What am I doing wrong?

[Thewrathofodysseus]

Ok. I bought one of those mini strings of led Christmas lights at the dollar store, the ones that run off two double As. I ripped it apart and wired my Kelvin saucer with five of them. I wired them in series and tested each one with a nine volt to make sure they worked and I was connecting the correct post of the led to the next led. Carefully measured the wires and soldered them all together in series... now only three will light up at a time. Doesnt matter where I test connection only the three will light. I then tried using the battery box with the two double As that the LEDs were originally connected to, and that wont even light one led. I then tested each led individually and the all work. What have I missed?

[Zubie]

I presume the situation is the following

(+)----✿---✿---✿---X---X---(-)
✿ lit, X unlit
where the easiest explanation would of course be the 4th one is wired wrong. Only other thing I can think of is poor contact under the solder on the 4th. If it's another pattern, I'm personally not sure what it could be.

[Ant]

I think that's because they were wired in parallel, but you have now wired them in series. This places different voltage/current demands on the circuit.

[Kylwell]

Did you have a resistor on them when you tested them with a 9 volt?

[naoto]

Hrm...
* Two AA batteries would give you 3 volts ( 1.5V x 2 ) -- not sure if it's a good idea to be running those LEDs at 9V. If anything, you'd probably want to add a bigger resistor to further limit the current (due to applying the higher voltage).
* In general you'd want to be using a resistor in series with LEDs to limit the current going through them
* Not sure if the individual LEDs you extracted from the string light have built-in resistor or not. Furthermore not sure how they were wired up originally (i.e. were they in series, or were they in parallel, or some combination).
* Might also want to check if these string lights have a "wink" or other effect -- if there is some sort of effect, then you may need to trace where that effect is being implemented (i.e. is the wink circuit within the battery pack, or perhaps a "nodule" on the string, one of the LEDs may be acting like a "master" to cause the rest to blink, etc.).

[Thewrathofodysseus]

I'm reasonably sure they were in parallel. No resistor attached. Connections were clean. All les posts were facing correctly.

[naoto]

You might want to check the original circuit -- see if the current-limiting resistor could be found there (perhaps it's in the battery box or the cabling). If there is no such component - then perhaps the current-limiting resistor is integrated into the LED. In that case writing them in series (when they were originally wired up in parallel) would mean that you probably have too high of a resistance, resulting in insufficient current to light up the LEDs.

[Ant]

What usually happens with many low cost mass produced items such as these battery operated strings of LEDs is they rely on side effects & physical properties of components within the circuit to allow the elimination of what would otherwise be required components for a high quality solution.

Typically this involves taking account of the internal resistance of the chosen battery source to prevent over-current destroying the LEDs that would be provided by series resistors in a "quality" circuit, and also when connected in parallel with no series resistance, then relying on the conduction properties of the semiconductor die of each LED to be virtually identical - something not at all guaranteed across different production runs of the same device.

The rule is: every LED in parallel needs its own resistor, unless you can explain to an electronic engineer why you don't (and yes, there are exceptions). If you can't explain, then you need a resistor.
When you do get a minute difference between LEDs it can lead to a cascading failure that can take out all the LEDs. It may not happen immediately. It may not happen for months. But it may happen once your model is sealed up and finished. Is that a risk you want to take?

It's a bit like checking for traffic before you cross the road, vs crossing with your eyes closed. Do it right, and you will always succeed. Take a risk and sometimes you'll be fine. But now and again, you'll be dead.

As for the series circuit not working. Each LED needs voltage across it to function (it's "forward voltage"). For a white LED this is around 3v. So 2 LEDs need 3v each, or 6v. 3 LEDs need 9v. 4 LEDs need 12v... and therein lies the problem.

[EVApodman]

I have modified the lighting circuit for my K'tinga cruiser. to fit my needs. All the lights on a particular string are wired in parallel with a single current limiting resistor on the connector. This means that all the LEDs have source voltage across them but minimal current per LED, yet they seem to work.

