Need a driver

[Fallingwater]

I used to EDC this flashlight. It was good, but the lack of a low mode quickly became a nuisance, so I replaced it with my current EDC, the NDI.

I still keep the MXDL as my backup EDC in case I lose/break the NDI, but if I had to actually use it the single-mode-only operation would again become a problem.

So I decided to mod it.

I unscrewed the pill:



Then I levered out the circuit board:



And this is the net result (RCR123 cell for scale, and yes, I know it's not really 800mah):



I don't know why that hole is there, it seems all it does is impair heat transfer in the place where it's needed the most. Oh well.

What I need is a multimode driver that will fit in that pill (no, I don't want to upgrade the emitter - I'm happy with the P4).
The light is a twistie, so hyper-multifunctional 20-mode drivers are out of the question. Ideally I'd want a two-mode driver, but since I can't find any I'll settle with a 5-mode without memory, as long as the first two modes are high and low.

I've had a look at the offers on DX, and I've come out with this one and this one.

I think the second driver is more efficient (all that extra circuitry has to count for something), but I'm not sure if it'd fit in the MXDL's small pill. Also the first one seems to deliver higher output from a single LiIon cell.

What do you think?

Oh, and keep in mind that I'm in Italy, so I can't buy from US dealers such as the Sandwich Shoppe as US international shipping (assuming it's even available) tends to be ludicrously expensive.
I need something from either DX or KaiDomain.

 

[VidPro]

depends on what you want, and what you are going to drive it with.
(and if it will fit of course)

what cell type will you use? a primary a li-fe a li-ion ? meaning what is your target total battery voltage?

what is the exact inner diameter of the hole your trying to shove the driver board into?

 

[Fallingwater]

LiIon or LiFe. The original board works with both, but is noticeably brighter on LiIon. If I could get a regulating board that gives about the same (high) brightness for both it'd be good, but it's not essential.
I tried to stick a 123 in the pill and it goes in with a tiny amount of space left, so it should be 17 mm.

 

[TorchBoy]

I've had a look at the offers on DX, and I've come out with this one and this one.

I think the second driver is more efficient (all that extra circuitry has to count for something), but I'm not sure if it'd fit in the MXDL's small pill. Also the first one seems to deliver higher output from a single LiIon cell.

What do you think?

Have you already got them? I see that second one is backordered (don't hold your breath) while the specs for the first one says it's for 1xAA. Apparently that should be changed, but how does it work?

 

[FRANKVZ]

I just ordered these: https://www.kaidomain.com/WEBUI/ProductDetail.aspx?TranID=3256 3 modes 10%/40%/100% output

 

[VidPro]

when working with li-ion voltages neither of them will work.
only thing i can find that will work is the wizzard GD500 or something like that, and it doesnt have PWM levels control stuff :-(

the low voltage ones, will go into direct drive with more than ~3.8V, the higher voltage ones fall out of regulation below ~3.8 and go into direct drive.
sooo (that is why i asked what you wanted)
do you want a nice overdrive for a while, and the problems that can have, with full constant regulation after that, and NO ability to use unprotected cells. without protected cells these type of drivers will discharge even a li-fe to far down, and you wont even see it happeneing.
or
do you want it to never overdrive, and dwindle after the voltage droops, which will keep everything in line with specs, last a long time, and have the advantages of direct drive, when it gets low , it will slow down and run longer. these can more likly work with unprotected cells, because when the voltage goes to far down they stop operating, and they are nicer on the battery when the battery is at its end of capacity.

without a boost&buck (both) driver most of the drivers being sold dont "match well" with the li-ion voltages.

on the Li-FE , well heck that is just perfect for the lower voltage ones, and would be DD with the higher voltage ones almost right from the start.

so for the primary cell (3.3Vmax) , and the lower voltage rechargables (3.8Vmax) go with the "buck" type of drivers , ones that will do 0.9V-3.6V , which are LISTED as 1.5-4.??, but really do NOT work proper with the 4V+.

as far as efficency with more parts i think efficency with more parts can best come from a DC-DC type of curcuit , because it will adapt to the li-ion voltages without being to high and to low 1/2 the time.

that is all i know, and it might not be fully correct, i only know it because i am still searching for the optimal LI-ION driver, and they all have some caveat Today.

