My first question is how is maximum steady state forward current for a LED determined? That is, what are the criteria?

Second, in electronics we use superscripts and subscripts quite a bit. Such as in the "I-sub-f" (or perhaps "I-sub-f-max") that I'd use in posting this question if it was easy to do.

I can waste time playing with fonts to get a halfway decent result for subscripts, but, superscripts...

I tried using HTML but it didn't seem to work / maybe I did something wrong.

???

Thanks!

 

Maximum LED forward current, and all the other parameters, are given by the manufacturer in the datasheet. The processes of design and manufacture of the product determine the values.

 

And, like everything else, it is a compromise.

The manufacturer could rate it higher, but would then have to list a shorter lifetime and/or lower ambient temperature, etc.

 

And typically, for good reliability, you don't wont' to operate a device much above 50-75% of its maximum ratings.

 

My first question is how is maximum steady state forward current for a LED determined? That is, what are the criteria?

Most likely the maximum power dissipation to keep the junction temperature at an acceptable level. Why? Is your intention to dispute manufacturer specifications or second guess them?

Second, in electronics we use superscripts and subscripts quite a bit. Such as in the "I-sub-f" (or perhaps "I-sub-f-max") that I'd use in posting this question if it was easy to do.

You can use Latex.There's a sticky in the Homework Help forum.
\(I_f\)

 

• You can use Latex. There's a sticky in the Homework Help forum.
  \(I_f\)

Noted & thanks! I'd tried dropping in formatted text from, say, Wordpad, which has worked for me in the past on some forums, but it doesn't work here.

I took a look @ the sticky regarding LaTex. It appears to be "sorta" like HTML --Giving it a quick try here:

\(I_f = 100 mA.\)

Ok, great! Interesting about the additional formatting added. ("medium bold", Alphabetic characters italicized.)

Most likely the maximum power dissipation to keep the junction temperature at an acceptable level. Why? Is your intention to dispute manufacturer specifications or second guess them?

No, I'm not really trying to "second guess" manufacturer's specs, but the reseller's information is very limited. I'd call it more my trying to gain more info. Or maybe it's just my curiosity / background and... some spec sheets, even long ones, seem... incomplete? For example (scroll down to "Reliability") :

www.bridgelux.com/sites/default/files/resource_media/DS55%20SMD%202835%200.2W%203V%20Data%20sheet%20Rev%20B.pd

So, the "High Temperature Operating Life" test in particular appears to be an accelerated test (JESD22-A108D) done in a short time (1k hours) compared to the expected life of such a LED, with limited variations from nominal performance as stated in the notes.. IIRC the actual reseller merely claimed 50k hours but without reference to any conditions other than the rated wattage and indicating use in "12v" applications including automotive.)

However, the Bridgelux testing was done at 105° C instead of JESD22-A108D's* 125° C°, and zero of 22 units failed. This is "weird" to me: Back in my days of "usually one or two loudspeakers in the power test lab 24/7" I almost always tested to destruction, sometimes checking along the way to observe or measure performance variations, nab voice coil or magnet structure temperature data, and so on. In the case of this LED spec, I have no way to know if one of these LEDs will make it / meet specs to maybe 1/2 of the reseller's rated life claim, operated as the reseller suggests. Essentially, I have no way to judge how "tough" the manufacturer's test really was, or even if the manufacturer has a clear grasp on how demanding the test was. I could probably research into how JESD22-A108D was developed, likely forking over some money for various papers, and hazard a guess, but I really shouldn't have to do that.

*https://www.cirrus.com/company/quality/product-development/reliability-qualification/

Maybe this is where part of the "operate at half the spec'd maximum" general guideline comes from, although I always thought that had more to with things like voltage variations, device variations, ambient variations, heat sinking unexpectedly impaired, and so on.

Ah, well, at least in my case I don't have to worry about trying to push a product in a highly competitive market, so when unsure I can overbuild by quite a bit in most cases, and be liberal, when practical, with heat sinking if need be. Sometimes that's a fair piece of extra work though: For example, the rear cargo area light in my wife's car is problematic. The 13w OEM incandescent bulb is a bit weak yet gets hot enough to discolor (darken) the internal surfaces of the bulb housing, resulting in even less light. I tried a couple different LED bulbs: They don't get as hot, but still get hot enough to melt their own bases: the bulb housing really limits convection cooling. I could cut more of the housing away, but that reduces what reflection from the housing still occurs. The best bet looks to be cutting away the entire rear 2/3 of the housing (retaining the door-off-on switch) and fabricating a simple mounting / heat sink for 2 or 3 of those LED modules I've been talking about in my thread about power supplies for those LED modules. Those give a more self-directed light. OR I can try to find a complete aftermarket generic LED cargo light assembly that is shallow enough to fit, and (almost certainly) cut into the car interior (ceiling.) Obviously too small an assembly means some sort of adaptor. Ah, what fun.

(It was more fun when I was 40 years younger!)

