[PART 1] Official BLF GT Group Buy thread. Group buy officially closed! Lights shipping.

I could really care less either way.

If we were doing serial numbers, I think 1 of 555 - 2 of 555 - 3 of 555 would be cool for this run. Then 00556 - 00557 - 00558
Or maybe just put “First run” on this batch, or “1 of 555” on all of the first batch…

And, I dont really care if there is no serial or special markings on it. Would be a neat extra, but does it really add to the light? To be honest, I am on the list for one and would like to honor that, but depending on the final price and time of release, I may not even be able to afford to jump on it… times are tough in the Pulsar household. So my opinion may not matter

All opinions matter :wink:

Yes if we do serials it is gonna be ” 001 of 555 ” and so on
Nobody will be able to choose a number they will be send random

Good news, DEL made the first driver prototype
It works at 2,24A with good 90+% efficiency
Parasitic drain is acceptable at 156uA and possibly a bit lower.
Heck of first proto driver results!

Good to hear that, the runtime will be very long
Btw: i love the idea engraving the number on the light, it will be more valuable in my collection

MARK core list updated

Thanks. Cant wait for it to be finished.

I’m in for one piece of this great flashlight!

+1,000,000

Good news, DEL made the first driver prototype
It works at 2,24A with good 90+% efficiency

Parasitic drain is acceptable at 156uA and possibly a bit lower.

Wow this is pretty freak’n awesomely excellent !!! :wink:

Great work DEL!!! :+1:

Man , i can’t wait to get one :slight_smile:

As The Miller mentioned, the prototype driver is built and mostly tested. It is running a tweaked version of Tom’s NarsilM and it has been performing pretty well.

Some numbers:

  • Switching frequency is at 800 kHz with the selected components and at 16.8 V input.
  • 100% output is 2.25 A, but this easily adjusted by changing Rsense. Aim is to have >= 2.5 A stock.
  • Everything is running cool. ‘Hottest’ components are the buck chip and 100 mohm sense resistor, and both have a Trise of barely 20 C.
  • Efficiency is 91% at 16.8 V, peaking at around 93% as the battery voltage approaches the output voltage.
  • With my ‘LED load’ (4 x 3 V emitters on a beefy heat sink), it stays in regulation down to about 13.8 V (3.45 V per cell). An actual XHP35 should do better.
  • As the battery drains further the LED current fades gracefully. About 2 A left at 13 V, 0.8 A at 12 V.
  • Parasitic drain is 156 uA. The bulk of this is the buck chip, unfortunately. Will bring it down to about 125 uA by upping the LVP resistors.
  • The firmware has been set to get pure analog dimming between 100% and 20. From there it is further dimmed using 15.6 kHz PWM. Butter-smooth Narsil ramping! (The 20 analog dimming limit is not set in stone, but some efficiency is sacrificed if trying to go much lower.)
  • Getting a moon mode of 0.44 mA, using a combination of analog and PWM dimming.

Beautiful, DEL. Do you think a lot of efficiency will be lost by pushing the components to 2.5A? What about longevity?

Is it just me, or did the project get a much faster pace in the last few weeks?

Orsm work DEL. :+1:

That is one nice looking driver :slight_smile:

That’s a beauty DEL, remarkable work! :beer:

The heat form this driver should be a non issue in the massive light! :+1:

hey im interested, can you put me down for one.
dont want to sound impatient but how long before its going to be available. Or even see a working prototype to be able to get real world figures and outputs from?

I wish there was a buck driver like this that could do 12A+ :frowning:

NICE DEL!
beautifull driver for sure!
And awesome how you write stuff like: “but this easily adjusted by changing Rsense” and “Will bring it down to about 125 uA by upping the LVP resistors.” it’s like magic and you’re a grand wizard for sure!

Will update list later

Thanks for the kind words everyone!

Regarding capacity, the design is very conservative and it should be able to run 5 A with minor modifications. Doing that will have some components run pretty hot though.

2.5 A is only a 10% bump from where we are now and is not an issue. Only reason I am at 2.25 A is the availability of good current sense resistors. Would have been nice to stick with 0.1 ohm.

Very nice driver setup there. Only thing I could see to improve it would be a 1617 mcu.

+1

I will start a NarsilM thread soon. Might not have DEL's buck driver mods in it for the first release but that's the plan. Hoping it will be just another compiler switch in the setups header to choose between 1, 2 or 3 channels or buck.