Led junction temperature over stress?

[burningmosfet]

Hello everyone I'm working on a project that involves the construction of a torch with a 1W white LED Xlamp Cree XP-E and I have some dissipation problems.

The datasheet of the component is here:

**broken link removed**

The LED is driven in constant current 300mA and was mounted in the center of a 1mm-thick pcb and area of 0.4 square inches. On this pcb (to save money, it was decided not to make a metal-core pcb) many vias were placed to let warmth pass even on the other side.

We then put the torch in a climatic chamber (without fan running) and noticed that at an ambient temperature of 30 ° C, on the thermocouple (placed very close to the LED) the temperature goes to 90 ° C.

Now if the temperature of the pcb near the led (and thus the temperature of the case) is 90 ° C and the thermal resistance of the component is 9 ° C/W, with a power of 1W should be, the junction temperature should remain below 100 ° C.

Tj = Tc + RCJ * W * 1 + 9 = 90 = 99 ° C.

The maximum junction temperature allowed for this LED is 150 ° C.

In your opinion, with a junction temperature of 100 ° C, will the LED be overly stressed and his life shortened, or can I consider it a normal operation?

Thank you in advance

B.M.

 

[alexan_e]

Page 13 of the datasheet says:

"Lumen Maintenance Projections
Based on internal long-term reliability testing, Cree projects royal blue, blue, green and white XLamp XP-E LEDs to
maintain an average of 70% lumen maintenance after 50,000 hours, provided the LED junction temperature is main-
tained at or below 135°C and the LED is operated with a constant current of up to 700 mA."

You are using half of that current and the temperature in the junction is probably a little higher from the 90 degree measured
but should still be much lower from the 135 so i think you are OK.

Alex

 

[leo_o2]

Tj=30° C+(9 ° C/W)*(1W)=39° C. Is your calculation correct?

 

[engineerpervez1]

same type of job I want to do, but more
LEDs and with 12 volts battery.
5 nos of LEDs and each 5 watt. to get reasonable room light at areas where the home power
supply is not available.
I saw some ICs from supertex, but these are not available in our country,
I bought some Bright lights LED without data sheet. but I tested single LED is fine with 6 Volts and 350 mili ampere.if i connect these leds in series 5 nos or six leds in two parallel lines
say 3 leds in each string. what will the circuit diagram to manage their current rise and control.

 

[Genomerics]

Referring to the land patterns on page 14 of your data sheet then I would expect that ideally rather than measuring 'close' to the LED you might wish to maintain one of your vias in the centre of the middle pad clear whilst soldering and place your thermocouple through it to make direct contact with that pad on the device itself. Of course you will have to make that particular via large enough to accept the thermocouple. Fill any void with thermal compound.

This should give you the best possible idea as to what the device temperature is at the point quoted in the data sheet, 'Thermal Resistance, Junction to Solder Point'.

As mentioned elsewhere page 13 suggests,

Lumen Maintenance Projections Based on internal long-term reliability testing, Cree projects royal blue, blue, green and white XLamp XP-E LEDs to maintain an average of 70% lumen maintenance after 50,000 hours, provided the LED junction temperature is maintained at or below 135°C and the LED is operated with a constant current of up to 700 mA.

and makes reference to other supporting documents.

Genome

 

[burningmosfet]

Thank you Alex per le your reassurances. Pervez1 your calculation is mistaken because the junction temperature in my case is:

Tjuntion = T[case-led] + Watt x Rth[case2junction]

---------- Post added at 20:56 ---------- Previous post was at 20:53 ----------

Thanx also to you Genome but i'm unable to place a thermocouple as you say