Oshpark Projects

I ran an Efest IMR18650V2 through it for .02, .34, 1.63, and 4.15A
So then I ran my badass 14500 charged to 4.19V through it and it did .02, .35, 2.16, and 4.52A.
Wanting to see what else she could handle, I broke out the Samsung 25R charged to 4.17V. It did .02, .38, 2.34, and 5.13A. Seems to me it’s pretty capable of pretty much any cell I have. :slight_smile:

How hard is it likely to be pushed in the lights in which it will be used? I think we’re already past that but you know you want to find the ceiling first. Push it ’til it fails or set that board aside and make another to kill. Chips are cheap, much cheaper than an equivalent number of 7135’s.

High current or high heat?

Scott, I’m betting the only thing that could kill that 10440 is a combination of all things evil…high current draining it to very low levels while heating it to boiling point. 5 minutes under an FET driver is brutal for that small cell. But, djozz DID say it still works! :slight_smile:

I’ve ordered some 15’s for testing. I have a few of the 2502’s left for this.

I do not know for sure, my first reflex is: high heat, the light got so hot that I burnt my fingers twisting it off. but it had been running at 2+ amps for most of the five minutes by then.

It’s important to know the failure points of our little light bombs so as not to give away a tragedy.

It is indeed important to know what these lights are capable of, but it’s difficult to go by my findings if I manage to fry it cause I do all kinds of other things wrong. :wink: Nature of the beast and all that.

But you know, if it can put out over 5A then I’m going to be pretty happy with it just under 3A. Might even have to change some things in the firmware to limit it to say 80% or something, just so it doesn’t hit those really high numbers.

If this is a FR4 PCB, I would like to suggest you to put at least 20 thermal vias on the pour copper and near the both sides of the LED’s thermal pad.
The recommended thermal vias size by Cree is 0.3mm diameter. Here enclosed the Cree’s FR4 thermal study for LED. Dropbox - XLamp_PCB_Thermal.pdf - Simplify your life

PC = phase correct (9.4kHz, 'TCCR0A = 0x21'), FP = fast PWM (19kHz, 'TCCR0A = 0x23'). Levels for both are 0,2,6,18,54,130,255. Both turn on and off and change modes exactly like they should. The phase correct board makes a slight whine on level 4, moderately loud whine on level 5. Fast PWM board is silent.

Why do these boards work fine without the gate/pulldown resistors? These are the exact same components used on the 17DD, which does not work without the resistors.

Thus I think STAR V1.1 and a timeout is the safest way to go

And/or have a approx 50% PWM always on…this way you don’t just run it wide open, I wouldn’t use that battery, it has vented…just not “explosively” yet!

Being an IMR battery it is unlikely to vent explosively, obviously something has happened to it, but it just now was still able to deliver 3.5A to the light. But I guess you are right, I will stop using the battery and dispose it .

I successfully rewrote my UI! :slight_smile: I made a new .hex file for the BLFTiny10 that replaces the 255 Turbo with 190, and the previous High setting with 90. After flashing the little guy I ran it through on the IMR10440 and it does .03, .31, 1.16, and 2.20. Perfect!

My TP has a Nichia emitter in it, don’t think it could’ve handled the full out direct drive. So now I get to try to make a contact pad over the FET and diode and then see if I can put it in my light. This is exciting! :slight_smile:

Woohoo! I now have a BLF Tiny10 FET in my Texas Poker! It’s running, from an IMR10440 at 4.15V, 3.45 Lumens a .02A, 34.50 Lumens at .18A, 127.65 Lumens at .67A and 262.20 Lumens at 1.40A! :slight_smile: Finally!

I had changed the driver in it before to get 188 lumens OTF, at 1.1A. So I’ve nearly doubled it’s output level on High or Turbo now, as it has 4 levels now to pick from. I’m stoked! :slight_smile:

Thanks Matt, for all your help, and Scott…this little FET is the charm we’ve been looking for for almost a year! :wink:

And the cool thing here? I’ve got the FET running at 75%! Can change it at will now, thanks to a lot of folks here that took the time to show me how, and got to laugh at me while I stumbled through it all. :stuck_out_tongue:

Is it known if there’s any adverse reaction to controlling the FET through the firmware? Say, setting max to 75? This is how I’m running my TP and it seems to be fine. And of course, the flashaholic in me really wants to find out what 80 looks like…90? :slight_smile:

It's no different than using a lower mode that uses the same PWM value (around 192?).

I set Turbo to 190 and Hi to 90, the lumens and used amperage work out pretty nicely. High makes 135 and Turbo makes 273.

Thanks for all your work with these MOSFET’s Comfy, really appreciate it, making this already loved light that much better! :slight_smile:

Oh, and for the record…no resistors for the gate or pulldown, works like a charm without em. :wink:

Which firmware? Fast PWM or phase correct? (if it's one of the STAR versions and you didn't change it on purpose, it's using fast PWM, hence why it doesn't squeal like a stuck pig). ;)

SWEET!

I am thinking the smaller 14500 and 10440’s will need a “choke” aka the 190 or 74.5% of maximum output…

So you always have a PWM control but it flows MUCH more power than any stock driver they can shoe horn in 10mm!

Awesome job!

oh and no laughing bro…we (especially me) are still stumbling around :smiley:

Oh and throw us up some shots of the Texas Poker light man :stuck_out_tongue:

I just noticed that this board is not listed here, so here the 7 XM-L board is.

http://www.brightlightmodules.com/Images/7xml_scaru.zip