For my lighting I have replaced some LEDs and wired them in series with a resistor designed for a 9v power supply. One thing I found out was that as I increased the number of LEDs in series the voltage required to turn them on increased. They were wired correctly with each one tested individually with a DMM. I'm guessing the voltage drop across each LED limited the available voltage for the others, so I limited myself to only 3 LEDs in series.

Another thing I noticed was that standard LEDs had a lower turn on voltage than other ultrabrights I was using in the circuit. Probably what caused the aforementioned effect.

I don't know exactly what is happening with your circuit without using a meter, only that whether wired in series or parallel you need a current limiting resistor if your power supply is above 1.5 volts. I also think that some of your LEDs have been damaged and will have to be replaced.

Based on what I have seen in the lighting kit I now believe that LEDs are voltage and not current driven seeing that in some strings each LED only gets a few milliamps of current while having full source voltage across it,

[Thewrathofodysseus]

I figured it out. I rewired in parallel and it works now. Strange thing though, the nine volt lights everything up, but the original two double A power pack will not. I’ve got five leds wired together with four, possibly five to go. Btw I checked the power pack for anything that remotely resembled a resistor and found nothing.

[Zubie]

I figured it out. I rewired in parallel and it works now. Strange thing though, the nine volt lights everything up, but the original two double A power pack will not. I’ve got five leds wired together with four, possibly five to go. Btw I checked the power pack for anything that remotely resembled a resistor and found nothing.

The thing to keep in mind when wiring in series or parallel is that the relationship between the battery and resistance to the circuit is different just like the voltage drop across the LEDs is also different. I still think it is important to have a multimeter handy to sort of understand what is going on, not just base it on whether things light up or not.

In series, Rtotal=R1+R2+R3+...+Rn. Across any individual R, the voltage drop is a fraction of the total V. If all R is the same, then Vi for any Ri is equal to V*(1/n) where n would be the total number of resistors (i.e. whatever is doing work in your circuit), and i is any given individual R in that set (Vtotal=V1+V2+V3+...+Vn).

In parallel, Rtotal is more complicated as it is an inverse relationship formula which is
1/Rtotal=1/R1+1/R2+1/R3+...+1/Rn
The voltage drop over any given one is the same for all, but this will require that the current increases over the circuit as the relationship of V=I*R remains true. The total R for the whole circuit in parallel could end up quite low and drain your battery quite quickly or run too much current through each LED-R combination (I'm kind of assuming the R is somehow built into each LED).

I've lived in quite a few old mill towns in NE, and the classic analogy is canals for powering water mills. You can run a single canal to power several water wheels one after the other (series), but then each wheel will have it's own individual drop (Vi) from the high water level to the town's low water level (V total). You could also drop multiple feeder lines off the main canal so that each wheel can take advantage of the total drop (parallel), but then you are going to need a lot more water to keep all feeder canals fed with water to turn those wheels (and if you were to add the width of each of those feeder canals together you get a sense of how much extra water is needed to go into them).

So if you understand the actual resistance in each LED element, how much voltage and current they were designed for in the original circuit, you are much more likely to re-use these elements in a way that will last as long and as effectively as originally intended. This is particular to your situation since I doubt you will find spec sheets for repurposed LEDs as opposed to buying them from a parts shop.

