[1900k edition is in!] WTS: 1900k-5800k 5mm LED 95+ CRI

I promised rngwn several weeks ago I would do a destructive test on some of these LED’s. I’ve been very delayed due to life being busy. However, I did this week finally have time to load ceilingbounce on my wife’s old smartphone and do one test on a 2300K sample.

It ran at 150mA for several hours before I had to stop the test because I couldn’t monitor my sketchy setup any more for safety’s sake. When I let it cool down and restart, it was still nearly as bright as at the start of the test. I was surprised it survived longer than 30 minutes!

I think the test is unrealistic, however, because at only 1/2W, even the legs provide significant air cooling. To better replicate a flashlight, I plan to repeat the test with the legs trimmed short, and the LED facing vertically so convection is more limited. Depending whether it dies quickly 150mA, or shows no signs of dying in a reasonable test duration, I will repeat it on another sample at higher or lower current.

Hopefully I will have results this weekend. These are really nice low power LED’s for tinkering with, so I think it is good for the community to have reasonably accurate information about how far they can be pushed.

I think I misunderstood the answer to a previous question about this.

Is that PCB the guts of one of the Ikea lights? If you swap in an rngwn LED, does it still flicker?

The tea lights my wife has do not have any dedicated controller. The 5mm LED itself has some sort of internal flicker control. If I replace the LED, the tea light no longer flickers.

I want to replace the monochrome amber LED’s, but not lose the flicker effect.

The 2300K LED’s significantly outperformed my expectations when abused. I trimmed the legs so there would be no excess exposed to the air to help cool them when inserted in my solderless breadboard, then hooked them up to bench power supply while recording output with Ceiling Bounce.

I was hoping they’d last 30 minutes or more at 150mA. Normal specifications for a 5mm LED are 20mA rated and 30mA max.

After 12 hours of gradual decline, I stopped the test, let the emitter cool down completely, then started it again. It restarted at 80% of new.

Given that excellent performance, I tested a second sample at 200mA. It took 10x rated current to actually cause relatively rapid damage. After 6 hours, we were finally down to about 10. I let it cool down again and it restarted at 18.5 of it’s initial peak. Forward voltage declined by about 0.2V over the course of the test.

Keep in mind that your results may vary, especially if they are allowed to get hotter.

I’m including 2 graphs of the results. Ignore the noise in the 150mA test. I’m pretty sure it was a setup issue. Please let me know if the graphs don’t display properly.

As a baseline, here’s djozz’s output measurement.

Nice data! These can for sure take some abuse.

Thx for the test!

Now i love my stash even more :)

Thanks for the test. :slight_smile:

Picked some up a while ago and they look great. Work very well in old mini-mags! I’d actually be very interested in a narrow beam version. I have some photons and some Gerber IU’s that are just begging for a high CRI upgrade!

As far as the situation goes, I only have 19 countries left that can be shipped through regular postal service. All these remaining countries except Russia (and China) must be shipped to via Surface Mail (2-3 Months minimum).

Alternatively, there’s DHL service still available for most countries. The shipping cost for that is from $35 to $100 varying by countries.

Thank you for the 2300K test data. We know that the these flea powered LED’s won’t last too long at 150mA. I would be interested in knowing what the max mA they can handle with no degradation in output for maybe 100+ hours. No joy for me in toasting these little gems. I’m guessing a safe limit may be 75mA or so? For night lights and candelabra the 5-7 lumen range is good, if I need more from them I would add more LED’s.

If you want to run this continuously then keep it at 30-35ma or below. Otherwise, the degradation might become apparent after 1000-2000 hours of operating or so.

For flashlight use, pushing it to 50-80ma range might be okay for occasional use. The loss in output might not become apparent over several years.

You only need 20ma for these to produce 7 lumens.

I finally had a chance to graph my data for the 3400K version.

150mA was too high for this version, but it performed better at 150mA than the 2300K version did at 200mA, so I tried a second sample at 125mA. At this level, performance was very comparable to the 2300K at 150mA. As with the 2300K, this was with the legs trimmed short, installed in a solderless breadboard. Power was by a constant current bench supply.

I view this as also very good performance, although the 2300K seemed to have a slight advantage.

One other 3400K sample failed getting setup for the test. I’m not sure what to do to investigate it further. I was having trouble with a bad connection causing it to flicker on and off. There’s a chance that caused a large transient from my power supply, and regardless, for the price, I’m not worried if I find a few duds.

2300K Baseline: 5.8 lumens at 25mA, 22 lumens at 150mA
3400K Baseline: 7 lumens at 20mA, 22 lumens at 100mA

I’ve got a couple “warm white” nightlights that looked rather green even with no other points of reference. Auto white balance suggests they’re around 2300-2400K.

When I swapped one of these 2300K LED’s in, the difference was stark. They went from one of the worst tinted LED lights I’ve seen to one of the best.

I guess Nitecores Tube exceeds any safe current, right? I made a couple of them but never gave one of them away because I thought high mode could kill the LED if used longer than one minute or so.

My Tube has a 3200K Rngwn led and it is doing fine. I believe I measured it and it came at 80mA (not sure it was 80 but under 100mA for sure), while these little leds are even fine for hours at 100mA. They will not reach 50,000 hours lifetime probably. :wink:

Thanks, djozz. So, now I have some gifts :smiley:

Update: Some countries are opening up as a result of easing the lockdown measures. You can check if your location is eligible here (Last updated June 2).

Surface mail or Surface parcels don’t really count, I tried that a few times and none of the attempts were successful.

Air Parcel post is pretty expensive and will cost you $21 minimum. For U.S. location, it will cost $33 at the current exchange rate.

TL;DR - Standard shipping now available for:

  1. Austria
  2. Belgium
  3. Bhutan
  4. Brazil
  5. Denmark
  6. Estonia
  7. Finland
  8. France
  9. Germany
  10. Greece
  11. Italy
  12. South Korea
  13. Malaysia
  14. Poland
  15. Portugal
  16. Russia
  17. Singapore
  18. Spain
  19. Sweden
  20. Switzerland
  21. UK

Countries not in the TL;DR list will either cost more or must use courier post only.

Btw, the Ledlenser K2 is a candidate for an easy swap. No soldering required. AG13 button cells, unfortunately. Works better than those $2 zinc alloy lights. Something for the car as those cells last forever without leaking. Or exploding. I’ve seen 30yo pocket calculators still running on original AG13.

Made it today while pan was heating up for the i3E emitter swap :smiley:

Any tips/tricks for getting the driver out? I’m doing something wrong apparently, and keep ending up with dead drivers. I’ve had a few swaps work, and a few more that didn’t :frowning:

You need to unsolder the LED first, dk. It is held back by a plastic retainer.

Thanks! I’ve tried that, and I’ve also tried carefully working and angling the PCB/driver up sideways, then unscrewing the retaining ring and desoldering. I guess my soldering must just suck and I’m doing something to the boards, they seem to die easily.