PDA

View Full Version : OT Electrical Problem/Question on LEDs



BigBoy1
12-05-2012, 08:36 AM
My question is an electrical one. I bought 5 LED bulbs, each rated at 24 volt. I have connected them in series, so the voltage drop across all 5 is 120 volts. They won't light. What did I do wrong? I check the resistance across each bulb and it is infinite. Are the bulbs bad? Did I screw up? Appreciate any wisdom.

ranger302
12-05-2012, 08:50 AM
We need more infomation, perhaps take a picture of the setup and we can go from there.

browne92
12-05-2012, 09:11 AM
A link to the bulb specs would also be helpful. If these LED bulbs are rated for AC, then I suspect there are power supplies in each bulb that will not allow it to work this way.

vincemulhollon
12-05-2012, 09:14 AM
A link to the bulb specs would also be helpful. If these LED bulbs are rated for AC, then I suspect there are power supplies in each bulb that will not allow it to work this way.

Probably the most exciting part is a resistive thing like an old fashioned light bulb doesn't care (much) about DC vs RMS but a LED probably will.

120V AC RMS is more like 170V AC peak.

MaxHeadRoom
12-05-2012, 10:18 AM
You did not state if using 120vAC?
Simple 24v LED's are usually rated for 24vdc, they have a suitable resistor internally, but they are not designed for AC if this is what you powered them with?
Max.

jnissen
12-05-2012, 10:25 AM
Yes more info is needed. AC or DC rated voltage for the lights? Got a link to these things?

Jaakko Fagerlund
12-05-2012, 10:35 AM
I suspect that you have just blown out 5 LED's, because if they were for 24 VDC, they just got that in reverse when you connected them to 120 VAC (as I'm suspecting). Hope they didn't cost you much.

danlb
12-05-2012, 10:58 AM
My question is an electrical one. I bought 5 LED bulbs, each rated at 24 volt. I have connected them in series, so the voltage drop across all 5 is 120 volts. They won't light. What did I do wrong? I check the resistance across each bulb and it is infinite. Are the bulbs bad? Did I screw up? Appreciate any wisdom.

Please post the model number of what you bought and we can reverse engineer it for you. It might be something as simple as a grounded case on the body of the "bulb".

LEDs are interesting in that they do not light at all until the voltage reaches a threshold. As the voltage rises above that they quickly reach full brightness, then very quickly hit overload. Did you measure the voltage drop across each one?

Lastly, the term BULB implies a ready to use fixture( often with electronics) , as opposed to an LED, which is a discrete component.

Dan

mayfieldtm
12-05-2012, 02:22 PM
Probably one is dead (open) so that stops current to the whole string.
I'm not sure, but, possibly for a fraction of a second when you first apply power, one of the LED's gets too much voltage and opens.
LED are not purely resistive and I can see this happening.
Try using a single LED with a resistor for 120V. (doesn't matter if AC or DC).
Then parallel several of these circuits for the desired brightness.
This is not a very efficient circuit.
If it were me, I'd use a small low-voltage transformer and design from it's voltage.
Tom M.

Evan
12-05-2012, 02:52 PM
You cannot run LED bulbs like that. 24 volts is an odd rating for LEDs bulbs but they do make them for use in China. Many cars and most trucks in China have 24 volt electrical systems but they are DC.

If you are trying to run them on AC you may have cooked them all. Most AC LED bulbs will not read anything other than open circuit. They use internal switch mode power supplies or on the cheap ones capacitor current limiters. DC LED bulbs may or may not read open circuit depending on how they are powered internally.

This is all changing quickly. The next step in AC LED lighting will come out next year. These are LEDs with multiple emitters on one die along with rectification. They run directly from 110 vac or 240 vac.

BTW, the current LED champion is a 300 watt unit about 1.5 inches square. It puts out somewhere around 30,000 to 35,000 lumens.

BigBoy1
12-05-2012, 03:00 PM
Here are the spec. for the bulbs I purchased.


