if it's a decent power supply, it's 3.3V rail is alread filtered and supressed with good size caps.
Also most electronics (but not all) have a 10% variance rule. You are just inside that at 3.3V. But I wouldn't be surprised if you got a premature failure.
Don - Specializing in layout DC->DCC conversions
Modeling C&O transition era and steel industries There's Nothing Like Big Steam!
A diode, which is what an LED is, will pass only a certain amount of Voltage. If the supply voltage is higher, it's usually not a problem. That's why diodes can be use as rectifiers and in place of resistors in some cases. The current that an LED can tolerate is what burns them out. If your power supply has a 10 Amp capacity and the LED is rated at only 30 mA, you need to put resistors in the lines to reduce the current flow.
Use Ohms law to determine the resistor value needed: R=E/I. Resistance in Ohms = Voltage (E) divided by Current (I) in Amps.
For a 3 Volt LED that draws 30 mA with a 3.3 Volt power supply, the calculation would be R=0.3 divided by 0.030. You need to drop 0.3 Volt at 30 mA, which would yield a resistance value of 10 Ohms.
scubaterry wrote:I got several great informative respones on the other thread. The consensus was that it would be prudent to put a 27 ohm resistor in series with each LED. Well the whole reason I choose this particular power supply was so I did not have to mess with resistors. I havd several hundred 1k ohm resistors so I could use the 12v power supply output if resistors where in fact going to be a requirement. Any other thoughts on this.
What happens if you put two LEDs in series?
If the light - and that is a big IF it would be a very safe solution.
Two LEDs in series: Add the Voltages; i.e., two 3 Volt, 20 mA LEDs would need 6 Volts at 20 mA.
Two LEDs in parallel: Add the Amperage; i.e., two 3 Volt, 20 mA LEDs would require 3 Volts at 40 mA.
LEDs are current-driven, not voltage driven like a regular light bulb. The voltage rating is the forward voltage drop in the diode, the minimum voltage needed to operate the LED. And the current rating of the LED is the absolute MAXIMUM current it can handle. Given sufficient voltage to pass the diode junction, the LED will run away with no current limiting resistor and quickly be destroyed.
Most white LEDs are 3.1-3.5 volts, so the 3.3v output of the power supply might not be the best choice. You're probably better off usign the 5v output with a 150-180 ohm resistor to limit current to around 10ma.
There are 5v and 12v LEDs, usually in the color varieties - these are just regular LEDs with the resistor already included to operate safely ont he rated voltage.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
nfmisso wrote:What happens if you put two LEDs in series?
Gandy Dancer wrote: nfmisso wrote:What happens if you put two LEDs in series?They each take about 0.7 volts and pass as much current as possible just like two normal diodes would. If the light - and that is a big IF it would be a very safe solution.?What!? Safe from what?
LEDs arctually drop (usually) between 2 and 4 Volts. Since they are current driven, rather than voltage driven, any solution that does not allow for control of the current is most likely going to result in their early demise. In any particular case, I don't think anyone knows what that really will mean.
Jeff But it's a dry heat!
Chip
Building the Rock Ridge Railroad with the slowest construction crew west of the Pecos.
One each is better, at least for me, because I change my mind more often than I change my socks! But, technically, one resistor would work. For 12 LEDs, you could try a 47 Ohm, 5 Watt resistor. (I think 2 W is enough, but it's real close, so I'd go). You are going to be disappating a the power over one resistor rather than twelve. I wrote a bunch of stuff about why a resistor for each is better (and it isn't that hard to do), but I think you've heard all that. If you use one, just remember that if you remove or add LEDs to the circuit you are going to have to think about it every time, with individual LEDs there is no thinking required. I think that's an advantage!
scubaterry wrote:Well it is obvious that I need to do something different to prolong the life of my LEDs. So much for trying the easy way. How about this, rather than going in and installing a resistor on all of the LEDs can I just use one resistor out of the pwr supply if the type and # of LEDs are the same. Another words this particular section I am talking about has 12 LEDs and I don't anticipate anymore. Can't I use just one resistor to limit current to all LEDs. Also when I converted the Pwr Supply I also included a terminal strip for the 12v supply. If I use the one resistor method would it be better to use the 12v? Thank you for everyones help in this. Terry in FLorida
Not recommended, if you have bunches of LEDs in parallel, the current drop will change as you add or remove some, meaning that single resistor will have to change. If you use basically the same LEDs everywhere, you cna buy the resistors in 100 packs (or larger), and they cost about a penny each. Especially for structure lighting, make each building self-contained with just a + and - power supply hookup, and each LED inside with it's own resistor. Then it won't matter how many structures you have, to add a new one, you just hook up the proper power supply wires and go.
