Large Scale Central

LED help

Electronics
#21

Pete Lassen said:

why not just go to your local Wal Mart and use the real small strings of lights they have that can be either battery powered or hooked to your power suppply? Maybe it would help with your brain flashes? Wish I knew more about how LED’s worked I need a electrical lesson, and not the touch this see if we have power kind of lesson

They are actually to big for O scale. I have a couple sets of them that I was going to play with in 1:24 but the bulbs on the ones I have, while small, are quite a bit larger than these nano LEDs.

#22

Dan Hilyer said:

Pete, don’t give him an easy way out (https://www.largescalecentral.com/externals/tinymce/plugins/emoticons/img/smiley-foot-in-mouth.gif)We’re supposed to make it harder for him, that’s what friends are for (https://www.largescalecentral.com/externals/tinymce/plugins/emoticons/img/smiley-wink.gif)Like, I could tell him to go get an arduino and some nano pixels and program them himself so he could have varying colors and themes but I wouldn’t do that to him … or would I (https://www.largescalecentral.com/externals/tinymce/plugins/emoticons/img/smiley-tongue-out.gif)

Okay damn it I had to look. . .I hate you

So, I have no idea what I am looking at but it looks very cool. And when we are talking Christmas lights I am thinking a mini LOR light display

Now how cool would this be on a layout

#23

Pete Lassen said:

why not just go to your local Wal Mart and use the real small strings of lights they have that can be either battery powered or hooked to your power suppply? Maybe it would help with your brain flashes? Wish I knew more about how LED’s worked I need a electrical lesson, and not the touch this see if we have power kind of lesson

Maybe we should do a thread on basic low voltage electrical theory.

An LED is a diode, a one way gate. Power will only flow one way through the thing, so it is polarity sensitive. The voltage drop across the thing is a curve, but its assumed to be constant for simplicity. So if it drops a given voltage, at something close to its rated current, it will give off light. So if we have a 5 volt supply, and the LED voltage drop is 2.4 volts, that means we need to loose, or drop, 2.6 volts across a resistor. (2.4 + 2.6 = 5) We know we want 20 milliamps (.020 amps) flowing through the LED (if that is its rated current). Current being the same in all points of a series circuit, we know we want .02 amps flowing through a resistor dropping 2.6 volts. Ohms law says that voltage equals current times resistance. We know the voltage we want across the resistor (2.6 volts) and we know the current we want (.02 amps) so we divide the voltage by the current to get…2.6/.02=130 ohms.

Clear as mud? Right?

Personally I would go a bit higher on the resistor value, its better to under-drive an LED then to overdrive it and burn it out.

#24

Can we all just agree that we give Devon 24 hours and then he will wonder off into another topic… :wink:

Sorry buddy, I just couldn’t resist.

#25

No apologies, I agree with that statement myself. I am already distracted by arduino. Its called adult onset ADD. . .and putting a 24 hour time limit on it with the MIK challange only a few hours away is insider trading.

#26

I have a similar problem Devon, but I blame my scattered brain on the kid chaos…

#27

I have read this thread a bit, and not in detail, so some of what I say may have been “corrected” already.

  1. Do NOT depend on the limitations of the power supply to supply “safe” current or voltage to the LEDs.

  2. unless you have a current limited power supply (and I would wager you do not, since this usually only happens with a laboratory supply - I can explain), do not take a “100 ma supply” and calculate the number of leds it can drive.

  3. if you want to run leds in parallel, the best way is a current limiting resistor each, gets to be a lot of resistors, but they can be small… if you try to do it with a single resistor, it can get big and hot.

  4. the most economical way (wiring and cost) is to use as many in series as possible, it will use less current and a single dropping resistor, and with a fair number of LEDs, and careful calculations you can sometimes not use the dropping resistor… not a great idea, but can be done.

Hope this helps.

Greg

#28

Greg Elmassian said:

Hope this helps.

Greg

No Greg. . .it doesn’t help. . .not one bit (http://largescalecentral.com/externals/tinymce/plugins/emoticons/img/smiley-tongue-out.gif). See that is where I was headed at first and then all this talk convinced me to run in parallel. So you have taken me back to square one. BUT even though I couldn’t get the lady at Evans to say it, when I looked at the nano string they sell IT IS in series with one resistor.

What you describe is exactly what I was originally thinking. Or at the very least a couple of stings in series with each string being in parallel with the others.

#29

Hurts my brain all that math and such.

I use these: https://www.mouser.com/ProductDetail/689-CL2N3-G

They go in my battery locos @ 14.8v. I also run 12v around my layout for my buildings and street lights and use these.

I buy them 100 at a time. They are magic. No math required. I like that :slight_smile:

Dave Bodnar wrote about these way back when: http://www.trainelectronics.com/LED_Articles_2007/LED_104/index.htm

#30

the cl2 is very useful, since it will work up to 90 volts. So the idea of one per led is nice, simple, no calculating of resistor, no matter what your supply voltage is.

