Cree XHP70 up to 4022 Lumens and XHP50 up to 2546 lumens - Multi-die leds.

When I first tested it the cells were at 3.95V, almost dead, after the lightbox run. I didn’t run it just tested it. The first reading I got was 12,523 lumens.

So of course I charged the little cells!

13,903.5 lumens out the front, even low does 1408.

I lined up the emitters such that the line between dies pointed towards the middle. This put them diagonal to each other. And the donut hole is there at low, but decreases as the brightness goes up. Turbo is a stupid amount of light!

These are the 6500K variant, the first ones out.

This Toshiba is a new release, it has a plate on top like the one on bottom that solders to the driver. This is supposed to help with heat. It’s rated at 150A, 30V.

Can’t help but wonder what a good set of top end 18650’s would do….

Full out belly roll laughter! Damg, my face hurts! :bigsmile:

:bigsmile: :wink: 0:)

What is the exact name for the toshiba? so i can check if it is available on stores thats available to me like farnell.

But it still is a LFPAK56 sized FET? have there been any discussions in any thread’s on it here on BLF?

I saw it when in announcements when I was looking for something else, looked good, bought 5 of em. $3 each.

TPWR8503NLL1QCT-ND

I have no idea if this performs any better than the SIR800 or PSMNO9R or anything else. I’ll be using the SIR800 mostly now but thought I’d try this out. :wink:

That FET might actually do better on a driver like the LD-1.

Thanks i found it, i guess i buy some of the various like the SIR800 & PSMNO9R because farnell don’t got it yet.

I hope one our spec gurus like wight, comfy or Rufusbduck will weigh in on if TPWR8503NLL1QCT-ND is a good option, i have a really hard time wrapping my head around the spec and the math comparisons on MOSFETs……

EDIT and here he comes…… :slight_smile: thanks Rufusbduck

Would you say the SIR800 & PSMNO9R should still be the better options?

I wants it! I wants the precious!

I put that down to your retinas catching fire :wink:

Very impressive looking FET btw, the RdsON graph is solidly at 1mOhm for our usage range. Low threshold voltage as well, looks really good to me.

Is the “thermal pad” on top electrically connected to any of the other pins? I can’t find any info on that in the datasheet.
Cheers

In a much higher application than ours it’s to be set up in contact with the housing or heat sink for maximum cooling. I know when we were working on the Knucklehead, there was a lot of conjecture about how to cool components so this looked like a great idea to me. I intended to make a small copper heat sink to solder to it, but realized even pushing 3 XHP70’s it’s way down on the bottom end of it’s intended use. Maybe 30A? In a component designed to do 150A? Yeah, not really an issue.

So the tab is connected to the gate? That’s really strange, I would have guessed/hoped it was connected to the source/gnd pin.
I was looking for a package like this a while ago and couldn’t find anything.
That would have been very handy for improved heatsinking when used in linear drivers (where heat dissipated would potentially be much higher than in a DD setup like this)

Edit: Just had a quick look and that DSOP Advance thermal pad does seem to be connected to the source pin. That’s awesome. I’m going to have to get a few of these things! :slight_smile:

I said gate, meant that it’s not to be used as an electrical connection.

I just thought the tech was neat. :stuck_out_tongue:

I mean, I read the datasheet and it looked like it would be good for our application and I had this triple XHP70 in mind and was thinking, if there was ever a place… so I tried em.

I don’t know electronics. I don’t know why with the same firmware that yields a sub lumen moon mode in my triple XM-L2 this one makes 1400 lumens. That came as a surprise! The lowest position on tap is 1400, not much of a low huh? lol

Ah gotcha. Yes I see that they don’t suggest using it to connect GND directly, but as long as the only heatsink path isn’t soleyl through the drain tab I still see this being very useful in flashlight applications.
Having to keep the drain tab electrically isolated from the body of the flashlight limits efficient heat sinking options quite a bit.

Is that with the same mosfet on both or compared with the standard ones?