As always you folks were very informative. I haven't began mylayout yet and will not for another two months or so. In theinterim I am building the structures to give me somehing to do. Ijust completed scratch building the Swift Soybean Complex in the MR"realistic operations" special issue that came out several monthsago. I will go back and rewire the LEDs for a single resistor LEDas rec. It will be pretty easy as All the roofs are removable andI used a two wire bus along the ceiling of each bulding to attach/wirethe LEDs. I also think I will switch to the 12v pwr supply outputrather than the 3.3v.
Again thanks for everyones help and advise.
Terry from FLorida
If you go with individual resistor per LED (recommended), consider surface-mount (SMT) resistors. Photo shows 5mm LED with an 0805 (0.08" x 0.05") SMT resistor.
For 3.3V, the power dissipated in the resistor will be negligible so 0805 will work fine. SMT resistors cost the same as regular resistors but you may have some trouble finding just a few - though the electronics chain store Frys on the west coast has them in hobbyist quantity packs.
I would stay with 3.3V since at 12V you'll burn more power in the resistors than in the LEDs but I guess you already have the resistors for 12V and the space.
Unetched copper clad board (e.g., Radio Shack) can be scribed with a hobby knife to mount the parts. I find using boards makes it easier to mount LEDs to walls, ceilings, whatever. I find it takes some re-positioning to get the right lighting effect so a small Nd magnet can help. An unbent paper clip attached to a ceiling allows something akin to home track-lighting to slide the board back and forth. These tiny but amazingly strong magnets are less than a dime each (e.g., http://www.kjmagnetics.com/products.asp?cat=10).
A long narrow scribed board with SMT components can make easy work of a strip of lights since the copper ‘islands' become zero height wiring which is of course the point of printed-circuit boards. Photo shows this technique to make a thin multi-LED strip to model the catwalk on a billboard but works for outdoor building sides, station platforms, and so on.
I find a slight coating of Tamiya clear-orange paint demotes the blue-tint of white LEDs. It makes the color appear more incandescent-like if that's what you're after. This is not needed if you have so-called warm-white LEDs.
Finally, I recommend getting a digital meter. If you watch the ads, this Harbor Freight meter http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=90899 regularly goes on sale for less than $3. Obviously it's not professional grade, but is suitable for simple hobby tasks. I'm pretty sure Harbor Freight is nationwide.
HTH
You can get 0805 SMT resistors for Digikey, 50 for $2.07. I'm sure shipping would be more than the parts. If you use the 3.3V supply, and since you are using white LEDs, the resistors should be pretty small, like 20-30 Ohms. Your supply voltage is really close (maybe even less) than the typical forward voltage drop for a white LED (at least the ones I looked at), so you are really just trying to limit current.
I was thinking some more about how to explain why you need the resistors, even with the supply voltage so low. The horse is probaly dead, but I want to beat it some more. If the LED drop is a half a Volt less than the power supply voltage that half Volt has to be dropped somewhere else. What that means is that the small resistance in the wires, connections, and even the power supply itself all get added up, and enough current has to flow to drop that voltage. If the resistance is small, that current could be very high. By controlling that resistance (using a resistor), you can limit that voltage. Now he goes to the glue factory!
I know we all look for shortcuts for the parts of this hobby that we don't feel comfortable with (at least I do)! And a lot of them are good. Sometimes, not so much! But keep looking for the good ones, 'cause I'm really lazy!
I'm going to go against the grain here and suggest that since the LEDs haven't blown yet, leave 'em alone! The power supply being filtered and regulated seems to be stable, as the LEDs are still operating, (no start up surge). As LEDs are rated for 100k hours or more, the few tenths of a volt over their rating is negligable and may shorten their life by a few thousand hours, are you still going to be alive to care?
Electronic theory is one thing, the real world is another.
Jay
C-415 Build: https://imageshack.com/a/tShC/1
Other builds: https://imageshack.com/my/albums
modelmaker51 wrote: I'm going to go against the grain here and suggest that since the LEDs haven't blown yet, leave 'em alone! The power supply being filtered and regulated seems to be stable, as the LEDs are still operating, (no start up surge). As LEDs are rated for 100k hours or more, the few tenths of a volt over their rating is negligable and may shorten their life by a few thousand hours, are you still going to be alive to care?Electronic theory is one thing, the real world is another.
Actually, at 3.3V his supply is probably below the voltage rating on the LEDs, but LEDs are current driven animals, and there's no deliberate control of the current here. It could well be they will last for 100s or 1000s of hours or more, it's pretty much unpredictable. If I was starting from scratch, I'd use a 5 V supply with a resistor on eash LED, and know what I had. I guess since we know these LEDs are not drawing a huge amount of current we could stick an ammeter in and see what's happening, and then decode based on that whther to make a change or not.
