Why are my LEDs burning out?

[Brouhaha]

I'm not an electrician, just a hobbyist trying to build a small sculpture which incorporates LEDs, but they keep burning out. I am using a 8w/24v Meanwell power supply, a single 20mA Dynaohm resistor, and 3 strings of four 5mm warm white LEDs each, in parallel (see attached diagram). They all lit briefly, but very quickly one string after the next went dark. What am I doing wrong, and how to fix it? THANKS!

 

[telectrix]

probably excessive voltage. change the power supply to a 12V one. (assuming that the LEDs are "normal" 12V.it will be stated on the box and maybe on the lamp also).)

 

[James]

What is the spec of the resistor?

 

[DPG]

I'd use one fixed value resistor per chain of LEDs.

 

[James]

Point is, we do not know if the resistors resistance is correct for the load to bring the voltage down within limits of the led’s

 

[telectrix]

can'tunderstand if the LEDs are 12V , why OP is using a 24V PSU.

 

[DPG]

OP, give us some details of the LEDs you are using.

 

[plugsandsparks]

What are the specs of the LEDs?

 

[telectrix]

OP, give us some details of the LEDs you are using.

2" diameter round which give a circumference of 6.25" and a surface area of 3.174 sq. in. 2 bits of wire stick out the back to show the way into the connectoer.

 

[Avo Mk8]

The OP's "resistor"is a 20mA constant current device.

https://www.ledsupply.com/content/pdf/led-driver-luxdrive-dynaohm_documentation.pdf

Surely you shouldn't put three strings of led's in parallel like that?
Any slight difference in the forward voltage drops of the diodes will mean the current is not equally shared between each of the strings. I'd think all the current will go through the diodes with the lowest drop, and the others won't light?

Broughaha - I think you need a Dynaohm in each of the three strings of LED's for your scheme to work.
Can you give details of the led's - recommended running current, forward voltage when running?

 

[ColinACJL]

I'm not an electrician, just a hobbyist trying to build a small sculpture which incorporates LEDs, but they keep burning out. I am using a 8w/24v Meanwell power supply, a single 20mA Dynaohm resistor, and 3 strings of four 5mm warm white LEDs each, in parallel (see attached diagram). They all lit briefly, but very quickly one string after the next went dark. What am I doing wrong, and how to fix it? THANKS!

 

[Brouhaha]

Thank you all for the very prompt replies. I see that I did not provide enough detail in my original post. I should also mention that I had previously tried using a 12W/15V Meanwell and with that 11 of the 12 bulbs illuminated but the 12th did not. Folks at LEDsupply.com thought I needed more voltage, but then they all burnt out. Below are the product details for the parts I am using:

DynaOhm (mine is the 20mA)

Product Details​

Output Current: 20mA, 25mA or 30mA	Input Voltage Range: (2.6VDC + Vf LED) to 50VDC
Connection: 2 Wire	Output Protection: Open & Short Circuit
Maximum Voltage Drop: 30V	Size: 0.25" (Diameter) X 0.5" (L)
Control: Pulse & Strobe Capable	Power Dissipation: 0.6W

LEDs:

Product Details​

Luminous Intensity: 4,700mcd typ. @ 20mA	Power Dissipation: 120mW
Max Forward Current: 30mA	Operating Temp: -30 to +85 C
Pulse Current: 100mA for <= 10ms, duty <= 1/10	Soldering Temp: 265 C for 10 secs
Forward Voltage: 3.2V typ. 3.5V max @ 20mA	Max Reverse Current: 50uA @ 5V
Max Reverse Voltage: 5V	50° Degree Half Angle

Meanwell Power source (I am using 8w/24v)

Product Details​

INPUT POWER: 90 ~ 264Vac	CASING: Fully Isolated Plastic Case, IP42 NON PFC
OUTPUT VOLTAGE: 5-36VDC	WATTS: 8W up to 35W
CONNECTIONS: UL Rated 18-20 AWG Wire	Low Cost, High Reliability
WARRANTY: 2-Years	SAFETY: Class 2 Power Unit, Short-Circuit, Over-Load & Over-Voltage
Super Small & Compact Size!	APPLICATIONS: LED Lighting & Moving Sign Applications

 

[Brouhaha]

The OP's "resistor"is a 20mA constant current device.

https://www.ledsupply.com/content/pdf/led-driver-luxdrive-dynaohm_documentation.pdf

Surely you shouldn't put three strings of led's in parallel like that?
Any slight difference in the forward voltage drops of the diodes will mean the current is not equally shared between each of the strings. I'd think all the current will go through the diodes with the lowest drop, and the others won't light?