 

The specs that a manufacture gives come from a number of different perspectives. Using your LED forward current as an example, the manufacturer is designing the part to a certain minimum lifetime under continuous operation, to not exceed a certain maximum temperature under certain specified conditions, to not have the forward voltage exceed a maximum level, again, under specific conditions, to have the emitted light meet certain specifications, such as lumens/watt or spectral purity. There may be others. Each of these imposes a different limit how how high the current can be and meet that spec. Whichever one imposes the lowest current is the one that is going to determine the published max forward current. Now, no manufacturer wants to design parts that could handle 100 A except for one factor that limits it to 100 mA, so naturally they are going to design the part very carefully so that all of these limits are at about the same level, because that will usually result in the most economical part to manufacture (but not always). They are also going to build in a margin of safety and take into account manufacturing tolerances, plus they will typically target widely used specs, such as I_fmax being 20 mA or 30 mA, and not something unusual like 27 mA.

 

It has already been mentioned a bit, and certainly a great many limits are based on temperature rise and the resulting reduction in product lifetime. And it is always a compromise.
The one exception is the "Absolute maximum" value, because when that is exceeded, immediate damage is very probable.

 

The specs that a manufacture gives come from a number of different perspectives. Using your LED forward current as an example, the manufacturer is designing the part to a certain minimum lifetime under continuous operation, to not exceed a certain maximum temperature under certain specified conditions, to not have the forward voltage exceed a maximum level, again, under specific conditions, to have the emitted light meet certain specifications, such as lumens/watt or spectral purity. There may be others. Each of these imposes a different limit how how high the current can be and meet that spec. Whichever one imposes the lowest current is the one that is going to determine the published max forward current. Now, no manufacturer wants to design parts that could handle 100 A except for one factor that limits it to 100 mA, so naturally they are going to design the part very carefully so that all of these limits are at about the same level, because that will usually result in the most economical part to manufacture (but not always). They are also going to build in a margin of safety and take into account manufacturing tolerances, plus they will typically target widely used specs, such as I_fmax being 20 mA or 30 mA, and not something unusual like 27 mA.

That all makes sense, but in the case of the spec I linked to, surely the minimum lifetime is not 1000 hours... Or is it?

I just don't see how a lifetime or minimum lifetime can be established without incurring some failures. "Failure" being defined as an actual cessation of operation or a deviance in in performance in excess of a stated amount (as noted on that same spec, but none occurred in the stated testing.)

 

That all makes sense, but in the case of the spec I linked to, surely the minimum lifetime is not 1000 hours... Or is it?

I just don't see how a lifetime or minimum lifetime can be established without incurring some failures. "Failure" being defined as an actual cessation of operation or a deviance in in performance in excess of a stated amount (as noted on that same spec, but none occurred in the stated testing.)

Which spec? The first link (Bridgelux) yields a page-not-found error.

 

That all makes sense, but in the case of the spec I linked to, surely the minimum lifetime is not 1000 hours... Or is it?

I just don't see how a lifetime or minimum lifetime can be established without incurring some failures. "Failure" being defined as an actual cessation of operation or a deviance in in performance in excess of a stated amount (as noted on that same spec, but none occurred in the stated testing.)

I recall seeing some incandescent light sources with claimed lifetimes of 500 hours, and 1000 hours claimed for some other lights.

 

Which spec? The first link (Bridgelux) yields a page-not-found error.

the link is missing last character (should be PDF not PD).
https://www.bridgelux.com/sites/def...ia/DS55 SMD 2835 0.2W 3V Data sheet Rev B.pdf

it shows 60mA 3V LEDs.

 

...how is maximum steady state forward current for a LED determined? That is, what are the criteria?

just read the datasheet. manufacturer already did all the testing and wrote the results down.
the datasheet in above link shows that LEDs are rated for 60mA. absolute maximum (edge of destruction) for continuous current is 80mA. obviously you do not want to operate it in such regime bit this is useful info to know.
if powered by pulsed current, peak can be larger than 80mA. datasheet tells that for pulsed mode, maximum duty cycle of 10% and pulse width of 10ms is recommended. And that absolute maximum peak current even if pulsed is 120mA. of course it is not a good idea to push them to the limit of they will really not last.

 

LEDs, 555s, Flashers, and Light Chasers

 

That all makes sense, but in the case of the spec I linked to, surely the minimum lifetime is not 1000 hours... Or is it?

I just don't see how a lifetime or minimum lifetime can be established without incurring some failures. "Failure" being defined as an actual cessation of operation or a deviance in in performance in excess of a stated amount (as noted on that same spec, but none occurred in the stated testing.)

Just looked at the data sheet using the fixed link provided by panic mode.

I don't see any claims about minimum life. They are just reporting the results of some tests that they did. They took 22 devices at operated them at a current that stresses them at the edges of the temperature specs and, after 1000 hours, saw no failures. Consider that informational content, not a guarantee or a specification at all.

If you really need detailed data regarding the expected lifetime of the device under various conditions, you might (might) be able to get that from the manufacturer -- but you might also have to pay for it or convince them that you are a serious potential volume customer.

 

The "edge of destruction" rating is often a value at which a fair percentage of the devices will have failed. For others it might only mean that a few will have failed.
So a wise design will not approach that value.
Those data sheet numbers are provided to assist in successful inclusion of the devices in products, and almost always they are correct.
The much shorter lifetime is based on heat, specific to temperature rise, because LEDs do deteriorate with excessive heat.