Go Down

Topic: LEDs losig brithness at a different rate, am I damaging them somehow? (Read 8970 times) previous topic - next topic

Grumpy_Mike

I thought you wanted to add one resistor to each of the existing resistors, that is why I said it would be fine. Yes with 150mA you burn just under 2W so you need a bigger resistor.
However as I said before it looks like the LEDs, did you measure the forward voltage drop of them?

byteofthat

I thought you wanted to add one resistor to each of the existing resistors, that is why I said it would be fine. Yes with 150mA you burn just under 2W so you need a bigger resistor.
However as I said before it looks like the LEDs, did you measure the forward voltage drop of them?
Well, I went and got a fuse today.

The I get ~9mA running through an LED in the phrase "KLUKKAN ER".

I get ~7mA running through an LED in the phrase "MINUTUR". I changed an LED and measure again, and still ~7mA.

I placed my probes on each side of an LED in both "KLUKKAN ER" and "MINUTUR". I get the same reading, ~2.9V. 2.95 in the former, 2.91 on the latter. This is the voltage drop/forward voltage right? I am not sure if the terms are synonymous, I am still trying to figure that out. Any chance you could clear up my confusion with this terminology? I get the impression forward voltage/voltage drop is the volts consumed by the component in question.

I wonder if the reason the amperage reading is different is because there is a small amount of resistance in the transistor arrays? I can't seem to account for the difference any other way.

I am starting to wonder if this is just a cause of lousy LEDs? These are the LEDs I bought:
http://www.ebay.co.uk/itm/50-x-White-LED-5mm-Diffused-1st-Class-Postage-UK-/121419713404?pt=UK_BOI_Electrical_Components_Supplies_ET&hash=item1c452dcb7c

Just a reminder that "KLUKKAN ER" is not connected through the transistor arrays, and is simply just connected to ground.

Grumpy_Mike

Voltage is not consumed, that is current. The voltage across the LED is the forward voltage drop.

There is less spair voltage on the transistors because there is a permanent saturation voltage across the emitter / collector of about 0.7V, so you loose some there and so the current is less for the same resistor.

byteofthat

Well I guess we are back to square one then. I suppose it either has to be poor LEDs, or the junction temperature is too high. If the latter, not really sure how to test that. I suppose I would need to setup something on a breadboard and enclose an LED in a similar situation. Even still, not sure quite how to measure the temperature effectively.

ghlawrence2000

If it were me, I would take everything else out of the equation and just go for Regulated 5V supply and 68R resistor per LED for say 2 or 3 LEDS and 75R resistor for another 2 or 3 LEDS and just run them constantly on a breadboard until/if you see any difference.

The Spec on the ebay advert says 3.2-3.8 forward volts 20-30mA forward current...

So :-

5 - 3.2 / 68  = 26.5mA  //  A little higher than typical but below max rating
5 - 3.2 / 75  = 24mA    //  A little higher than typical but below max rating
5 - 3.8 / 68  = 17.6mA  // Approaching typical rating
5 - 3.8 / 75  = 16ma     // Approaching typical rating

5 - 3.2 / 220 = 8.2mA   // Expected to be quite dim
5 - 3.8 / 220 = 5.5mA   // Expected to be even dimmer

This will at least drive them closer to their rated spec and prove if it is the LED's or something else....


Regards,

Graham
UTFT_SdRaw now included in library manager!! ;) High speed image drawing from SD card to UTFT displays for Mega & DUE.
UTFT_GHL - a VASTLY upgraded version of UTFT_CTE. Coming soon to a TFT near you! 8) Shipping April 1 2016!

byteofthat

If it were me, I would take everything else out of the equation and just go for Regulated 5V supply and 68R resistor per LED for say 2 or 3 LEDS and 75R resistor for another 2 or 3 LEDS and just run them constantly on a breadboard until/if you see any difference.

The Spec on the ebay advert says 3.2-3.8 forward volts 20-30mA forward current...

So :-

5 - 3.2 / 68  = 26.5mA  //  A little higher than typical but below max rating
5 - 3.2 / 75  = 24mA    //  A little higher than typical but below max rating
5 - 3.8 / 68  = 17.6mA  // Approaching typical rating
5 - 3.8 / 75  = 16ma     // Approaching typical rating

5 - 3.2 / 220 = 8.2mA   // Expected to be quite dim
5 - 3.8 / 220 = 5.5mA   // Expected to be even dimmer

This will at least drive them closer to their rated spec and prove if it is the LED's or something else....


Regards,

Graham
I actually happen to have my breadboard and resistor kit out, I think I will give that a try.

