Well, the lantern is supposed to be using an absolute max power of 3x7135 per channel, so a max power output of 2.1A is doable without any throttling at all.
4,9A on the other hand would require thermal regulation.
Back in the day people like MTN were cramming in loads of 7135s onto bespoke drivers for SRKs etc. This has got, I think, 16 total. 5.36 A. Which is plenty for a usable big Q8 sized 4 emitter torch, well heatsinked, that you just want to use indefinitely, rather than in short blasts with a DD FET.
For this lantern, which was never intended to be a mega-blaster, the quantity is excessive. But it is there in the driver design, if wanted.
Remember, it will only have two LEDs of each colour temperature. Total 4. To be modulated for tint-shifting. Averaging only two perhaps. With a questionable thermal path. And is supposed to be efficient and long-running and dependable at what it does. For many hours, or days even, providing useful illumination, even being recharged from a small solar panel between times (design aim, lets see how well the charging circuit manages that).
Quite different from typical BLF pocket-rockets.
It is supposed to last for many many hours. Draining over 5 A is not compatible with that, even with four good 18650 cells in the tube.
That said, I would have a use for this driver in my Q8, in the real world, with suitable firmware. Two banks of 7135s connected in parallel through through the LED wires could work nicely. And if the charging input would also efficiently accept input from a small 12V solar panel (e.g. car battery maintainer), that would be icing on the cake.
That was all part of the original concept. Has that all been forgotten ? Or what a lantern is for ?
Maybe I’m the one who missed it, but IIRC, the driver design is limited to 7 chips per channel, total of 14, for 4.9A (unless the 380mA chips are used, then 5.32A would be possible). Several of the long-time members have done stacking in the past to get more, and they could presumably do that with this driver. But most of us current members wouldn’t.
For this lantern, 7 chips per channel is much more than is needed. And the charger is limited to a low rate, despite many members wishing otherwise, because many solar chargers can’t handle too high a current load and will shut off, failing to charge anything at all. So yeah, that concept (using a solar charger for this light) is very much still there.
The present PWB has two 7135s hard wired to drive the two banks of lights. Up to 7 can be connected by applying a solder bridge to the pads on the battery contact side of the driver. Yes, if you do this then you will need to address thermal concerns also. The battery side also has programming pads, and some other bells and whistles I don’t recall presently.
Yes, but it’s 2 channels of 4. Not 2 channels of 2. At least that’s what it was supposed to be. You can see the picture of the board in the OP, or click the img in my last post (not sure why it isn’t showing).
Correct on that… things have been very rough as of lately for me with life issues, financial issues, stresses, & preparations to sell my house and sell off a lot of things to move to a different province, but i still will make time to work on this lantern project here with the team here to do what i can to make it a reality as much as i possibly can.
I think they should be able to get good parts. If not though, the effect would be a bit of change in brightness while changing tint:
With ideal parts, brightness should stay rock-steady while adjusting tint, like a flat board placed across two support beams.
With lower-quality parts, particularly with slower activation times, brightness would sag a bit at middle tints, like a cable strung between two telephone poles.
Given how visual perception of brightness works though, even the non-linear version would probably be a pretty subtle effect. And if it shows up in testing, perhaps the firmware could try to correct for it.