A 7135 will work, all 5mm leds that I checked can handle 40mA fine. So using an 7135 could be a solution to drive this light on a li-ion battery.
where do I get these? I`ve looked on eBay several times (for another project) and there`s nothing 
Riiight! nobody ever mentioned the `AMC` part at the beginning, it`s always just `7135`, Now it makes sense! LOL
20mm LEDs are usually limited to 20mA or thereabouts. Sure, they’ll work at 30mA or even 40mA… for a while.
I “pushed” a set to 25mA in a N×M array board, and it worked beautifully… ’til a few strings started flickering.
After a while, I had “lightning mode” in that panel long before andouille was even written.
I have modded a bunch of those 5mm 9 led lights using the 3200K LEDs rngwn sells here:
I used the 7125 regulator, was able to fit it right on the LED board…. can’t quite remember the details, but clipping the middle supply pin on the chip, straddle the output and positive pins across the LED feed traces, then isolate the spring from the outer negative rail, add a wire from the spring to the center negative supply/heat sink on the 7135…
Arghh sounds more complicated than it is to implement. I would have to pull one apart and photograph to simplify that jumble of explanation. If you have the parts in hand though, it will be obvious how it could work.
They work great with these wonderful 5mm LEDs, a nice flood light. I use them with 18500 cells as well as 3 AAA.
…… as I recall now, some of these lights did not use a spring for the negative feed, but a small wire that contacted the body across the edge of the LED board. Same basic mod though, the 7135 still fit directly on the board.
so perhaps putting resistors After the 7135 would be better? (I`v got 20 of them on order now).
However you do it - please document it and take pictures.
I ‘direct drive’ my 9 and 12 LED lights since there’s not much space (and so am limited to 3xAAA), but have constant current regulators laying around since a year or so. Just need an idea how to cut up traces and place the chip safely.
Good luck. If it works, I’ll copy it.
The LEDs I have from rngwn have no problem being driven to 40ma. Check the test data in the thread reference above.
If you are not using them…I don’t know, maybe order some. The 2300K are a delight as well for night light, candles, etc. I’ve used over 100 already. A handy item for the toolbox.
…I generally keep the current below 50ma for lights that will be on for hours at a time. Have not had a single failure yet.
Thinking out loud: A carrier with one protected 10440 and the ‘driver’ in one of the other two bays should be possible.
I guess it would depend on the parasitic drain of the driver because it wouldn`t have the benefit of the tail switch, that would be used to to engage the already conditioned output of the carrier module rather that cut power entirely.
I had a further look at the PCB and there`s even less room than I thought, perhaps cutting the track from the Pos spring to the rest of the leds MAY afford me 2x 1206 peices side by side at best.
and to top it all off, the PCB is a little bit smaller than it should be to fit the shelf in the main body of the light so if it`s not falling right through it`s not mking a propper contact, it was compression fitted originally, and it looks like pure luck that it ever worked at all!
and as a test since i`m going to be getting rid of the nasty cold white leds, I thought I`d try see what would happen with a 18350 powering it, well it Does work, but I wouldn`t trust it for more than half a minute at a time, it gets kinda toasty! LOL
A resistor after the 7135 will do not much good, the 7135 will raise the voltage to keep delivering 350mA (that is basically what a current regulator does), and if the voltage overhead is not enough to deliver the 350mA, it will work as direct drive with the resistor as current limiter, so no use putting the 7135 there in the first place.
You’re right. Only OUT would be switched. Not sure what the 7135 will do then.
But the resistor could be parallel to the LEDs (as if it were a 10th LED)?
Ah, I see, never thought about that possibility, the resistor value will be critical (my feeling is that small changes will affect led current a lot) but it sounds good.
basically This is what I have to work with!
Nope, it’d just drop out of regulation sooner.
Seriously, circular piece of thin plastic, a bit of tape, axial resistor. Done.