*BLF LT1 Lantern Project) (updated Nov,17,2020)

So far in my tests at 1.4 amps at the cells on maximum mode with the four LH351D LEDs the top of the lantern has not got to hot even after a hour running on max. ( no higher than 45 ~ 50 degrees) i doubt it still wouldn’t be a concern even with 1.6 amps considering the top is a large solid mass of aluminum.

This is something I do not need but really want… I’m not sure how I’m going to explain this one to the misses but please put me down for one!

It’s an emergency light for power outages - (for her safety and convenience) :wink:

I am increasingly mystified at what people are expecting from a useful practical efficient lantern.

The very concept of having a FET, and a “turbo” mode is utterly ridiculous. I hope I have mis-understood and that was just a naive poster spouting nonsense.

Higher power and better heatsinking could perhaps have been achieved by upward-firing LEDs from the body, the same as most other LED lanterns do it, but this is different. It fires down from the top. And it is a good decision I think. It ought to be capable of easily supporting four LEDs, if driven sensibly at their most efficient points. Not over-driven to ludicrous levels.

Which has been a key design decision, based on practical experimentation. And the correct way to go IMO for a practical lantern.

Don’t listen to the “max power, push the LEDs to the max. 30 seconds on turbo is great, let it burn” idiots. Balance in everything. This is not a torch, it’s a lantern, designed for the long-run. Don’t over-stress anything, keep it to manufacturers’ data sheet parameters, then it might be supremely reliable and out-last most of us.

PS: that does mean that a FET has no place in this driver, in case you missed that.

I think I used that excuse already, not sure if she’s going to fall for it a second (or is it third) time… :stuck_out_tongue:

@Tom Tom, we are using the FET as a linear regulator, not in a direct drive setup.

Do you have schematics for that circuit?


Lexel posted the driver schematics for both a few pages back.)

The Emisar D4 is a hot rod. It can run at like 80 Watts with the right combination of battery and emitters. Its “turbo” is way beyond what it could actually sustain with such small mass. It could easily damage itself or start fires.

The BLF lantern is not a hot rod. It maxes out at around 6 to 10 Watts (TBD), and it has significantly more thermal mass. It may not have a “turbo” level, but if it does, it will probably mean that the temperature goes up to 25 C below the emitter’s rated operating temperature instead of 35 C below. It could potentially get hot enough to be uncomfortable to touch, but it’s very unlikely to cause any damage.

The FET-based driver uses a linear FET design to achieve constant regulated current. It is not direct-drive. Instead, it’s like two of led4power’s constant current drivers. They’re also based on a FET.

I personally prefer the 7135 version of the driver, but it’s not because of anything related to heat or brightness. It’s mostly because the 7135 version uses fewer pins, while the FET version sacrifices or complicates some features. For example, the FET version has no pin to control the button LED, because there was no room for it.

I wish I was earlier on this list… but me on please!

Off to drill out some orifices on the ’ole 2-mantel Coleman… :wink:

Put me on the list if I am not already

Look’n good.

Yes practical illumination with efficient run time for this lantern please.

Max brightness lumen monsters can be trouble… I actually got chided by a camper once, they complained that my lantern was too bright for them in the next campsite like 75-100 feet away. :blush:

(Which is also why I seconded the adjustable shield request.)

Sorry I missed that development. Previously the driver was a 7135 based design AFAIK.

It’s good to see another controllable linear driver developed from discrete components, instead of just slapping down a bunch of AMCs and PWMing them.

How well does it behave ? e.g. dynamic range, moonlight or firefly ability, etc.

A schematic diagram would be of interest, but if that’s giving away a proprietary design, at least a functional block diagram.

From that layout you posted, my initial reaction for Lexel is that there are far too many, and too fine, thermal vias. surrounding the FET, and the USB charge regulator. As well as being costly to produce (drill wear and time) perforating the pcb like a sieve is counter-productive and mechanically weakening. If taken to the extreme, as here, you end up with more air than copper, and I predict one small bump would be enough to fracture the PCB, given that it has to support the full mass of the cells.

Also not happy about relying on the solder resist to insulate the brass ring from the ground plane beneath it at the top, once soldered on. It might work if you specified the solder resist to tent the vias, but it’s not good practice, given the possibility of a hard short to the battery, and most good PCB manufacturers will not tent vias, for legitimate reliability concerns.

Edit: PS: there are also vias in the pads of e.g. C7 and C11, and nearly one in R18, which is usually a no-no. Particularly in this case, when if solder enters the via it will increase the chance of it penetrating to the brass ring when it is fitted.

From my experience top part defenitly not have enough surface area for 10w now.

Lamp may be used also indoor. So it should work normally at least at 30C and no moving air. I don’t want to check every 30 if is right temperature.

It is stupid not to use bottom part for cooling if it is possible. It allows to decrease overall mass and use thermal control.

It is stupid if that is easy to design into the lantern and does not add cost, but I would not call it stupid if it does complicate the design and costs and if tests prove that it is not needed for function.

I’ll take one

No :cry:
That is a picture of the PCB design

Brass ring in latest version has also silk to cover the viases, practically the 0.3mm and 0.35mm viases will be for sure covered with 2 layers ink, there is no solder paste on the area so no risk if the bosrds have not huge silk and mask defects
We can always let in production also let them use Kapton tape for more security than that

Dynamic range with PWMing the FET OPAmps will be higher than with AMCs

Who cares if the fab has increased drill work
Even if it looks like swiss cheese those viases conduct heat much better than a simple copper plane on both sides, you see it a lot on boards with thermal load on it

I have let made MF03 board with this via density and it works, a test with MF02 with 0.3mm viases and less space between then did not fully cover with copper as it was more than 1000 per square inch, also 0.35mm viases work a lot better

also MCU with 14 pin might be an option to get all done