24 Volt.T3 1/4 Miniature Screw (E10) LED Light Bulb 0.72 Watt Color White

Product ID# LM1024MS-W
LED BULB
T3 1/4 Mini Screw LED
Color: White
Plastic Envelope Shape: T-3 1/4
Max. Bulb Diameter (in): 13/32
Base Type: Miniature Screw (E10)
Design Volts: 24.00
Mili Amp (mA): 30.00
Design Watts: 0.72
Bi-Polar: Yes
Viewing Angle Degree: 360
Luminous Flux: 3.10
Wave Length: 0
Average Life Hours: 100,000
1–5 units 6 & Above
$8.00 $7.60

The wiring was to place the five LEDs in electrical series so that each bulb is a candle in a candelabra. The bulbs in the picture are the old bulbs, NOT the LEDs. The white power cord plugs into 120V AC.

http://i163.photobucket.com/albums/t308/i422twains/PICT0197-1.jpg

Evan
12-05-2012, 03:36 PM
The bulbs are "bipolar" which means they are designed to run on ac. HOWEVER, it also means they have a rectifier of some sort as well as a current limiter. The bulbs will not present a pure resistive load and cannot be run in series with predictable results.

The product is here: http://pearlandlighting.com/proddetail.php?prod=LM1024MS-W

Those bulbs are rated at what looks like 3 lumens. That isn't even as bright as a night light. The bulb form factor is so small that there is only room for a diode and resistor in series with the LED.

Barrington
12-05-2012, 04:09 PM
Doesn't a 'Bi-Polar LED' contain two LEDs back to back (in parallel) rather than additional rectification ?

If these ones are 24v 30mA they they may just have an internal ~700ohm resistor in series.

Putting 5 of these in series would then be effectively the same as 3500 ohms in series with a chain of 5 LEDs - i.e. should work o.k. at 120v - for a while at least. They may however be susceptible to mains transients (voltage spikes).

They should light up enough with a 12v battery to check if they still work (testing each one both ways round )...

Cheers

.

Paul Alciatore
12-05-2012, 04:26 PM
LED stands for Light Emitting DIODE. Emphasis on the word "DIODE". A diode will only conduct in one direction. So if you connect it to AC, it will act exactly like a rectifier and only conduct (and light) when the Voltage/current is flowing in one direction. On the other half cycles, when it tries to flow in the other direction, it will not conduct and it will not light. LEDs can be connected to AC and they will work: like all other diodes, they simply block conduction in the reverse direction. This does not burn them out unless the reverse Voltage is greater than the rated reverse Voltage of the LED.

My first thought is if you are running them in series, they MUST all be connected in the same polarity. That means all of the plus (+) terminals on the LEDs must be connected to minus (-) terminals on the next one (cathode to anode). Just like batteries. Even if only one of them is connected in the reverse polarity, they will not conduct and none of them will light. Since you are using screw based sockets, this SHOULD mean that the center terminal on one socket is connected to the outer terminal on the next. Recheck your wiring and try it again. Chances are you have not destroyed any of the LED lamps.

All the above being said, you may still have some problems. There is no guarantee that the forward current or conduction (the inverse of resistance) rating of each individual LED lamp is the same as the next. If they are sufficiently imbalanced the actual Voltage across the individual ones in a series circuit will not be the same: some will be higher than 24 Volts and others will be lower. Since it is a series circuit, the current will be the same for all of them, but it may be low and that would mean all of them may be dim.

BTW, you can not check a 24 Volt LED with an ordinary Ohm meter because most Ohm meters use a 1.5 or 3 Volt battery to check the resistance. Those Voltages will not be enough to turn on a 24 Volt LED so no conduction will occur in either polarity. It will read very high or infinite resistance. You would need a meter with a higher test Voltage, at least 24 Volts. I do not know of any such meter. 9 Volts if the highest I have ever seen except for meggers.

danlb
12-05-2012, 04:27 PM
A door bell style transformer will provide you with 24v AC so you can run them in parallel. That's your best bet.

Dan

Barrington
12-05-2012, 04:40 PM
A diode will only conduct in one direction. ...But these are bipolar LEDs. Two LEDs in parallel such that one or other conducts depending on the polarity of the applied voltage.

Cheers

.