Vail and Southwestern RR wrote: modelmaker51 wrote: I'm going to go against the grain here and suggest that since the LEDs haven't blown yet, leave 'em alone! The power supply being filtered and regulated seems to be stable, as the LEDs are still operating, (no start up surge). As LEDs are rated for 100k hours or more, the few tenths of a volt over their rating is negligable and may shorten their life by a few thousand hours, are you still going to be alive to care?Electronic theory is one thing, the real world is another.Actually, at 3.3V his supply is probably below the voltage rating on the LEDs, but LEDs are current driven animals, and there's no deliberate control of the current here. It could well be they will last for 100s or 1000s of hours or more, it's pretty much unpredictable. If I was starting from scratch, I'd use a 5 V supply with a resistor on eash LED, and know what I had. I guess since we know these LEDs are not drawing a huge amount of current we could stick an ammeter in and see what's happening, and then decode based on that whther to make a change or not.
A computer power supply regulates both voltage and current for each voltage tap. That 3.3v tap is most likely the supply for a computer's indicator LEDs for Power and HD. Most everything else in a computer uses either 5+- or 12+- volts.
But what's it regulating the current to? Is he lucky, and it's a 250 mA output, and he just happens to have 12 LEDs, and wants 240 mA. If he had one LED, Poof! And if he had 50, no light? I believe that there's a lot more than LEDs hooked to 3.3 V in newer computers. The ratings that I saw in doing a little research showed specs of 15 or more Amps. Anyway, my point is that with no resistor there is no positive limitation of the current relative to the LEDs. The supply does not know what's out there, or what they want. I'm thinking he may be saved by being a little below the forward voltage of the LEDs. They are not fully on, so they might be limiting the current themselves. Been way too long since I though about such, though, so don't quote me!
scubaterry wrote:...I would like some more info on your Bill board light set up if you don't mind. The pic was kind of small to see any detail. I just built a couple of bill boards but haven't decided how to do the lighting yet and I like the way yours looks...
This is from pix I had uploaded for another forum so there's some extraneous stuff but should give you a better idea:
The modeling problem is the catwalk is very thin and wires between LEDs would be unsightly. The scribed board is not the exact board used in the billboard, but it shows the concept of a long board. In this case the board used SMT white LEDs (available from DigiKey). Thin styrene sheet with holes for the LEDs can cover the scribed board. The particular board shown with the child has LEDs in parallel with a single resistor protecting the entire strip. But that horse has been declared dead so I won't go into that.
Since you have time, request a DigiKey catalog thru their website; the current one is over 2000 pages and you can learn a lot about different parts just by flipping through it. They sell copper clad boards in all kinds of sizes and thicknesses; I use the ones by Injectorall. The thinner ones such as those less than 1/32" thick can be easily cut with scissors which is nice. They have a $25 minimum else a $5 handling fee. If you order just a few light parts and specify USPS First Class, postage might be under $3 as they charge actual USPS rates to the penny, and they ship instantly - usually within 1 hour! I highly recommend them.
HTH (Hope This Helps)
stan2004 wrote: scubaterry wrote:...I would like some more info on your Bill board light set up if you don't mind. The pic was kind of small to see any detail. I just built a couple of bill boards but haven't decided how to do the lighting yet and I like the way yours looks...This is from pix I had uploaded for another forum so there's some extraneous stuff but should give you a better idea:The modeling problem is the catwalk is very thin and wires between LEDs would be unsightly. The scribed board is not the exact board used in the billboard, but it shows the concept of a long board. In this case the board used SMT white LEDs (available from DigiKey). Thin styrene sheet with holes for the LEDs can cover the scribed board. The particular board shown with the child has LEDs in parallel with a single resistor protecting the entire strip. But that horse has been declared dead so I won't go into that.Since you have time, request a DigiKey catalog thru their website; the current one is over 2000 pages and you can learn a lot about different parts just by flipping through it. They sell copper clad boards in all kinds of sizes and thicknesses; I use the ones by Injectorall. The thinner ones such as those less than 1/32" thick can be easily cut with scissors which it nice. They have a $25 minimum else a $5 handling fee. If you order just a few light parts and specify USPS First Class, postage might be under $3 as they charge actual USPS rates to the penny, and they ship instantly - usually within 1 hour! I highly recommend them.HTH (Hope This Helps)
Since you have time, request a DigiKey catalog thru their website; the current one is over 2000 pages and you can learn a lot about different parts just by flipping through it. They sell copper clad boards in all kinds of sizes and thicknesses; I use the ones by Injectorall. The thinner ones such as those less than 1/32" thick can be easily cut with scissors which it nice. They have a $25 minimum else a $5 handling fee. If you order just a few light parts and specify USPS First Class, postage might be under $3 as they charge actual USPS rates to the penny, and they ship instantly - usually within 1 hour! I highly recommend them.