Also a string in series is likewise simple…

I jumped on the thread because some of the posts were clearly in error. There are some ways you can “do it wrong” and it might work sometimes, and might fail at some random time.

The cost of the components is negligible, the drivers are the work and like you say, wires everywhere… but in series, like 10 houses, one wire in one wire out, that’s simple.

Greg

#31

Am I understanding that the driver replaces the resistor? And if so can a single driver run several LEDs in series?

We need to get back to the original plan. While LED theory is nice, I am a dunderhead when it comes to this stuff. I don’t know why as I consider myself to be reasonably intelligent. I know at some point while messing with them it will sink in. But I do have some sort of block when it comes to this so dumb it down for me.

I want to run a string of say 15 or so pico or nano LEDs around a tree to make it look like a string of Christmas lights. Having some dohicky like a resistor attached to Every LED will mean a heck of a lot of “stuff” that won’t look very pretty in my tree. Or a pile of wire going somewhere else and attached to all the dohickies. I don’t see that as at all practical. Now if a single resistor or driver will run say five (or even all 15) with a given power source then I could see them in series or even in parallel and having the dohickies elsewhere. Either way, in series or in parallel, running 3 strings of 5 or even 1 string of 15 would look much prettier.

Now with all that everyone has said, It seems that if a given LED runs at 2.4v and 20mA and I have 12v/100mA source then theoretically I could run a string of 5 LEDs with no resistor in a series. Now I understand that adding a small resistor would add protection but lets leave that out for a second. Each LED being a diode and not a light bulb I understand that I start with 12v and in series knock down the voltage by a factor of 2.4 at each with each LED, so 12-2.4=9.6-2.4=7.2-2.4=4.8-2.4=2.4-2.4=0. And since I have 100mA and each needs 20mA then 20mA X 5LEDs=100mA of required current. If I have this correct then I further assume that the first LED in the series is not really receiving 12v which would burn it out but that the collective of [email protected] “evens out” the voltage to be 2.4 across the series? Do I have that at least sorta right?

Now lets work on parallel. I will use the same LED at 2.4v 20mA. No matter what a parallel circuit maintains the given voltage, right?. So regardless of 1 LED or 100 LEDs at the end of the run I still have 12v. Because the LEDs are all wired singly to the positive wire they won’t rob voltage from the next one because they are not really in line. !2v is carried across the entire positive feed clear to the final LED, right? As far as voltage is concerned they do not “see” the other LEDs in the string. Each one gets 12v. But wouldn’t current still be a factor and 5- 20mA LEDs would still require 100mA source to provide enough current. If I understand this then each LED would need a resistor so that it only sees 2.4v and only 20mA of current instead of the full 12v/100mA being provided at the feed?

Yes, no, sorta? If I have this anywhere near right then It would almost seem to make sense to me if I want say 15 LEDs running off a single 12v source then a 300mA source could be used to make a hybrid system of three (3) 5 - 2.4v/20mA Leds strings in series with each if those strings run in parallel would make sense. Each string would use up the 12v and 100mA and since each string is in parallel to the other each string will see 12v and each string will take 100mA of the available 300mA, right?

#32

Yes, one driver can run a string of LEDs, if the voltage is sufficent.

Current is the same in all points along the series part of the circuit, so 5 LEDs in series will not draw 5 times the current of one LED. So one driver, limiting the current to 20 ma, will keep the current in that “string” at 20 ma. That is what you want to happen.

Yes, the voltage to each parallel “leg” will be the supply voltage.

Yes, if the voltage across each LES is 2.4 volts, then 5 in a string (in series) should be able to be supplied with 12 volts and each LED gets the 2.4 volts it wants, provided that the LEDs are identical.

The max current of the supply only comes into play when figuring the total load on the supply. 5 strings of LEDs, drawing 20 ma each would be the max for that supply. I would personally not run a supply at its max rating. The supply will not “push” out 100 ma into one string, because that is what the supply is rated for.

#33

Thanks David,

So my drawing would work with one major difference in that since current it NOT cumulative what ever voltage I run with it needs to be 20mA (larger so I am not driving it to full capcity all the time but for understanding we will go with the current rating of the LED)? Should my drawing look like this:

Instead of resistor lets insert driver. that sounds like a better way to go.

Wait after rereading I am confused. I understand each string in series will have the same current demand as the individual LED in that string. So if my LED is a 20mA LED then the string as a singularity would be a 20mA string NOT a 100mA string. But now what about each parallel string on the total load? If Each string as a single unit has a current draw of 20mA do each parallel string become cumulative? or does over all current demand stay the same as the individual string? would my power source above need to be 12v/90mA or 12v/30mA?

#34

Another question about the drivers themselves.

It gives a very large range of operating voltage. And my intention is to have my main feed under my layout be 24v so that I don’t lose current over the lines. Wire has resistance so the longer the wire and the smaller diameter the wire the more resistance and loss of current right? Don’t you over come this by 1) increasing voltage 2) increasing wire size and 3) providing excess current and then at each “sub station” drop the voltage, decrease the wire size (if you need too) and drop the current? Isn’t this how the 1:1 power grids work?