Evan
12-05-2012, 04:48 PM
Bipolar could mean two LEDs back to back but it probably doesn't. It simply means that the direction of the applied voltage doesn't matter which implies ac. The bulb is intended as a replacement for an indicator lamp, most likely one that runs on 24 vac.


They should light up enough with a 12v battery to check if they still work (testing each one both ways round)...

Maybe but probably not. White LEDs may not light on only half the voltage. It depends on the amount of current limiting.

Evan
12-05-2012, 04:51 PM
BTW, you can not check a 24 Volt LED with an ordinary Ohm meter because most Ohm meters use a 1.5 or 3 Volt battery to check the resistance. Those Voltages will not be enough to turn on a 24 Volt LED so no conduction will occur in either polarity. It will read very high or infinite resistance. You would need a meter with a higher test Voltage, at least 24 Volts. I do not know of any such meter. 9 Volts if the highest I have ever seen except for meggers.

It isn't a 24 volt LED. It is a single white LED and that means it operates on a voltage drop of around three volts. Throw in a diode and that makes it close to 4 volts.

LEDs have a very low reverse voltage rating, 5 volts at the most. If you apply 24 volts without a diode that will be exceeded even with a resistor in series since it won't conduct until the back voltage is exceeded. Then it is toast. Blue LEDs are also static sensitive so some care should be taken to avoid static discharge when handling. That includes white LEDs since they are really blue inside.

Barrington
12-05-2012, 05:01 PM
As you say, this is just an indicator LED, so the forward voltage is unlikely to be more than 4v, i.e the series resistor is going to be about 20v/0.03A = 667ohm.

At 12v that means a current of about 8/667 = 12mA.. should be plenty to get a visible output.

Cheers

.

Evan
12-05-2012, 05:08 PM
It probably will be but if it doesn't light it should be tested on full voltage before pitching it.

Jaakko Fagerlund
12-05-2012, 05:08 PM
This does not burn them out unless the reverse Voltage is greater than the rated reverse Voltage of the LED.
Usually the reverse voltage rating of an LED is quite low and in OP's case they would have burned the second they were reversed if they would not be bipolars.

darryl
12-05-2012, 07:01 PM
I think Paul put his finger on this one. Chances are there is one diode per 'light', and if even one is reversely connected, there will be no current flow in either direction. Bipolar in this case means that it doesn't matter which polarity of voltage gets to the module- it will only use the polarity from one direction. If you have all in series, each will have to be connected so it uses the same polarity. Otherwise one will block the voltage from the others, and any of the others will block the voltage from that one. Chances are pretty much 100% that you haven't destroyed anything.

The way to figure this out and still be able to have them work directly from the ac line is to test each one from a 12v source such as a battery. You might even get them to light from a 9v- it would be dim, but the voltage is enough to exceed the leds requirement and so is valid as a test. When you find the polarity which lights them, mark it on the base. Then wire them all up + to-, + to -, etc. The string will work on 110ac then.

Barrington
12-06-2012, 05:22 AM
Looking around for data around for these types, from what little I've found I think the term 'bipolar' has now become distorted from its original useage.

It looks as though these are likely to be what many manufacturers term 'non-polarised' and probably do include a bridge rectifier as Evan suggested.

Shouldn't make much difference on a 12v test though - they should still draw nearly 11mA.

Cheers

.

Barrington
12-06-2012, 05:58 AM
LED reverse breakdown isn't always fatal - it depends on the reverse current.

Following a thread a couple of months ago I wired an ordinary red LED in series with an 82k resistor across 240vac. Although it breaks down on each cycle (at about 16v), the reverse current is limited to less than 3mA - and it's still working...

In this case the forward current is only 3mA as well, so it's not very bright...

If forward current was more a typical 20mA for example, the equivalent reverse current would cause burnout because of the much greater dissipation at the relatively high breakdown voltage. (i.e. 16vx20mA = 320mW)

Cheers

.

Black Forest
12-06-2012, 08:02 AM
My question is an electrical one. I bought 5 LED bulbs, each rated at 24 volt. I have connected them in series, so the voltage drop across all 5 is 120 volts. They won't light. What did I do wrong? I check the resistance across each bulb and it is infinite. Are the bulbs bad? Did I screw up? Appreciate any wisdom.


