Single 18650 XHP50 Driver! Self designed driver, check it out!

It’s all packaged into the pill, and installed into the host (a 6 dollar eBay “3000 Lumen” XLM flashlight) I’ll hard link a pic in a sec here.

I think I’m going to be looking into a copper pill/maybe 2oz copper pours for the pcb. I did my best with the 1oz pour, and I wasn’t really able to get the trace width where I wanted it. It isn’t a “mode” light yet, but I think I can add that later, it might end up being a double side 4 layer board, which sucks for price. I am too scared that there is almost NO room left to add a tiny to the board.

I have a final revision that was sent off that I think will be able to hold heat properly and not go into thermal (another feature I forgot to mention).

Each board right now is 6.50 for the chipset, 2.30 for the coil, about 1-3 for the random resistors/caps, and roughly a couple bucks in pcb manufacturing, she these suckers aren’t cheap unfortunately. I believe it operates at about 2.3KHz? I designed it 2 weeks ago. lol, can’t remember. I attached the scope but I didn’t take any real solid readings, now that it’s all pilled up I’ll not be able to test the current draw or get any screenshots of the WF.

Current draw at tailcap was measuring a little above 4A I had a hell of a time with the first board, first time hand soldering 0201s maybe caused some thermal damage to a resistor or a cap… Oh well, first one works, and that’s the important thing :slight_smile:



Very cool and definitely something I would be interested in.

Definitely interested in one. I really like my XHP50 drop-in in my Solarforce L2 but the runtimes from two 18350s aren’t the best

Nice…0201 is teeny tiny!

Welcome aboard…a boost circuit is very awesome, any idea on the inefficiency of the boost?

Great 1st post man…you will go far here!

Waiting for a sales thread :smiley: :party:

Cool. Very good work!! If you post what you have you may get extra eyes to help with a Attiny13 or 25.

Very very nice!

BTW...welcome to BLF!

So why 3.2v instead of say 2.8v for a cut off voltage?

And is that 3 amps at the emitter aka 6v or at the cell at 3.6 ish? I think these XHP50 are tough so is is possible to get 4-6 amps at the emitter at 6v+?

Wonderful accomplishment! Pic mcu might save some board space but the firmware available isn’t as extensive. How about a stacked mcu board?

interested

Awesome! Keeping an eye on this one :)

Low Voltage cutoff is at 2.7 threshold, sorry about that.

The 4A was measured at the tailcap, I pulled the emitter and was only able to get .84-1A at the emitter, I think I need a high draw battery, or I’ll need to force PWM instead of the PFM, will do more research on it, but that means another board cut and wait time. Ugh.

Received a question in my inbox, wanted to open clarify it. If there are any questions guys, please throw them to me!

MESSAGE RECEIVE
Hi,

Are you sure you’re not direct driving the emitter from the battery? Are you able to measure/confirm ~6V voltage across the emitter?

I don’t mean my questions to sound the way that they might be sounding, but I noticed the current you showed measured in the 2nd to the last pic, and started wondering if maybe something is shorted such that the emitter is being driven directly from the battery.

Good luck!

By the “2nd to the last pic”, I meant the pic with the meter showing the current of 4.19 amps.

MESSAGE RESPONSE

Hey no worries I need skeptics :wink:

I measured 5.9-6.4V across the output of the driver. I had already pilled up the star/driver before I thought to take pics of it. I’ll rip her apart tonight and get a pic of it, under load.

Furthermore, a XHP will not even emit @ 4.2VDC. I did try to directly solder a XHP50 to a fully charged 18650 and got nothing.

I’ll c&p this question, so others can see my answer if you don’t mind!!

Efficiency is above 90% I believe it was calculated @ 97%.

great project and thanks for sharing. I am interested; I understand it is a proof a concept atm but - if at some point in the future - you are making it available commercially or maybe via DIY platforms (i.e. Oshpark) I am willing to support/pitch in.

Once I get it nailed down I think I’ll go commercial, do a few small runs maybe 100 or so. But I wanted to share the latest revision of the board so you guys can get a picture of what I’m working with :smiley: White ring is the 17mm ring that it is routed out at. with a 2oz pour this should be able to fully support the 3A.

Designed it this time to add a solder on spring and copper to the edge. I still am labeing it REVA. lol

You cracked the code.

You guys have me all feeling so flattered. :8) I think I’m going to go hit the lathe and turn myself a new light, I’ll do it in SS. Happy days.

After a little more reading, I am going to have to force PWM. This sucks a little bit, as my board will now need an additional ground trace. Ugh.

Any reason why this great driver shouldn’t work with a MT-G2?