think you are going to find that you need a fairly small resistance to get the LEDs to light properly. If you run them a bit dim your batteries will last longer, so there’s not really much of a penalty (and actually kind of a benefit) for under-driving them. They will probably be “bright enough” at 10 to 15 mA, even though they might go to 30 mA, for example.
There’s nothing wrong with using a current driver like an LM317, but it’s a bit of overkill IMHO.
Use regular el-cheapo (non-alkaline) batteries. They don’t vary their voltage as much as alkalines while they discharge, and LEDs are much more sensitive to changes in voltage than they are to changes in current, which is the whole point behind Crafter Man’s LM317 idea.
Cripes. They’ll give a patent to damn near any simple stupid idea, won’t they?
The LM317 - which is sold at Radio Shack for $2.29 - is actually a “programmable” linear voltage regulator. The term “programmable” means you set the output voltage using a couple resistors.
The neat thing about the LM317 is that you can also configure it to be a series current regulator, which is ideal IMO for powering LEDs. And it only requires one programming resistor when configured as a series current regulator.
Here’s a schematic of the LM317 being used as a series current regulator to drive one LED.
The current can be calculated as follows:
I = 1.25/R
Where I is the current in amps (A) and R is the resistance of the programming resistor in Ohms. If you want (for example) 56 mA through the circuit, then you would use a resistor of around 22 Ω.
Here’s what I like about using a series current regulator vs. a simple series resistor: As the battery voltage decreases, the current through the LEDs will also decrease if you use a simple series resistor. If you use a series current regulator, the current will remain steady (down to a certain battery voltage). You can think of a series current regulator as being a variable resistor, and the resistance is automatically lowered as the battery voltage goes down to maintain a constant current.
The LM317 only represents one more component over a single series resistor, and the improvement in performance is well worth it IMO. That’s why I always recommend people use a LM317 configured as a current regulator when powering LEDs.
I have found that most garden-variety LEDs are specified to have a maximum current of around 20 mA, but they can still be seen fairly easily when run at 5 mA. So for battery-operated projects, I always recommend “20 mA” LEDs be run between 5 mA and 10 mA. It makes the battery pack last longer.
If they are all white LEDs (which means the voltage drop for each LED will be around 3.4 V), then each set should have 4 LEDs.
I would suggest the route Crafter_Man suggests, or the voltage drop may quickly leave you without lights when the voltage is insufficient to drive the leds. He even provided needed details for you.
I don’t know why you think having all white LEDs would make an easier design. As long as you put them all in series they’re all going to have the same current anyway, and red and yellow LEDs are cheaper than blue ones, so they’re almost certainly going to go with the true-colored LEDs rather than tinted white ones.
The green and blue LEDs in the strands may have different forward voltages, in which case putting a ROYG and ROYB strand in parallel would tend to light the green-containing strand more than the blue-containing one unless you use an extra resistor in series with the green strands.
Again, thanks to everyone for their responses. I’ve got the soldering iron out, and I’ll be giving this a shot over the next few days. Thanks especially to Crafter_Man. Pix to follow, if this works out…
Crafter_Man, for the record, your post wasn’t there when I started typing my last one. I hope I didn’t give you the impression that I was arguing with you over the best way of doing this. That wasn’t my intent.
You really shouldn’t drive parallel circuits with a current regulator. If one strand faults then the current driver will still try and push the same amount of current through the remaining strands, which in this case would cause too much current to go through the remaining strands of lights. In other words, if you have four strands at 5 mA each, you want your current regulator set at 20 mA. But, if one strand fails, the current regulator will try and push all 20 mA through the remaining 3 strands, which will be 6.6 mA per strand. If you want to use a current regulator, then you should have a current regulator on each series strand. That way a fault on one strand isn’t going to cause an overcurrent on another strand.
That said, I wouldn’t worry too much about it because if you aren’t pushing the LEDs to the limit, then if one strand breaks the others may get a bit brighter, but not enough to really damage anything.
A voltage source with a series current limiting resistor for each series strand is simlpe and cheap. A current driver on each strand is going to be more expensive, but also works.
I was referring specifically to the OP. If you divide the total voltage by the number of LEDs you don’t have enough LEDs for them to be anything except white or blue, or else something is going to be making smoke somewhere. If there are different colored LEDs present then there must also be dropping resistors, and then you’ve also got the issue of trying to get the brightness levels to match between different types of LEDs, which means either choosing your LEDs very carefully or driving each color from a different circuit.
In light of the fact that this is for a costume for a 30-month old child, and it only needs to be powered for roughly an hour, I was trying to come up with the best balance of simplicity, performance, cost, and efficiency. IMO, using parallel/series combinations of LEDs, and using an LM317 + 1 resistor as a current regulator, offers the optimum in terms of simplicity, performance, cost, and efficiency:
Simplicity: Only two components for current regulation. (A single resistor would be the simplest, but performance suffers.) And a high-voltage battery or DC-to-DC converter is not required, as would be if the LEDs were all connected in series.
Performance: Better than a single resistor, since the brightness of the LEDs will remain constant as the battery energy is depleted (down to a threshold voltage). Of course, better performance and reliability can be had by regulating the current for each set of LEDs, but at greater cost, complexity, and (maybe) less efficiency.
Cost: A resistor is only $0.20 at RS. And an LM317 is only $2.29. While not as cheap as a single resistor, it’s still very cheap in the big scheme of things.
Efficiency: Good, since the voltage drop across the current regulator is pretty low. Even better efficiency can be had by using a switching converter, but at greater cost and much greater complexity.
If this were for a consumer or industrial application, the design would be very different.