THIS THREAD IS WORTHLESS WITHOUT A PICTURE! I saved you the trouble Jerry !

J Tiers
12-06-2012, 08:12 AM
THIS THREAD IS WORTHLESS WITHOUT A PICTURE! I saved you the trouble Jerry !



We need more infomation, perhaps take a picture of the setup and we can go from there.



Yes more info is needed. AC or DC rated voltage for the lights? Got a link to these things?

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Evan
12-06-2012, 11:07 AM
I provided a picture. It's at the link.

Paul Alciatore
12-06-2012, 02:37 PM
On the subject of the reverse Voltage rating, I need to add some thoughts. First, these LED bulbs are rated at 24 V. I have never seen any LED that has a forward Voltage greater then 5 Volts and most are 2 - 4 Volts so I have to assume that there are several of them in series, perhaps six or more. The other alternative to this would be to have a rather large value series resistor and that would be unlikely in this energy efficient world. Such a series string would also increase the reverse Voltage rating of the series string. So, in lack of more information, I stand by my assumption that no damage was done.

As for the bipolar spec, I admit I overlooked it, but I look at it as an almost meaningless term, just like the term "all natural" that is used so indiscriminately. I have never seen a product with supernatural ingredients, have you? Bipolar can mean almost anything so it is worthless. I find the fact that the specs do not make any reference to AC or DC to be very telling. Basically, the person who wrote the specs didn't know what he or she (lets be PC here and include both genders) was doing. Probably written by an advertising type person.

I like the 24 Volt doorbell transformer suggestion. Get one and test the individual lamps. Then try using them in series if they are still OK. If they are blown, then get more and use a parallel circuit with the transformer.

Paul Alciatore
12-06-2012, 02:49 PM
I have to disagree with Evan when he says that bipolar means that they have a bridge rectifier. There are other ways of creating a LED lamp that will operate on AC. Imagine two series strings of individual LEDs with a total Voltage rating of perhaps 20 Volts or more. They are connected in parallel, one for conduction in each direction. Then a single resistor to limit the current. If the reverse current is a problem, which I doubt, then a single diode with a higher Voltage rating can be added to each series string to block the reverse Voltage. All of the above could probably be incorporated into one or two components for mass production.

A single series string can also be used with a resistor if the reverse Voltage is not a concern. It will only conduct on one phase but it will work on AC so it can be called "bipolar".

I am sure there are other ways to do this and each may have different consequences when a series string is attempted. As I said earlier, I do not like series circuits for lights of any kind. Go parallel. If there is no room for a transformer in the candle fixture, then use a wall wart to power it.

Evan
12-06-2012, 03:49 PM
I have to disagree with Evan when he says that bipolar means that they have a bridge rectifier.

I didn't say that. I wrote:

The bulbs are "bipolar" which means they are designed to run on ac. HOWEVER, it also means they have a rectifier of some sort as well as a current limiter.


Imagine two series strings of individual LEDs with a total Voltage rating of perhaps 20 Volts or more. They are connected in parallel, one for conduction in each direction.

We already discussed that, as in a pair of opposite polarity diodes in parallel. That is very unlikely as no advantage accrues for an indicator and a plain diode is much cheaper. It could be a bridge though since then the LED is truly "bipolar".

Barrington
12-06-2012, 04:35 PM
It seems there are both 'bridge' and 'back to back' types around.

From Dialight: T1 3/4 incandescent indicator bulb replacements: http://www.dialight.com/Assets/Brochures_And_Catalogs/Indication/MBELCRFX002.pdf

http://i564.photobucket.com/albums/ss82/MrBarrington/LED002.png

- and from Solitronics: http://www.solitronics.com/products/multichip-led/CAT-BASED-LED-PACK01.pdf

http://i564.photobucket.com/albums/ss82/MrBarrington/LED001.png

Difficult to say which type the the OP has, but the bridge type seems far more sensible for a device listed for both AC and DC.

Cheers

.

Evan
12-06-2012, 04:44 PM
LEDs are now cheap enough to make the dual LED approach economical.

