I’m amazed at what several of the guys on this forum are doing with designing and building all these custom drivers! I’m going to need a driver for one of the projects I’m working on, but I don’t know what size or even shape it will be. But, anyway, this one looks awesome! Great job, wight!
wight - this is awesome btw... A 20 mm AtTiny based buck would be pretty awsome too!
I hear that Alex, unfourtinately I don’t have the option to take a break with how my business has evolved…
Btw I have a new 2-4 cell buck driver based on the MAX16824 chip. Output is selectable to around 8A (for 3-9v emitters/strings) via sense resistor. PIC controlled but it could be easily adapted to AVR (with the addition of a vdivide circuit for batt monitoring and a few cap changes). Very stable operation with the inducer I selected, much easier on 3v emitters than the HX-1175b’s at high currents and with pic Very low component count so cheap to produce. I’m working on a larger one initially, hopefully I can get it to a 20mm single board (not sure tho) or a 17mm double stack <12mm tall.
When you do…tell me how you did it!
I think I might get testosterone shots or something…so dang demotivated…
I need someone to put on a DI hat and yell at me to get my butt moving as of late…sux, so tired, so moody, so sluggish and unmovtivated…ARRRRGH
I also am eagerly awaiting the 20mm buck driver, I have one of those 3x “C” Defiant flashlights, could definitely use a good driver and stack some 26650’s w/ spacer in the badboy 
Will do ;-).
DI hat?
IIRC 26650’s wont fit in a DST, and boring it removes the tail threads. I think the normal solution is to permanently attach the tail after that and load the batteries from the front of the tube?
DI = drill instructor.
Wight I’m searching for your post on calculating voltage divide resistors for multi-cell Zener drivers….
Edit:’4wheelr found it - link below.
So it’s taken me a while to post this, but another member pointed out a mistake. A careful examination shows that the 7135-A markings are clearly wrong in v016. At least one 7135 in both the “A” and “B” groups is required for the driver to function, I had the markings for A mixed up for two components. I have updated a few other things as well.
Changes:
- Updated 7135-A silkscreen to reflect actual “A” positions.
- Updated silkscreen with boring name: “A22-7135”. Added silkscreen for “ZENER” and “OTC”. Added polarity markings for D1.
- Moved my marking from top to bottom.
- Changed Zener to correct (more narrow) SOD-123 footprint rather than an incorrect wide (SMA?) footprint.
- Revised several traces and vias to correct design rule violations!
- Voltage divider now feeds directly from Vbat, please switch to 22k resistor or adjust your LVP values in firmware.
- Added an emergency escape route for those swapping over a Nanjg-105c. There is a scrapable solder jumper to allow the use of a Nanjg-105c diode and 19.1k resistor.
- Fixed all the mistakes I made while doing the above stuff.
Very nice i just ordered a few boards to try out my new hot air re-flow station
And now I see for some reason I never marked that the 7135 next to R2 must be populated for LVP to work? Ugh! Well now we know. :~
Thanks nitro!
Wouldn’t you be able to squeeze the off time cap in to the left under LED+ with a little redesign? Flashing might be difficult with the off time cap in the way?
I measured in Eagle, the off time cap should not block the clip. I have >8mm of vertical clearance overall for the clip. The top left 7135 and the Zener are both a little closer than I would like, but I’m pretty sure a stock Pomona clip will fit. A shaved Pomona clip should fit for sure (stock they are very wide compared to the actual chip, 1.4mm wider IIRC).
> moody, so sluggish and unmovtivated…
Winter blues?
I’ve found cet.org has a good self-evaluation for timing bright lights; and a $25 dawn simulator is a great help in our household
But back on this board, is CK or anyone else building them?
I melted the dome off a Nichia-on-Noctigon (my first mod of an F13) after less than 20 minutes’ total use, and I’ve been wondering whether it was just cooked and a driver swap would help before I replace that.
I’ve been thinking of doing a few to sell.
I dont have the newest version (wight’s pretty quick at constantly pumping out new revisions of all his stuff, always tinkering) but I have some older one’s. If you have an immediate need it may be faster to get some from him (if he’s got the boards on hand) but you’ve always been a good customer of mine sir, if you need one I can order the new boards today (I think it’s always best to use the newest version’s available so I’d rather do that than sell you an older one) just let me know if you want to go that route and I’ll get them ordered now.
Good to know. I’ve been way to conservative with mine.
Congrats wight, that is a very nice upgrade to this already great driver.
Could you elaborate on that one?
If a person needs or wants to use the ‘traditional’ setup with 19.1k/4.7k + a diode, I made provisions for that. They can cut the trace which goes from R1 to B+, then scrape the jumper next to R1 and solder over it. If properly scraped surface tension should allow for an easy connection with no need for a wire or anything.
No hurry on my part because — as I think I melted the Nichia dome off the Noctigon just in intermittent use with the stock driver — I’m not quite sure what combination of emitter and driver will make sense.
(I got two F13s, put a Nichia in one for better color, while it lasted)
I know I want to replace the drivers to get rid of the blinkies — but I’ll wait til someone else has more thoughts on what combination of emitter and driver will work.
(I’m more interested in length of use than brightness, these will go into the earthquake kit eventually)
The really big plus here is a driver that will work with the retaining ring.
EDIT — meanwhile, I can’t stand to have an unusable light, so I stuck in a 3W Osram
http://www.fasttech.com/products/0/10001895/1136403-osram-3w-4700-5000k-150-160lm-neutral-white-led-em
and one of these inappropriate but seems to work tolerably well spacers
http://www.fasttech.com/products/0/10007066/1576101-xm-l-reflector-to-xp-g-reflector-converter
(go ahead and laugh)
and the F13 is again a usable light, nowhere near as bright as the Nichia I melted or the stock blue-white emitter, but decent color rendering.
But yeah, I need to understand what pair —emitter and driver — will do better. I liked the high CRI til I melted the Nichia.
I’d say an XM-L2 at ~4A is decent. As I recall I swapped the stock centering ring for a normal one (with wings). This yielded a pretty large hotspot.