Originally Posted by Diode Dynamics
We will be introducing 4300K and 5000K versions of the XML2 within the next few weeks. The XML2 is currently available in 6000K only.
The LED bulbs that we cross-tested in that video posted above were from K5Optima store
Thanks Nick. I prefer those warmer tints over the cooler 6000K version. The 4300K would be a perfect match for my HIDs too.
I checked out your video and was surprised to see the sharp drop in current from the H8 bulb you say was from K5Optima store. I performed my own bench test and, although the current did drop over time, it was not as rapid nor go as low.
My bench test was set up as follows (see pic below for reference)
Test Results (see graphs below)
- Current and volts were measured using a digital volt/ammeter.
- Power was supplied by 3 x LiMn batteries in series charged to full capacity (4.2V each)
- Light output was determined by a lux meter and readings were taken with the meter probe place in such a way as to capture the indirect light from the bulb (also known by hobbyist as "ceiling bounce). An integrated sphere is really the proper device to use for accurate light output, but this was the best I had. Thus, lux readings are just meant to be a relative comparison of output change over time.
I took measurements at the very start (time 0) and every minute for the first 10 minutes, then at 5 minute intervals up to 30 minutes, at which time the test was terminated.
As you can see, current starts at ~800 mA, then steadily drops to ~500 mA during the first 10 minutes at which point it stabilized until the end. Voltage sag was minimal (12.2V starting to 11.9V ending), so the drop in current was most likely due to heat build up and not voltage drop.
The lux readings dropped with current (as expected), but did stabilized once the current settled down to ~500 mA.
Overall, this appears to have performed well, given the high number of Cree XBD emitters (16) and relatively minimal heat sinking. I don't think much can be improved on this, except using a more efficient emitter (XML or XPL) and perhaps more mass for heat sinking. However, unless the heat can be removed, once the unit heats up, current sag will prevail as there is no place for the heat to go. This is really a case of striking a balance between output and minimizing heat build up, much of which is controlled by the electronics.