Is it possible to damage an LED by driving them with too low of a current?

Zapro


Grumpy_Mike

Is it possible to damage an LED by driving them with too low of a current?
Not only will you not damage them they will last a lot longer.

byteofthat

I came across this:

http://lumeniquessl.com/2008/11/30/5mm-leds-generally-unsuitable-for-general-illumination/


Maybe this is a heat issue? The writer seems to hint that the 5mm bulbs I was using aren't really for continuous use. I'm guessing we are talking at about the 600 hour mark when the light brightness started to dim.

The writer also suggest using "LED packages". If I was to do another project using LEDs in such a continuous use fashion, what packages might be suitable? Are all LED packages SMD that I might possibly use in a situation like this?

Able to recommend a better dealer of LEDs?

Grumpy_Mike

Quote
The writer seems to hint that the 5mm bulbs I was using aren't really for continuous use.
He has a axe to grind.

Quote
Are all LED packages SMD that I might possibly use in a situation like this?
There is no difference between an SMD LED and the ones he is talking about in this page.

He wants to sell you:-
Quote
Packaged LED devices, with proper thermal control,

arduidiot

data log the analogRead(x) values by connecting one of the analog inputs of the data logging device to the output of the power supply board, but with a resistor that will keep the current below the max operating for the data logger, eg if you have programmed an uno with an SD card shield to log the analog input on the pin connected, see the max operating current for the uno in its spec sheet.

get a good number of samples, say leave it writing to an SD card all day. copy and paste your data into an spreadsheet and use the stats formulas to spot any surges/transient voltage spikes that have occurred. if its being rectified well enough and there is no decent spikes, well excluse the power supply and focus on the led circuit.

in this case, and If you are  not getting a few naughty little buggers at  the fringes of the bell curve of kinetic energy distribution that are damaging your circuit, im not sure exactly what your problem is, ( in my case its almost always because ive rushed the pin out and its not wired like i wrote in my book or the person who wrote the project guide has recommended, ive soldered two IC pins together that short the MCU, SOMETHING silly)

but ive attached  photo of on of the  component types i like for surge protection (range of this design, not the exact one i have attached a pic of) and they seem to be very popular
im a good observer. not the brightest but very curious.

that poor innocent dog died from AC. he died.

I do not refuse my dinner simply because I do not understand the process of digestion."

Coding Badly

Quote
Able to recommend a better dealer of LEDs?
Cree for white.  Lite-On for colored.


byteofthat

data log the analogRead(x) values by connecting one of the analog inputs of the data logging device to the output of the power supply board, but with a resistor that will keep the current below the max operating for the data logger, eg if you have programmed an uno with an SD card shield to log the analog input on the pin connected, see the max operating current for the uno in its spec sheet.

get a good number of samples, say leave it writing to an SD card all day. copy and paste your data into an spreadsheet and use the stats formulas to spot any surges/transient voltage spikes that have occurred. if its being rectified well enough and there is no decent spikes, well excluse the power supply and focus on the led circuit.

in this case, and If you are  not getting a few naughty little buggers at  the fringes of the bell curve of kinetic energy distribution that are damaging your circuit, im not sure exactly what your problem is, ( in my case its almost always because ive rushed the pin out and its not wired like i wrote in my book or the person who wrote the project guide has recommended, ive soldered two IC pins together that short the MCU, SOMETHING silly)

but ive attached  photo of on of the  component types i like for surge protection (range of this design, not the exact one i have attached a pic of) and they seem to be very popular
Wouldn't I need to use some sort of potentiometer for this? If I was getting spikes beyond 5 volts, wouldn't that damage the analogue input pin? Or at the very least, would I ever get a reading above 5 volts? I think that's where its measuring range maxes out.

I also have access to an oscilloscope at a local fab lab, I could try and bring the project down there in the coming week and observe it for a while. 

arduidiot

note that i said you must connect it to the analog input with an appropriate resistor inline to drop the analog input voltage to a safe level. if you are getting spikes, it will still spike but below 5 V with that resistor. so you simply calibrate the analog read data  to reflect how large the voltage spike WOULD be if you had not have put the resistor there.
im a good observer. not the brightest but very curious.

that poor innocent dog died from AC. he died.

I do not refuse my dinner simply because I do not understand the process of digestion."

Grumpy_Mike

at  the fringes of the bell curve of kinetic energy distribution that are damaging your circuit,
What on earth does that mean? Total rubbish to talk of kinetic energy distribution in the context of electronics.

Data logging this will tell you nothing.

Go Up