Sounds like there’s something interesting going on anyway. Is it possible the low you’re seeing is actually a medium mode and the lowest is completely off? That seems more likely to me if it was a mode compression issue related to the dynamic mosfet properties.
You’d have to hook it up to a scope to see what’s really going on with the PWM signal.

Bloody mosfets are tricky buggers to understand! :stuck_out_tongue:

I thought about trying to add a copper sink on this, enabling it to possibly have more cooling, but the way I built the driver it would have been too close to touching the back side of the emitter shelf, not good electrically OR thermally!

I’ve got 3 18ga positive leads and 3 18ga negative leads supplying the A17DD-S08. :bigsmile: Think it’s got enough access to the cells? Solid as a rock! An 18ga wire about 1/2” long and full of solder is quite the little support stanchion. 3 of em side by side is a veritable tree trunk of power supply.

I was just reminded that the original firmware was written to allow the stock 3ONO2 mosfets (the bigger boxy ones) on the original M6 driver to run 3 XM-L2’s. It’s waaay higher in the low setting because of this and I should reflash it to bring down the low end. But this is a blaster, 1400 on low is fine, it’s not like I expect to go from 0.3 lumens to 14,000 in one light here!

Just curious, but have you considered just for science, to see how offensive a 13000 lumen blinky mode is? Not sure if the driver can be made to blink.

I set up one of the XHP70 leds on a 20mm star and put it on my big heat sink. I tried out four different reflectors. I dod not raise the led off the srat with copper shims. I wanted to see what would happen for most people, if they just tried to put in a XHP70 and did not have the means to build up the led height above the star.

The heads of the screws are filed down about as far as they can go and the wire is 18ga, with thin insulation. All of that still sticks up higher than the die face and that would be about how you would find in one of the smaller lights, with little room.

The reflectors are a mix of odd stuff. From left to right in the back row - a 50mm Shallow stippled, a 50mm deep SMO, a 60mm very deep SMO. The small one in the foreground is about 25mm and shallow SMO.

As you can see, the reflector sits on the die face. This is what you might find if you are not able to open the reflector up enough to get around the base of the led.

Another reflector with the same problem.

The 60mm deep reflector sat over the die face and almost touched the face of the star. The small reflector also sat down low, but it had tabs that still kept it off the star by about 1mm.

(Sorry, that's SMO, not SOM. Dyslexia gets worse every single day).

Here's what I think. - I think that in order to get a good beam, you will have to use a large diameter deep reflector and you will have to sink that reflector down to the same height as the star, to get rid of most of the artifacts from the four die led. OR, you can use a shallow reflector and have it OP or Stippled, to blend in the defects.

The shallow stippled and the 60mm deep reflectors looked the best. The shallow one gives a large spot that blends out to the edge. The deep one has a defined spot and is almost right. I think if I could get it down that last .5mm to the star, it would be about perfect. Well, as perfect as it's going to get with a four die led.

I also think this will never be a thrower as such. It will give out tons of light and that will make some think it's a thrower, but there is a big difference between a true thrower and a mass of lumens going OTF.

That’s about what I’ve found as well. Floody and lots of it, difficult to focus into a beam.

The triple can be manipulated such that the artifacts are self cancelling. Keeping the “+” at angles goes a long way to make for a nice beam profile when using multiple emitters in this variant.

Nice pics Justin, thanks for the detailed exam on what works and what doesn’t. :wink:

The plan from the very beginning was to have a variable strobe with stupendous output. But the strobe firmware which has been worked on quite a bit didn’t function well with the Zener mod. Something in the timing of the switch presses is thrown off.

Still the plan….

Also working on getting a carrier made for 2S2P 18650’s for longer run time and higher current levels. As it’s running right now, each emitter is making 4614 lumens, so they’re getting over 4A apiece for sure. The little Efest 18350’s are performing far better than we have a right to expect. 35A’s would rock this bad boy! It’s already just 97 lumens shy of 14,000, figuring on the bigger cells getting 15K or more. :slight_smile: