Today I soldered an LED into a circuit - first time I soldered an LED.
On testing, the LED was dead.
Thought maybe the LED was dead to start with, so I took a new one, tested it and soldered it in.
It didn't fire up immediately, then worked ok for all of 5 minutes - and died.
So obviously the LEDs are not surviving being heated.
Common sense says for the next attempt be sure to use a heat sink.
FWIW I am only using a 40W iron, although not temperature controlled.
Is this a common issue with soldering LEDs? Anybody else have a similar problem?
Or is it just that the cheap eBay LEDs are showing their true quality.
Semiconductors have always had a sensitivity to being overheated. If you are holding your iron to the solder joint longer than 3-4 seconds you are likely to overheat it. It is common practice to use an alligator clip on the lead to be soldered as a temporary heat sink while soldering.
That being said... you should have an iron and solder (and possibly extra flux) that allows you to make these connections in a short time before heat damages the semiconductor.
If it worked for 5 minutes, the soldering is not the problem. How much current is going through the LED? What is the size of your current limiting resistor and what voltage are you driving the circuit with?
Are you sure it's the soldering? Do you have a current-limiting resistor in series with the LED?
I have seen "flakey" LEDs, but it's VERY RARE. Where I work, we sometimes get a bad batch of LEDs where perhaps 10% of the LEDs are bad after soldering. ...Problem is, we don't test them before soldering.
Of course my company doesn't buy random LEDs from eBay, and as a hobbyist I also buy from reputable distributors.
We just had a batch of 50 boards where there are 4 green LEDs, 4 yellow LEDs, and one red LED. Most of the time, there are no problems, but when it happens, it's always the green LEDs that give us trouble. About half of the boards had one or more bad LEDs. We had 1600 LEDs remaining in stock. I tested a bag of 200 and they all passed. So, the test was inconclusive... Maybe 10% of the LEDs in the previous bag were bad, or maybe the soldering temperature was a bit higher than usual....??? But, I'm sure it wasn't just the soldering temperature. If they are failing after solder, these LEDs are defective and too sensitive to solder.
LEDs are ESD sensitive devices. The other day I was a bit careless wiring up some RGB LEDs.
A few of the BLUE LEDs in the RGB LED had reduced brightness after that.
Yes, overheating and ESD kills LEDs.
If you work with a cheap/old-fashioned poker, clip on something that shortens the pins and keeps the LED cool.
Leo..
pwillard:
It is common practice to use an alligator clip on the lead to be soldered as a temporary heat sink while soldering.
Alligator clips are not likely to do much useful.
I wasn't sure whether they were still around, but thirty or so years ago, Radio Shack sold a set of heatsink clips for soldering components. Some were free-standing, a couple were attached to a magnet to hold onto some nearby piece of "furniture" (attached to my set of component drawers for much of that time!). These were made of 1/20 of an inch aluminium, and the jaws are flat in order to provide full contact to the component lead, so they were indeed designed properly for the purpose rather than a common alligator clip.
Research: They are still available in one form or another!
I have most certainly had failures of LEDs from soldering, and especially de-soldering and replacement. Whilst thermal degradation of the chip is certainly one mode of failure, more commonly the slightest torque on the wires as the plastic softens causes the weld on the bond wire to fail resulting in an intermittent connection which can be demonstrated by stressing the lead wire in various directions.
Common problem demonstrable in those cheap LED torches with 7 or so LEDs.
Hi,
Hope you are not trying to solder SMD LEDs, but LEDs with 20mm leads.
How are you testing the LED?
Have you tested the LED before soldering it in.
Are you cutting the leads before soldering, if so try soldering one with the full length leads.
What is the circuit you are placing the LED in? Voltage?
Are you using a current limit resistor in series with the LED,
Can you post a picture of the LEDs you are trying to solder?
Can you please post a copy of your circuit, in CAD or a picture of a hand drawn circuit in jpg, png or pdf?
But those were the days of germanium such as GET103 etc, when the slightest heat could damage and cooling clips were the order of the day... Along came silicon and life was bliss. Today's lads don't know they're born.
The circuit is not the problem.
It is a sensor with a transistor switch and no led.
I was using an external 3mm led on a breadboard to monitor the switch.
I simply added the external led circuit onto a prototype to avoid having to hook up the breadboard each time.
I have since soldered in another led and it works just fine.
This time used a small spring clamp as a heatsink.
I normally use the clamp to hold the pcb when soldering.
It is a common practice to recommend such a thing, but I have never done this since 1965 and have never damaged a component by soldering.
I agree. I never damaged a led during soldering but my electronics experience only goes back to 1979. I was a High School sophomore in 1965 and didn't even know what a resistor was (I may have been aware of LEDs but that was about it).
I need a constant temp soldering iron.
Why ? I've used the cheapest soldering iron you could buy at Radioshack up to expensive Metcals with the
tweezer accessory for removing surface mount devices and never had a problem. Are you sure you know how long to apply the iron ? Is it possible you're holding the iron on too long ?
I now think I now understand why people stop trying to be helpful in this forum. I was assuming that a crap iron was being used and that solder time was too long and then recommended that wicking some heat away would be helpful. I seemed to have only created a "dog pile"about 1965 practices.
I was assuming that a crap iron was being used and that solder time was too long and then recommended that wicking some heat away would be helpful. I seemed to have only created a "dog pile"about 1965 practices.
We are no closer to knowing the cause of the failure but the "dog pile" as you refer to is the most telling
evidence that the problem was user error, so while you may think it is relevant , it is in fact the most relevant information we have pertaining to this post. I it has worked for everyone else since 1965 then I think it's safe to say that the led was never defective and was in fact subjected to abuse of some nature.
Moreover, there is more to the forum than a dry answer. We are , after all , human, and not microprocessors or paid customer service reps. Our experience is the result of many many many years of working in the industry. If someone wants to post on the forum for free advice, I think they can sit through the commercials or they are certainly free to go somewhere else where the answers are crap but there's no commercials. We did our time and have earned the right to reminisce or tell a story now an then, if you will. If I may say so, you have a very negative vibe. I don't see anyone else complaining about anything. in this thread
Just out of curiosity, how many years experience do you have ?
jackrae:
5 seconds is far too long. Properly set up with clean tinned iron and components you should achieve a joint in about 1/2 second, 1 second at the most.
That was not an indication of time - it was me biting my tongue. Maybe I should have gone up all the way to ten.
@pwillard - yes it is a crappy iron, but for the most part it works ok. Only sometimes the temperature will drop radically and I get no solder flow. So I generally wait a bit to let the component cool down before trying again. I know this happened with the LED, so the cumulative attempts probably induced too much heat to LED and fried it. The LEDs were very cheap on eBay - that probably did not help. The heat sink helped.
