I’m fairly new to modding lights, managed to find some sw45ks, and just don’t want to take any chances with them.
Forgive any ignorance, but wouldn’t a new, fully charged, Sanyo GA still pose a risk? It’s rated at 10A, which is over the maximum current when split over three LEDs. Please let me know if this is way off base.
That’s not how it works. The 10A rating for the GA is the maximum safe continuous discharge current for the battery as rated by the manufacturer. It’s a safe operating limit, not the maximum possible current. A dead short would produce many times that current, probably until the cell got hot enough to be limited by its internal PTC thermistor, or until gas is produced, triggering the current interrupt device and permanently disabling the cell. If you’re unlucky, neither happens and it enters thermal runaway and explodes.
Current through a FET driver light is determined by three things:
The forward voltage curve of the LED. Here’s a test of the 219B R9080 with a forward voltage plot. Note the peak output of 700 lumens occurs at 4.0A, well in excess of the rated maximum, which requires 3.3V.
The resistance of the circuit, including any springs, leads, traces, and the FET itself.
The internal resistance and resulting voltage sag of the battery.
If the actual current is 10A or less, which seems likely with 219Bs and a GA in this light based on my experience with other FET driver lights, it will be fine. Most LEDs can take far more current than they’re rated for without acute damage, but with more than 3.0A per 219B, almost all the energy goes in to producing heat rather than light.
If I recall correctly, the old 219B emitters may have had higher Vf than the more recent ones, so the more recent ones are more prone to turning into smoke.
But mostly, I think it’s just because people are more concerned with general safety now, after reports of some Fireflies lights damaging their 219B emitters.
So… I’ll just make a 219 build for the FW3A firmware. It could either have the FET turned off entirely (max output of ~800 lm), or it could have the FET levels decreased to like 25% or 33% or 50%. Not sure which would be better.
Nice! I’ve got a 219B triple board waiting for my FW3A.
I’ve been tempting fate with 219B emitters in a D4. I have the ramp ceiling turned way down and mostly use it on lower levels, but I’ve run it on turbo for a few seconds a handful of times. :partying_face:
Will the FET % be a compile option, or something you determine during testing? BTW, I’ll gladly send you a 219B sw45k R9080 triple board if you need/want one for your FW3A.
I would really like it if the ramp has an indication of when it goes from regulated to hybrid modes like the Emisar D4 standard UI. Or if there’s a pre-defined percentage indication in a leaflet that allows one to adjust the ramp ceiling to max regulated current, leaving turbo for full FET.
The FET percent is generally just a hardcoded thing. I’m just not sure yet what percent it should be. I’ll probably put in a few options which can be selected at compile time, and make the default build something relatively safe. I definitely can’t turn down a 219B triple board though; those are hard to get these days.
It’s too late to change the firmware now, but fortunately that function is already included. It has a brief blip in the ramp when it changes power channels.
About the actual ramp shape, those values are in the config file. They’re also in the user manual.