Head2Head: LCK26K/N.CS/MJLED/LCK65K/LoV35 (Pt.2)

Re: Head2Head: LCK26K/N.CS/MJLED/LCK65K/LoV35 (Pt.

koala, unfortunately I don't have anything for PWM testing laying around. The application for most of these 5mm LEDs usually involve a resistive voltage drop, a small boost circuit or direct drive off a couple of button cells (LRI Photon excluded). I have been considering doing a cycled test at a higher current (say between 50-100mA). This would allow the die and phosphor to cool somewhat during the off part of the cycle.

Paul

Re: Head2Head: LCK26K/N.CS/MJLED/LCK65K/LoV35 (Pt.

Peak's Snow 29 burn-in plot (3rd post - this thread) has been updated.

The output reached 50% at 118 hrs @40mA.

Here is the final plot posting for Peak's Snow 29 LEDs. These certainly look like 20mA devices (could not get a spec sheet or manufacturer part number from Peak).



Paul

Thanks for testing the Snow. Doesn't look like Nichia is in any danger of losing the Overdrive award any time soon. I *am* a fan of Snow's beam, though. Now if they could just be made a bit more robust...

I also tried to get specs from PEAK for the SNOW 29 led with no luck. After seeing the less than stella results of chimo's tests, I can now understand why PEAK would want to keep the specs to themselves. Like they say....people that have nothing to hide, hide nothing.

I guess we shouldn't be *too* hard on them. After all, Nichia are pretty much the only 5mm that can be safely overdriven. I don't so much have a problem with the Snow 29 results - though obviously i was hoping for better - as i do with Peak getting rid of the Standard option and charging an extra $5 for Ultra.

In plotting all these results from various LEDs, I remembered that I did not have a light meter when I did the initial beamshot comparisons of the LCK26K versus the NichiaCS. In retrospect, the light reading from the Nichia also seemed quite low compared to some later current vs Lux plots. As a result, I have started another burn in on one of the NichiaCS LEDs (bin C/U) at 60mA. So far the output has dropped about 7% after about 44 hours (still outstanding for this current!). I will post a plot when I have enough data.

Paul

So far I have run the NichiaCS burn-in for 6 days at 60mA. The LED is holding out well considering that this is twice the rated current.

My MJLED started flashing. It had minimal usage on it. It was not even turned on for last month or so. I tried different set of batteries and it starts flashing within few seconds of turning it on.

- Vikas

Vikas Sontakke said:

My MJLED started flashing. It had minimal usage on it. It was not even turned on for last month or so. I tried different set of batteries and it starts flashing within few seconds of turning it on.

- Vikas

Click to expand...

Have you used Lithium AAs in it?

The 60mA burn-in on the Nichia has just completed day 6.

The result is that after 145hrs at 60mA (twice the rated current), the output has only dropped 15%. Quite impressive.

See the third post in the thread for the burn-in plot.

Paul

I have completed the 60mA burn-in on the NichiaCS LED. I halted the test after 192 hours (8 days). The output dropped only 15.5% over the test period.

For those just tuning in, it is important to note that these LEDs are rated for only 30mA steady state. These are the LEDs that are used in the new ArcAAA-P model and 60mA is close to the drive current of the ArcAAA.



Paul

i'd be interested in what the 35k led's produce after a similiar burn in test... anyone have any theories?

thanks for all the great info!

jar3ds said:

i'd be interested in what the 35k led's produce after a similiar burn in test... anyone have any theories?

thanks for all the great info!

Click to expand...

See post #2 above. They are listed as Ebay LoV 5mm 35K. Glad you enjoyed the thread.

Paul

This is fantastic stuff, thanks!

I just placed regular "BS" Nichias in my doorbell lights. I resistored them to what I think should be around 5mA. They are VERY bright. So bright, I kept thinking I had not calculated proper resistance.

I had been wondering how long they will last...based upon these nice graphs, I would guess a very, very long time!

Bob

chimo said:

See post #2 above. They are listed as Ebay LoV 5mm 35K. Glad you enjoyed the thread.

Paul

Click to expand...

thansk for the response.. i guess my real question was at what mA is the inova mircolight running the LED at?

Thanks!

jar3ds said:

thansk for the response.. i guess my real question was at what mA is the inova mircolight running the LED at?

Thanks!

Click to expand...

I believe that on full, most of these button cell lights start out at over 100mA but drop quite quickly. I have not plotted the discharge curve - it would vary a lot with different batteries.

If vs Vf update (Fox UV LED included)



Paul

chimo said:

I have completed the 60mA burn-in on the NichiaCS LED. I halted the test after 192 hours (8 days). The output dropped only 15.5% over the test period.

For those just tuning in, it is important to note that these LEDs are rated for only 30mA steady state. These are the LEDs that are used in the new ArcAAA-P model and 60mA is close to the drive current of the ArcAAA.



Paul

Click to expand...

chimo: :bow: :goodjob: Thank you very much for all the effort you've put into unbiasly testing these LEDs for us.

What I would like to ask is if it would be possible for you to extrapolate the above graph till the 50% output of the LED is reached. This will at least give us a hint of how long the LED would last at that current till 50% brightness.

Thank you.

This link is a great read and does an extensive burn-in on LEDs. Unfortunately, it's also about 5 years old and LED technology has advanced quite a bit from that time. The graph near the end gives a a good indication of the output decay plot shape. Although this is a bit of a swag, I would put the 50% point somewhere in the vicinity of 800-1500 hours.

Paul