Broughaha - I think you need a Dynaohm in each of the three strings of LED's for your scheme to work.
Can you give details of the led's - recommended running current, forward voltage when running?

Hi Avo I just posted a general reply with the details of what equipment I am using. I'd be very grateful if you take a look at that and then let me know if you think putting a separate dynaohm on each string would be the best approach, or if you think some other arrangement would be more suitable. Thanks!!

 

[Avo Mk8]

Hi Brouhaha
Thanks for the data. Everything we need is there.

Are your LED's embodied in the sculpture already, or are you still at the experimental stage?

If you still have freedom to play with the arrangement, I would suggest splitting the LED's into two strings of six, each with a Dynaohm in series.

Taking the 'typical' figures, with 3.2V across each lamp, thats 19.2V across six, plus the drop across the Dynaohm of 2.6V = 21.8V so with a 24V supply it should work - BUT if we take 'worst case' of 3.5V across each lamp, that's 21V across all, plus 2.6V = 23.6V, rather close to your power supply output. That's why I think it's best to try it!

You can of course keep to your original scheme of having three strings, but with a smaller number of lamps in each string, I think it's a good idea to have a Dynaohm for each. I was just trying to save you having to buy more Dynaohm's than you need!

If you have a multimeter it would be useful to check you do get these voltages and currents in practice.

Finally I'm not clear why the bulbs should "blow up" unless you connect one the wrong way round, or directly across the supply!
I can see why some may not light when you have parallel strings of a small number of lamps - that's what you need to avoid.
Unless your Dynaohm has blown up while experimenting. It's vital you have a current limiting device like that in each circuit, otherwise you will destroy the bulbs connecting them straight to the psu.
It would be really useful to have a multimeter to check you have 20mA flowing through the Dynaohm(s).

Good luck with the project

 

[Brouhaha]

Hi Brouhaha
Thanks for the data. Everything we need is there.

Are your LED's embodied in the sculpture already, or are you still at the experimental stage?

If you still have freedom to play with the arrangement, I would suggest splitting the LED's into two strings of six, each with a Dynaohm in series.

Taking the 'typical' figures, with 3.2V across each lamp, thats 19.2V across six, plus the drop across the Dynaohm of 2.6V = 21.8V so with a 24V supply it should work - BUT if we take 'worst case' of 3.5V across each lamp, that's 21V across all, plus 2.6V = 23.6V, rather close to your power supply output. That's why I think it's best to try it!

You can of course keep to your original scheme of having three strings, but with a smaller number of lamps in each string, I think it's a good idea to have a Dynaohm for each. I was just trying to save you having to buy more Dynaohm's than you need!

If you have a multimeter it would be useful to check you do get these voltages and currents in practice.

Finally I'm not clear why the bulbs should "blow up" unless you connect one the wrong way round, or directly across the supply!
I can see why some may not light when you have parallel strings of a small number of lamps - that's what you need to avoid.
Unless your Dynaohm has blown up while experimenting. It's vital you have a current limiting device like that in each circuit, otherwise you will destroy the bulbs connecting them straight to the psu.
It would be really useful to have a multimeter to check you have 20mA flowing through the Dynaohm(s).

Good luck with the project

I'm attaching a photo of the sculpture, taken when it was running on the 15V power supply, and a 30mA dynaohm, and one of the bulbs was not lighting up. I then tried changing to the current power supply, which is 8W/24V, and all the lights illuminated but then one string went dark (they did not blow up or even look different in any way, just stopped lighting. I then rebuilt it, replacing the four dead bulbs, keeping the 8W/24V power supply but changing to a 20mA Dynaohm (at the suggestion of someone at LEDSupply.com) but once again a string of bulbs went dark, followed in short order by the remaining strings. I wish that the bulbs were ok and that I just need to fix something about the power supply and resistor, because replacing the bulbs means a time-consuming rebuilding process, but I am assuming that in fact the bulbs are permanently dysfunctional now and will need to be replaced. So, in answer to your question, I guess I am back to the experimental stage!

