You beat me to it, thanks.
Link to 1st part is
HERE
Here is a summary from the Part 1 thread.
=====================================================
This thread contains testing data on various LEDs. It included lumen maintenance, spot output of various types of LEDs, a plot of forward voltage (Vf) versus forward current (If) as well as some beamshots.
One of the main thrusts of this post is to display the effects of overdriving 5mm LEDs. Some LEDs do not have very good lumen maintenance (how well a LED retains its brightness over time) when overdriven or even when driven at their rated current.
This first post will be updated with summarized testing data.
=====================================================
BEAM SHOTS - LCK26K vs NichiaCS
The Nichia (NSPW500CS - Color B Intensity U) LEDs were purchased from Grumpy's group buy. I had a chance to compare them to Smitty244's 26K LED (ebay 26K) group buy. The beamshot comparison setup was three CR2016 button cells in series with the two LEDs connected in series to ensure they got the same current. The output level was quite similar (perhaps a slight edge to the Nichias in center spot brightness) but I preferred the color of the 26Ks. The 26Ks seemed a more uniform white color and the Nichias had a blue center with a yellow ring.
Pre Burn-in comparisons
The first set of pictures are some pre burn-in comparisons against some known lights.
Before and After Burn-in
This next one is a side-by-side comparison of the LCK and Nichia LEDs before and after burn-in for 147hrs at 57mA. Note that at 57mA the LCK was run at almost 3x rated current compared to 2x for the Nichia. The reason I selected this current is that it is close to ArcAAA current.
After doing a quite a few other tests on the Nichias - the one I grabbed to test as a fresh one was either quite a poor performer or I didn't take a very good Lux reading (probably the latter ). A fresh Nichia's output at this current should be in the range of 43-52 Lux. SEE THE NEXT POST FOR AN ACTUAL BURN_IN ON THE NICHIA CS. The Nichia's output only dropped 15% after 145hrs at 60 mA!
LUMEN MAINTENANCE: LCK 26K at 30/41mA
Because I pushed the LCK 26K LED so hard in the previous test, I ran another couple of tests on a fresh LEDs to determine what a "safe" operating current would be. The LCK 26Ks look fairly stable at 30mA. Some of the variances/measuring error could be due to ambient light in the room, searching for the "hot spot", and changes in ambient temperature. Output drop is 8.8% after 48hrs continuous running. Good drive level for a flashlight but not great for a lantern or nightlight. It appears that 2x rated drive current is a little too much for the LCKs as well. There's a 35.5% drop in output after 48hrs at 41mA.
LUMEN MAINTENANCE: MJLED at 75/105mA
An uncut MJLED was tested at 105 and 75 mA. The results started out looking very promising but went downhill fast at 105mA. The data looks good for 75mA - there has been no noticeable degradation. Here's the graph:
LUMEN MAINTENANCE: LCK 10mm 65K at 20/30/40mA
Here are the results of the 20, 30 and 40 mA burn-ins of the LCK 10mm 65Kmcd LEDs provided by rdshores. These LEDs have a published If spec of 30mA, so I started them with a "relatively safe" overdrive of 40mA. After seeing the results I did another run at 30mA. After seeing those results, I tried 20mA.
The chart can be read as percent of starting brightness measured at 1 meter.
LUMEN MAINTENANCE: Ebay LoV 5mm 35K at 30 and 50 mA
Hookoo sent me some of these Light of Victory (LoV) 5mm 20mA 35Kmcd LEDs for testing. I have started at 30mA (because if it's good at 30 there's no need to go to 20mA /ubbthreads/images/graemlins/smile.gif). Looks like 30mA is a relatively safe drive level. Things go downhill quickly at 50mA - do not use these in your Arc or Dorcy if you value longevity. The epoxy lens on these LEDs are not tapered like the Nichias.
SPOT OUTPUT COMPARISONS - VARIOUS CURRENTS
This is an output test (Lux at 1 meter) of various LEDs. (the legend is stacked according to the plot lines - top to bottom). I ran the three LEDs from about 12 to 70 mA and continued up to 105 mA for the MJLED.
These numbers are for spot intensity at 1 meter and do not represent total output. The beam width is
narrower for the LCK and Nichia LEDs (and
much narrower for the 10mm LED) so they give higher spot readings, however, my impression was that the MJLED gave more overall output at the same current. I base this observation on being able to read labels much better from the bounced light off the ceiling. I may have to make one of those light boxes.
I ran another group of each LED.
This has been presented before, but here is my version.
Here is an interesting chart on LED efficiency vs current. The chart reads as percentage of rated current output vs percentage of rated current. Confused yet? The 100% mark is at the "rated" current for each LED. I placed a linear diagonal line to show what a linear relationship would look like. Brightness is not directly proportional to the current.
For example:
When the Nichia LED is run at 12mA (40% of rated current) its 48% as bright as when run at rated current. In other words, when the current dropped 60%, brightness only dropped 52%.
Also, when the Nichia current was double the rated current(200%) the brightness only 163% of the rated current brightness.
VOLTAGE vs CURRENT (Vf vs If) - VARIOUS LEDs
Here are the Voltage vs Current plots. This illustrates how careful you must be when driving the MJLEDs with a voltage source (just look at the driving voltage and project the current - imagine what current a Li-ion would produce).
Paul