 

[VidPro]

the 7135 type of drivers is a fairly good choice for the li-ion and the li-fe, but NOT for the primary.
(it wants there to be more battery voltage than led Voltage, like a buck)

its what they call a "linear regulator" it just chops off the voltage that is to high for the led (Vf) and tosses it out as heat.
this will maintain the current to your usual White led till the battery goes below about 3.6-3.8 (or the Vf of the led at that current).

the "glorified resister" is a good choice because it will stop the overcurrent ONLY untill the voltage drops to low on the battery, then it will fall out of regulation. once it falls below the voltage of the led, the light will basically be in direct drive.
this is a good way to preserve the led (current control of some sort) and its good for the battery, because it will not suck the battery dry. potentially in a single cell light even a unprotected cell could be used (very carefully)
BUT
it will not maintain constant regulation throughout the batterys possible usable voltages. (and that can be a good thing, cause DD is not so bad as it sounds)

so for a clipping curcuit, if your intended battery is Very close to, but ABOVE the Vf of the led it will work great.

so it is great for the single rechargable li-fe li-po or li-ion (>3.6v), but bad for 2 cell (>6V) or primary (<3.3V).

 

[eebowler]

its what they call a "linear regulator" it just chops off the voltage that is to high for the led (Vf) and tosses it out as heat.

I'm a little confused... I've read this statement in it's various forms before and it gives me the impression that no matter the level used, (high, med, low,) power consumption (and current draw from the battery) should be the same since energy not used by the LED is 'wasted as heat'.

Practical experience contradicts this concept (misconception?) because on low, current draw is 80mA, med: 270mA and high: 970mA. Isn't this a clear indication that power consumption is different for the three levels? (I'm using the 16 level, 3 mode driver with 3 NiMH cells.)

Maybe I need some sleep...

Thanks for any clarifaction.

 

[VidPro]

2 "parts" of the curcuit the regulator thing that controls the total max current, and the PWM levels adjustor thing which switches the led on and off. (both items can be combined as a single control item)

with the levels, which is done with a microcontroller type of chip that "pulses" the mosfets (switches) on and off (completly) , that is PWM (pulse width modulation).
with PWM the current you read with your usual DMM is an "averaged" current, and it can be quite far off too. if you switch your DMM to AC you will see it displaying AC current too, because of the chopped DC.

so with the PWM type Levels control, the LEd is pulsed on and off , and that is what makes it look dimmer to our eyes which average it out, and why we dont really know any facts with the DMM either.

Ahh pictures
.......................................... and averaged its _____
------------------------------ and averaged its ____________
__________________________ and averaged its __________________________

^ thusly, the led is on at the same current, but just not all of the TIME.
because its not on all the time, it indeed uses less Total current (per hour).


you gotta have a scope thing to read the actual stuff on the PWM.

there DOES exist current control in lights and drivers, but its par for them to use PWM.

Averaged pulses to the led over the voltage for the led still slowly damage the led, but the HEAT is removed between pulses , so it helps the heat situation, but an overdrive still hits the little molecular gate thingees.

umm, ask newbie he knows this stuff for real.

 

[eebowler]

Funny you should mention Newbie, I was just thinking about him.

I didn't think about PWM mostly because the light does not flicker on the fan (yes, that's how I check for PWM) like other lights. Even on low there's no change I can detect.

Thanks for the info :thumbsup:

 

[VidPro]

i dont build these things but a simple capacitor added to a PWM curcuit would totally average out the pulses things, and become actual current control.
the only caveat would be capacitors dont charge and discharge forever, they have a failure possibility . then there are much better capacitors than the 4cent ones they would put in these .

if the curcuit will still operate correctally, a good small cap in there would "fix" the pwm and make it great , and some of the drivers have them already or use them for making thier dc. potentially a modder could just toss one in if they have the room, and "improve" the design, and the way the led is driven. it would even improve the output of the led during pulsing.