That is totally cool! And (sorry horse) I'd say that is a perfect place to break the 'rule' and use one resistor. You know how many LEDs there are, there are not going to be more, not going to be less, the thing is a unit. Did I mention that it is cool?
PC Power supplies are usually switch-mode power supplies and don't have current limitting. Ohms law will take care of the current the LEDs are drawing. The amount of current that can be drawn will typically be determined by the ICs inside and their ability to provide current.
3.3 volts may be a little high but it depends on the LEDs, I've seen some with recommended voltages of 2.7V. If they seem overly intense, I'd say they are running at too higher voltage and they will eventually fail.
Use a multimeter to test what the power supply is putting out. It is probably putting out around 3 volts anyway, then it should be fine. As was mentioned earlier they use LEDS to show power and Hard drive status so why not give it a go!
Regards,
Nigel
Stan - Tks for the additional info. I had to laugh when I saw the HTH at the bottom. I didn't read the original post from you very well and thought it was your initials. I will def look into Digikey. I have a handful of white SMT LEDs laying around destined for my PCM 2-8-8-2. I bought extras as they are so small and my hands are soooo big I figured I would lose, step on , smush or otherwise screw up a couple in the process.
I looked on the pwr supply case and the 3.3v supply has an output of 20A. Thats a butt-load in my book. I think what I am going to do is all the bldg that are already lighted I will put one resister of appropriate size in series with the lights in each bldg. The most I have is five LEDs in any one bldg. That # will not change and if it does I will just have to recalc the resistive value. I found a handy LED resisitor calculator online. Any future bldg will have a resistor for each LED. I will def get the smt resistors. And I really like Stans idea on the Bill Board lights.
Terry in FLorida
Vail and Southwestern RR wrote: But what's it regulating the current to? Is he lucky, and it's a 250 mA output, and he just happens to have 12 LEDs, and wants 240 mA. If he had one LED, Poof! And if he had 50, no light? I believe that there's a lot more than LEDs hooked to 3.3 V in newer computers. The ratings that I saw in doing a little research showed specs of 15 or more Amps. Anyway, my point is that with no resistor there is no positive limitation of the current relative to the LEDs. The supply does not know what's out there, or what they want. I'm thinking he may be saved by being a little below the forward voltage of the LEDs. They are not fully on, so they might be limiting the current themselves. Been way too long since I though about such, though, so don't quote me!
To be clear here the power supply regulates against over current conditions and is not a current regulator. A true current regulator power supply would adjust the voltage (within the limits of the supply) to maintain a specific current output. These are rarely used in electonics today.
Engineer Jeff NS Nut Visit my layout at: http://www.thebinks.com/trains/
jbinkley60 wrote: Vail and Southwestern RR wrote: But what's it regulating the current to? Is he lucky, and it's a 250 mA output, and he just happens to have 12 LEDs, and wants 240 mA. If he had one LED, Poof! And if he had 50, no light? I believe that there's a lot more than LEDs hooked to 3.3 V in newer computers. The ratings that I saw in doing a little research showed specs of 15 or more Amps. Anyway, my point is that with no resistor there is no positive limitation of the current relative to the LEDs. The supply does not know what's out there, or what they want. I'm thinking he may be saved by being a little below the forward voltage of the LEDs. They are not fully on, so they might be limiting the current themselves. Been way too long since I though about such, though, so don't quote me!To be clear here the power supply regulates against over current conditions and is not a current regulator. A true current regulator power supply would adjust the voltage (within the limits of the supply) to maintain a specific current output. These are rarely used in electonics today.
Yes, what I was trying to say. The power supply is giving those LEDs all the current they will take at 3.3V. Since they are white, their forward rated voltage may well be higher than that, so they are current limiting themselves, sort of, which may be what is saving them. I'd much rather use the 5V output and have things under my own control through choice of resistors. (By the way, Terry, the idea on using one resistor per building isn't awful.)
Jeff
(By the way, Terry, the idea on using one resistor per building isn't awful.)
Great response! Are you a politician? Sorry that was low
I have three other computer pwr supplies yet to be converted. You mentioned you would use 5v. would that be a better voltage than 3.3 because it gives you more control over the LED current? IF so then I need not mess with the 3.3 on the remaining units. Tks again.