If that is correct an I am running 24v and these drivers handle an input of 24v and spit out the correct voltage for the LEDs and I can run 5 LEDs on each driver; then Wouldn’t it be better to maintain 24v through out my entire grid and then just let the driver do its thing and not worry about a substation transformer that takes the 24v and make it 12v?

#35

Martin Sant said:

Dave Bodnar wrote about these way back when: http://www.trainelectronics.com/LED_Articles_2007/LED_104/index.htm

God bless Dave. I have learned much from him. But he makes my head hurt. He operates on an entirely different frequency than I do when it comes to this stuff. I feel like a third grader trying to learn the solar system from a professor of astrophysics. That’s a complement to him not a dismissal. Its not that he can’t explain it, but when it comes to trying to explain it to me you have to sink to crazy ridiculous low levels. It really does bother me that I pick up fast on most everything but these electrical gizmos and I don’t mesh.

Thanks for the driver info. I think they look like they make way more sense than a resistor. Why drive a model T when you can have a 2020 cybertruck. I can see where resistors and their calculations are old school technology and the LED driver is purpose built new technology to specifically overcome our problem.

#36

WAIT WAIT WAIT!!!

Did I just read in Mr. Bodner’s write up that LED voltage is a minimum and NOT a maximum as well. I knew that LEDs needed a minimum voltage to operate; so a 2.4v LED needs 2.4v min to make it produce light. But will the same LED run with an unregulated 24v supply provided it does not exceed the 20mA of current?

If that’s the case then I think I see much more clearly when it comes to protecting the circuit and the use of these drivers.

If I am getting it through my fat head then It would actually be better to keep my 24v through out and use the driver and then I could run a bunch of LEDs (10@ 2.4v or 8 @3v) or 1 because the voltage doesn’t really matter as long as there is enough. as long as the driver is taking that input current (whatever it is) and dropping it to 20mA.

#37

each parallel string become cumulative?

Yes

Ok, wire losses come into play with long runs, small wire, and high currents. For LED lighting, wire losses would be minimal.

You are a water person. Voltage is like water pressure, current is like water flow. You don’t boost the flow along the way. So get your mind off current flows the way you are trying to think of them. If you have a water line with 20gph flowing through it, it doesn’t suddenly become 50gph 50 feet down the line. If you have 2 water lines side by side, each with 20 gph flowing through them, then your water source would have to supply 40 gph to sustain the flow in both of them. That is ok, since you water supply can supply a maximum of 100 gph. So in theory, you could run 5 separate (parallel) water lines from your water source, each drawing 20 gph and the source would handle it, but it would be running at its maximum.

LEDs, run at a given voltage and current. supplying them with a higher voltage and current will make them into flash bulbs. If, you push 20 ma through an LED, the ones you have will have a voltage drop of around 2.4 volts. If you hook them up to a higher voltage, without a current limiting device (resistor or driver) then you will push more current through them and burn them out. More voltage is more pressure, more pressure means more flow, more flow means more power and poof.

#38

Devon Sinsley said: I want to run a string of say 15 or so pico or nano LEDs around a tree to make it look like a string of Christmas lights.

Michael’s craft store has them on sale/discounted big time right now/dirt cheap and already figured out the wiring for you.

You just have to gut them!

#39

Rooster ’ said:

Devon Sinsley said: I want to run a string of say 15 or so pico or nano LEDs around a tree to make it look like a string of Christmas lights.

Michael’s craft store has them on sale/discounted big time right now/dirt cheap and already figured out the wiring for you.

You just have to gut them!

Funny you mention this.

As i was digging through my junk scrounging pieces and parts for the MIK I came across a string of Nanos with a 100 LEDs in the string. It has a controller to make them flash, fade, swich colors, stay stead on. They have an adapter for 110. Reading the adapter its output is 3.3v and 1A (1000mA). Throw the adapter away and provide a 3-5 v supply (what the controller is rated for) at 1000mA and viola I have a hell of a string of O scale Christmas lights. I must have bought these on sale last year, I am thinking my local Ace Hardware. I thought all I had was a string of 3mm. But this are tiny, like 1mm surface mount, with a some clear plastic covering them to make them bulb like. Absolutely perfect. I am going shopping. While its fun to explore I see no need to reinvent the wheel. Hell why even gut them. Run a 110 feed under my bench instead of 24v and then at each place knock it down with a transformer. 110 to low voltage adapters are a dime a dozen at thrift stores in the way of old electronics like phone chargers, cordless, phones, and so on and so on. I have about 10 of them in an electrical junk drawer. And then I would never have to worry about having ample power. A 50 foot extension cord and some J boxes and outlets under the bench. I was thinking all these little strings were made with “huge” 3mm LEDs. my wife has a ton of them. Then I saw mine and yeah baby

#40

Ok, I will go back to sleep now.