Barrington
12-06-2012, 04:52 PM
...as in a pair of opposite polarity diodes in parallel. That is very unlikely as no advantage accrues for an indicator and a plain diode is much cheaper.
LEDs are now cheap enough to make the dual LED approach economical.
Wow- prices are dropping fast !

Cheers ;)

.

darryl
12-06-2012, 05:20 PM
If there was a bridge rectifier involved, or two leds back to back, then it wouldn't matter in which direction they are wired- they should all work. I suspect that if they were all in working order in the first place, there is probably a single rectifier in series with the led, plus a resistor. This would be cheaper to make and thus is probably how it's done. In this case then, polarity of connection would matter. The potential remains though- if one unit of any series connected string of devices is/was somehow burnt open, the rest would not work.

From the Solitronics data page that Barrington posted, if that is the configuration within the 'device', then each led conducts for one polarity of the ac voltage- the result being that current will flow in both directions through the device. It would not matter in this case how they are connected. Each one of these would require a minimum of about 3 volts to begin to light, assuming they use typical while led technology. A series string of 5 would then need a minimum of 15 volts to begin to show any light. If rated at 24 volts, 5 in series would work from 110. If this string fails to show light, then at least one of the devices in the string is burnt open.

The chore is going to be to check each device in turn to see whether it lights from just one polarity of connection, or from both. If it doesn't light at all, it's defective. If all light, but just with one polarity, then they must all be connected in that same polarity throughout the string. If they all light individually with either polarity of applied voltage, then they can be connected without regard to polarity. You must test them one by one to know firstly if they are working, and secondly to know whether they are actually polarity sensitive, regardless if they are called 'bipolar'.

BigBoy1
12-06-2012, 06:14 PM
I appreciate what has been written so far, and I can see that I have strayed too far into an area which I known nothing about. My naive assumption was the LEDs were "sorta like" light bulbs and as such they could be just screwed in. I guess it was a good learning experience which cost $53 for the bulbs and shipping. Any solution proposed so is far is too complicated for me. If anyone would like the bulbs to "play" with, send me postage to ship them to you and they are yours. Knowing my luck, I burned them out!!!

Barrington
12-06-2012, 07:22 PM
Don't give up yet !

It may just be the wiring, have you double, double checked it ?

Otherwise just connecting each bulb to a 12v battery should tell you something - how hard can that be ???

Cheers

.

Evan
12-06-2012, 08:04 PM
Wow- prices are dropping fast !

You would not believe just how fast. I price LEDs by the lumen. I bought a few tapes of tiny smd LEDs that put out about 15 to 20 lumens each for $2.00 for 50. That is about 0.2 cents per lumen. That works out to the same amount of light as a 4 foot fluorescent bulb for around $8 and the LEDs will last 3 to 10 times as long while drawing 1/3 to 1/5 the power. Prices will continue to fall, to less than half what they are now. Prices for complete bulbs will fall to around $1 per 500 lumens within 5 years.

darryl
12-06-2012, 08:21 PM
Looking through the local stores, it's hard to find where you can buy a significant number of lumens in a screw-in bulb for under $20. Guess I'll have to start shopping at DealXtreme.

Evan
12-06-2012, 08:47 PM
Some examples:

152093 E27 7W 4000K 700lm 14-LED Warm White Light Bulb - Purple (AC 85~265V) USD$ 10.80 Quan 1

152110 E27 7W 6500K 700lm 14-LED White Light Bulb - Golden (AC 85~265V) USD$ 9.58 Quan 3

132306 E27 4W 360~400LM 5500~6000K White 4-LED Spot Light Bulb (AC 85~265V) USD$ 5.50 Quan 3

Those are the current best deals based on price vs lumens, LED type and heat sink type and quality. Also, the two 700 lumen bulbs have the same form factor as incandescent bulbs so they will fit any standard socket and lamp. However, they do not produce a lot of side light, they have about a 160 degree effective spread. They are perfect for a directed reading lamp or a ceiling fixture. The warm white bulb is a little higher on the colour temperature so the light isn't that sickly yellow that some are.