Do you think that there is any chance that the existing bulbs might still be functional and that they would light up again if I fixed some other problem (e.g. replace the Dynaohm or power supply with something different)?

If not, I will try to redesign as you suggest with two strings of six, each with it's own dynaohm (do you think 20maA is the correct size for those?)

 

[Avo Mk8]

Do you think that there is any chance that the existing bulbs might still be functional and that they would light up again if I fixed some other problem (e.g. replace the Dynaohm or power supply with something different)?

If not, I will try to redesign as you suggest with two strings of six, each with it's own dynaohm (do you think 20maA is the correct size for those?)

Given your bad luck with this, and the work you've put in, I feel it would be worth getting a cheap multimeter (if you don't already have one), get familiar with it, and every time you connect things up, put the meter in circuit on a suitable milliamp range to check the current is what it should be.
Also you would be able to check the resistance values of the lamps you have, to establish if they still might work before you connect them to power.

My concern is that once things have started to go wrong, you can never be sure what might have been damaged. Are the power supply and dynaohm doing what they should? And if a bulb is out, is it because the wires are shorted together, or the bulb is faulty etc. etc.

Your lamps have an absolute max forward current of 30mA, so I would be inclined to run them at 20mA, unless you need the extra brightness. And if you do, I suspect they won't last as long!

If you can get at the wiring to each lamp, you could check each in turn using a working Dynaohm and supply. Be aware that putting the power on the wrong way round, or supplying excessive current because the Dynaohm is faulty, will instantly destroy the lamp (you won't be able to tell by looking at it).

Your power supplies have protection devices in them, but activated at much higher current/ voltage than your lamps will survive. A safer way of proceeding would be to find a constant current 20mA supply that could give up to 48V, and connect all your lamps in series to that. But I confess I haven't found one yet!

I hope the above makes sense. I would be cautious connecting things up now, as I don't really understand why so many bulbs have failed unless your Dynaohm has too, or the power supply is doing something strange.

Regards.

 

[DPG]

I would definitely read Avo's post above. And I agree a multimeter would be very useful for you.

 

[Avo Mk8]

Hi Brouhaha

One thing occurred to me - have you connected the Dynaohm the right way round? There's a little "+" on the positive end, and, as supplied, the leads each end are different lengths to also indicate the polarity (don't remember which way round - sorry)

People call it a resistor, but it is a semiconductor device, and only does its job when wired the right way round.
Putting it the wrong way round could destroy the lamps.

And as far as testing/re-using lamps - is the wiring you've done marked with the polarity of the lamps? ie red and black wires, or a tracer mark on one wire, so within the sculpture you know which is anode and which cathode for each lamp?

 

[Brouhaha]

Hi Brouhaha

One thing occurred to me - have you connected the Dynaohm the right way round? There's a little "+" on the positive end, and, as supplied, the leads each end are different lengths to also indicate the polarity (don't remember which way round - sorry)

People call it a resistor, but it is a semiconductor device, and only does its job when wired the right way round.
Putting it the wrong way round could destroy the lamps.

And as far as testing/re-using lamps - is the wiring you've done marked with the polarity of the lamps? ie red and black wires, or a tracer mark on one wire, so within the sculpture you know which is anode and which cathode for each lamp?

Hi Avo, I have the positive wire coming off of my power supply going to the negative terminal of the Dynaohm--is that correct?

I did color code the wires to the bulbs.

I do have a cheap, old analog multimeter, but I've wondered about it's reliability, so maybe I should get a new one...

 

[DPG]

Hi Avo, I have the positive wire coming off of my power supply going to the negative terminal of the Dynaohm--is that correct?

I did color code the wires to the bulbs.

I do have a cheap, old analog multimeter, but I've wondered about it's reliability, so maybe I should get a new one...

I think this is incorrect. The Dynaohm positive should got to battery positive according to the data sheet.