I will be using muggle mode 99% of the time. For that reason, runtime graphs are only in muggle mode. I also wanted to see how temp limit settings affect the muggle mode's max brightness. ToyKeeper has confirmed that temp limit settings do indeed affect muggle mode.
To populate the graphs with meaningful lumens values, the Texas Ace Calibrated Lumens Tube measured the following:
Muggle mode – max output at top of ramp (with clear Carclo optics 10622)
377 lumens at turn-on
373 lumens at 30 seconds
Muggle mode - max output at top of ramp (frosted optic Carclo 10623)
319 lumens at turn-on (approx a 15% drop)
Before going to the graphs, I wanted to see what turbo-mode would give (after exiting muggle mode). The output measured right at the manufacturer’s prediction of 3,000 lumens for the Nichia 219CT option (note the meter auto scaled to X10). The 3,000 lumens reading wasn’t easy to photograph, it’s only there for about 2 seconds.
Four Runtime Graphs (at different temp limit settings)
battery; Samsung INR18650-25R, rated capacity 2,500mah
Note; TL = Temperature Limit
Runtime graphs were done inside my house at an ambient temp of 75F (24C)
In this first graph, I set TL at 38C (100.4F), which is pretty low. A TL of 38C causes the output to settle around 135 lumens. A temp limit of 100.4F is barely warm, so no surprise here. Note the end of the graph, where the light turned off for about 20min, then came back on and started blinking.
In this next graph, I increased the TL to 45C (which is the out-of-box default). Here we see that the output stays a bit higher than a TL of 38C. Note again, the end of the graph, where the light turns off, then turns back on (spikes), then starts blinking. I’m wondering if this is expected behavior, as it does seem a bit odd? There is no low-voltage blinking until the light turns completely off for a while (usually about 15-20min), then turns back on. EDIT UPDATE: The firmware developer (BLF member ToyKeeper) has confirmed that this is firmware issue, and has been fixed in the latest firmware. Only people who ordered this light within the first few weeks should see the anomaly at the end of my graphs. Furthermore, the manufacturer will eventually provide "kits" to reflash the first batch of lights.
In this graph, TL remained at 45C (113F), but I added external cooling (a clip-on fan pointed at the light). Due to the cooling, we see a significant increase in output level, but a shorter runtime. Here again, the light turns off for about 18min, then turns back on and starts blinking.
In this next graph, I increased TL to 55C (131F), no cooling. As expected, lumens values are higher than the default TL of 45C.
Note, this last graph is after the firmware update. As you can see, there is no longer an anomaly at the end of the graph. Note also, there was an increase in output with the new update. This graph was done with the frosted optic on the D4V2. The above graphs were with the clear optic.
PWM
As noted in the original D4, a PWM of 15.7Khz was also seen on the D4V2 (at all levels I tried).
OFF CURRENT (parasitic drain) in muggle mode:
Using a Fluke 17B+, we get 23ua. However, that value cycles between 23ua and 3ua about every second. This implies the driver board is performing some regular routine.
Temperature reading:
I took the time to measure some temp readings in muggle mode with TL set to 45C (default). We see that the temperature peaks at around 12 minutes (120F head, 109F body). That's pretty warm. I’ve personally set my TL to 40C as a compromise.
head/body (Fahrenheit)
5min: 108 / 97
10min: 116 / 105
12min: (peak): 120 / 109
15min: 116/108
Hope this data was meaningful. I like this this light enough - that I've ordered another one (dark grey), and with a different LED; Cool White XP-L HI V3 1A (6500K).