[Thewrathofodysseus]

I have a very small, simple multimeter. How do I check the resistance for each led? Put the probes on each lead and turn the dial to the omega symbol? Then what? Each string has eight leds, two were damaged, and I need five more lights throughout the model. I was planning on wiring a nine volt clip and switch in the base. Ordering individual components from a supplier isn’t really an option for me, and I won’t buy a a light kit unless there isn’t an DIY option, and even then it has to be really impressive. My “hobby” money is limited, and with a few exceptions, the kit and accompanying paint is usually is the extent of my expenditure. I’ve been haunting this board since before I even had a computer back in the nineties, remember WebTv? LoL. So I’ve always tried to be as frugal as possible cannibalizing as many dollar tree toys, led reading lights, and happy meal toys as needed to complete my projects. As of this writing, my wonderful wife surprised me with a Battlestar Pegasus for my birthday (a model I never thought I’d be able to get) so I am already watching videos to see how others have done theirs. After I finish the Kelvin, next up is the Nu Cylon raider, then that big beautiful Battlestar! What I like about these LED kits is they’re dirt cheap, and the ends are encased in a very thick clear cylinder, with an inverted cone on the end that throw a lot of light. Additionally, they’re so thick, I can drill into them to directly and insert fiber optics. As well as sand them to fit into tight spaces or scuff/frost them to diffuse the light. To be honest, I’ve only recently started to pay attention to the math behind the electronics, it’s always confused me, and been not fun. I had a vague understanding of wiring and series and parallel circuits (I literally watched a video to fix the Kelvin). But the important thing is it’s become much more enjoyable. Funny how perspective changes as you grow older. Please keep the suggestions/tips coming, I appreciate ALL the feedback, I’m learning a lot.

[jgoldader]

In a lot of cases, those kinds of accent light strings have resistors wired to each LED. If so, as long as you use only the rated voltage (e.g., 2x AA) and wire them in parallel, (so the wires on each LED are touching both leads from the battery pack) then you don’t need any more resistors. If you wire them in series (bat->LED->LED->LED->bat) then you’ll need more voltage, but you might also need to add an extra resistor (see below). Say a single LED needs 2.5V to turn on; in series, the total voltage needed is the sum of all the individual voltages, so 3 LEDs would need 7.5 V just to turn on; if you use a 9V battery, they’d be brighter than they would have been in the original string powered by 3V.

An LED is a device that has very little (nearly zero) internal resistance, but only turns on if the voltage exceeds a particular value (usu. 1-5 volts, depending on the color and composition of the LED; reds turn on at lower voltages, blues take higher voltages). Once an LED turns on, the more current there is, the brighter it gets. However, the LED also gets hotter, and they can be ruined (burned out, with charring and smoke, and I’ve heard they can burst) if there is too much current. (I generally limit LEDs to a few milliamps of current.) The way you control the current is to wire a resistor in series with the LED. The current (in amps, where a milliamp is 0.001 amps) is then (Vb-VL)/R where Vb is the battery voltage, VL is the LED turn-on voltage, and R is the resistance of the paired resistor in ohms. Larger values of R reduce the current (and brightness) and smaller values of R increase the current and brightness. The lifetime of an LED also depends on current: more current=shorter life. General rule is, make the thing as faint as you can to still achieve the desired lighting effect. There are LEDs made that can handle more current if you need.

You can’t check the resistance of an LED using a multimeter. If there’s a resistor prewired with the LED, the colored stripes on the resistor will tell you the resistance.

[Zubie]

Thanks for explaining the situation thewrath... as it makes it easier to think about understanding what you have to work with and what you may need clarification with. As jgoldader explained, LEDs are not the same as wheat lamps. They are also electronic components called diodes that have interesting electrical properties that allow current to flow nearly freely in one dir and not at all in the other. As a result the direction they are wired makes a huge difference, and the current has to be DC and steady to avoid flickering.

jgoldader's notes on cheap LED chains are also useful. When you buy LEDs directly you would know a lot of the properties of a given LED from its specification: operating voltages, maximum currents, inherent resistance properties, etc, and an identifiable anode (current in) and cathode (current out) indication on the element. In the case of cannibalized LEDs, you have to rely on discussion descriptions of how these things are generally made as opposed to tech. specs.

A simple test set up, a board to hold batteries, maybe in multiple sets (first 1, then 2, and so on), with wires on alligator clips should allow you to check in 1.5V increments (1.5v being typical battery voltage) what the minimum voltage required to get an LED to light up adequately would be.

