11
\$\begingroup\$

I am looking at the datasheet for the OSRAM Golden DRAGON® Plus LUW W5AM, a high power white LED.

It has a forward current of 20-1000 mA, surge current up to 2500 mA, and I think they want you to run it at 350 mA; that is what they use for most of the tests.

Notice that I emphasized the lower range of 20 mA. The forward current can not go below 20 mA. That is curious, but taken in isolation I would never have given it much thought.

Now, page 15 has a curious graph, titled Maximum Permissible Forward Current:

enter image description here

Here they iterate the minimum 20 mA operating current, even going so far as to exclude the area from the graph (it is covered with a grey block), and they use the strong words "Do not use below 20 mA".

Why can I not operate my LED at a current less than 20 mA?

Presumably they feel so strong about this that they have to warn me against it. Are there any electrical reasons, perhaps related to long-term reliability? Or is it just that they want me to use a more suitable LED?

I've tried to contact OSRAM about this, to no avail. Maybe it's not important to reply unless the customer is interested in a 100k purchase.

\$\endgroup\$
  • 1
    \$\begingroup\$ Maybe you should ask them why? \$\endgroup\$ – Passerby Mar 27 '17 at 23:00
  • 3
    \$\begingroup\$ Good question. I haven't seen this before. If you discover the answer elsewhere, please come back and let us know! \$\endgroup\$ – bitsmack Mar 27 '17 at 23:07
  • \$\begingroup\$ @Passerby I just might! \$\endgroup\$ – pipe Mar 27 '17 at 23:12
  • \$\begingroup\$ An oddball theory here. Maybe at less than 20 ma, some funkiness in the LED p-n junction causes too much heat or other damage, much like the non-linear region of a BJT? Or like a forward bias counterpart to a reverse bias avalanche breakdown? I still think their engineer support should have an actual answer \$\endgroup\$ – Passerby Mar 27 '17 at 23:37
  • 2
    \$\begingroup\$ I want to second that this is an excellent question. I'm curious now as well. Please let us know if OSRAM provide an official answer. \$\endgroup\$ – anrieff Mar 28 '17 at 0:15
8
\$\begingroup\$

To be honest, I don't really know.

But they also give this graph:

Chromacity Coordinate Shift

It may be that under 20mA the chromacity shift is much higher and the colour of the LED changes noticeably. As the other parts of the datasheet stress on CRI (Colour Reproduction Index) qualities, it may be that the LED wouldn't fit into the specs @ <20 mA.

\$\endgroup\$
  • \$\begingroup\$ Even if you don't know, I have a strong suspicion that you're on to something here! \$\endgroup\$ – pipe Mar 27 '17 at 23:13
  • 1
    \$\begingroup\$ that was what I would guess: in some way the LED does not meet the specifications guaranteed on the datasheet when used below 20mA \$\endgroup\$ – Jasen Mar 28 '17 at 4:36
  • \$\begingroup\$ This is very likely the answer. In photometry, if a low level light source is required, you don't just reduce the power to the lamp. The lamp must run an an optimum current to ensure the correct radiant frequencies. You mechanically or optically reduce the light level by a iris diaphragm or filter. \$\endgroup\$ – Paul Uszak Apr 1 '17 at 21:46
  • \$\begingroup\$ To follow up, I've just received some white LEDs. They're labelled 15500 - 44000 mcd. So they don't start at zero mcd which you might expect. \$\endgroup\$ – Paul Uszak Apr 4 '17 at 14:00
  • 1
    \$\begingroup\$ Actually, @pipe the warning is due to damage to the LED substrate operating high power LEDs below 1% of their rated current causing a great but gradual loss in efficacy. I recall reading this, but ref is not handy. It's not clear how permanent the damage is but may be restored part with pulsed high current or not. But it is definitely not due to visual CCT or CRI. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Feb 28 '19 at 18:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.