If you want to kind of back calculate the chain you bought as far as what was expected, check the way it's wired. If the chain has 2 wires going in and out of each bulb, it is very likely wired in parallel.
┬┬┬┬──(+)
┴┴┴┴──(-) LED's in the cross connections
If the chain only has 1 wire going in and out, it pretty much can only be series
(+)──ᗒ|──ᗒ|──ᗒ|──(-) where ᗒ| represents an LED (similar to the actual symbol, check the web for what it actually looks like)
Keep in mind, that some systems may mix and match, particularly if they have effects.

I know the math for circuits can look complicated, particularly since often it isn't actually shown but instead represented symbolically in a triangle or a circle to show how each element balances the other. For each element, the most important formula would be V=I*R, where I=current, R=resistance. Current should be in amps, resistance in Ohms. Note that with these low power circuits, amps should be on the order of milliamps(mA). So if I=.002a, V=1.5V, then R sb 750ΩOhms since V/I=R (similarly V/R=I).

Now take a 10 parallel light string with an operating voltage of 4.5 volts (3 batteries). Assuming all the elements are the same, then for a current to be .002 to .005 amps, the R to draw that current across each LED should be between 2250Ωand 900Ω just using the math above.

Same set, but running in series would imply that each LED+R element should be only using .45V (V/10) and for that .002 to .005 A rating, each R sb between 225 to 90 Ohms (R total 2250, to 900 Ohms respectively, div by 10).

Using a multimeter you can check if indeed the LED elements actually seems to meet that range. A cheap one sb sufficient since at these low powers all you need are things on the order of few volts and mA. This is important because you can wire these things commercially with variations, such as 3 sets of parallel LEDs wired in series (so each parallel set handles 1.5V drop each from a 4.5 starting point) or a single R at the battery unit could handle the current management for the whole circuit making the calculations above misleading for individual elements.

The main thing, is keep your units straight. Many of the info boards I've seen mix measuring things in V with mA and KΩ which can result in simple math errors. I'd recommend keep track of the decimals and just work with Volts, Amps, and Ohms. LEDs show low Ω values in one direction, and very high Ω in the other (that "gate" property discussed). The light emission does represent a certain amount of work, even if very slight, so some impedance/resistance should be expected.

I'd also assume that no LED is eternal, so be wary of putting them somewhere where you can't get at it or replace (that was D. Adams rule about the placement of things that "couldn't fail").

[EVApodman]

Does your meter have a diode check function? This is what I use to check diodes when working with them and making sure I have wired them correctly. In the forward biased condition the voltage drop is typically between 5.5 and 7v, with 7v being the ideal condition. Reverse biased they show an open.

Also when forward biased by the meter the diode will glow dimly without needing a resistor which will tell you if it is any good and the color of it if it a clear one.

[Thewrathofodysseus]

Posted some pics on the SM Facebook page.

[Mr. Engineer]

Is there any link for that Facebook post?

Facebook 'moves' very fast

[Thewrathofodysseus]

https://www.facebook.com/groups/starshi ... 281616319/

[Mr. Engineer]

Thanks, Thewrathofodysseus.

So, assuming it is the saucer Wiring, they're wired up in Parallel but there's no current limiting resistors for each LED.
If the 9V is confirmed to be the power source, you can use 330 Ohms resistor per LED.
Or, the other alternative (which I do not like) is to connect two LEDs in series and use one 150 Ohm resistor. Having 6 LEDs, you will then have three sets of LEDs which you can then connect in parallel.

Another alternative is to bring the power down to 4.5v (3x AAA batteries) or 5v (mobile phone charger via USB, not PC USB) and your resistor can run much cooler at 100 Ohms since there is only about 1.3v to resist.

[Thewrathofodysseus]

Hmmmm, Phone charger you say?