is it even feasible to maybe...

i figured i would run this by the guys here that are into it? to see if its even remotely feasible…

i am always looking at the programmable driver stuff… somthing like an evenly spaced “6 or 8 mode” with no blinkies would be cool all by itself…

heres what i wonder…

on my other forum, we do night vision… naturally, we have builds on the infrared illuminators we use. (think zoomie with an oslon emitter, lol…)

on the datasheets fr the oslon we use, the datasheet describes creating a pulse train, to drive the LED, to achieve more light. They take great pains to make sur you understand that your not getting ay more wattage out of th unit in this manner, but rather, you can use a higher amperage for the tops of the pulse train.

with the caveat that while your BRIEF BURST of light will be brighter? that there will be an OFF period, all constructed such that the total wattage is the same.

in my mind? i am relating this to similar of a “flash” for a camera… its just flashing mroe than once.

==

we use PWM for LOWER power modes… i was wondering if anyone had tried this with visible, such that it might be viabl for my IR illuminators…

our ight vision s almost all “digital”, and the CAM sensor is running at 30 frames per second, although maybe the way it works, it might be twice that.

=

i know this exact same thing, USED to b common with “regular LED” infrared illuminators? the kind that have a “ring” of little IR LEDS aroun the camera in the body housing… and theres a “lucid science” project to do this with a “LED ring”…

(the project demonstrates before and after pictures, whereby you can “see” the same camera in the same dark room, is noticably brighter after the MOD)

==

has anyone ever played with this strategy? (even visible…)

and…

ayone think the programmable microcontrollers might be capable of creating such a pulse train, that would accomplish this?

i had previously thought this might be a little more costly of an endeavor? back in the day, i was used to “proto boards” and 100 or 300 dollar “development kits” to program microprocessors…

…now you guys talking about 20 dollar programming of the little atmel 8 bits micro’s… has me thinking…

Seems easier to do with IR run from li-ion then white leds due to the higher voltage at high amps. Even at 5A a black oslon is suppose to have a vf as low as 3.4v.

By the way, do you guys over there have a source for high flux bin black oslons?

I'm not sure what your asking really, its sort'a hard to follow but it sounds like your talking about PWM. Thats EXACTLY how it works, there are constant current non-PWM drivers out there but the way the ATTiny13a works is by providing a PWM signal, the actual switching is done either in current regulating chips (7135's) or in a MOSFET which switch power on and off thousands of times a second and the "level" is controlled by pulse width (thats on time vs. off time of the pulsed signal).

IR leds can be run at much higher currents if pulsed then if not. Black oslon can only take 1A normally but pulsed you can push 5A. Results in more light. Its allowed & specified by the datasheet so its not even over driving the led. It can also be done with visible LEDs. Thats what the OP is referring to.

lucidscience - pulsed IR illuminator

i have been “there” for close to a year now?

i NOTICED the “binning” of th oslon black in the datasheet, and remembered the “binning” of the visible LEDs i had read about on th web (likely here and CP, lol)

i was going to ASK about that over there? then i didnt… later on, it “came up” on the forum… guys were complaining that even good suppliers like “mouser”… the person on the phone acted like there was no binning of the oslon black, they didnt know what ayone was talkg about… that there was only one oslon black… blah blah…

which made me HIGHLY SUSPICIOUS… out of 2 seemingly identical p60 emitter i ordered off of a member there? one appears to be slightly “better” than the other,

one or two guys? loudly proclaim “MY supplier knows to provide ME with the very best bin numbers!” which again, to me, sounds “suspicious” and a little bit like “ad copy”… i dont necesarily accuse a member of my other site of lying to get a few more sales? but more accusing the idea that the VENDOR simply said “oh, YEAH… we ONLY get the highest BIN, fer SURE, dude… your good !”

every couple months, i ASK on my site… if anyone has “solved the mystery” and located a vendor that knows, and guarantees… the bin numbers?? (I ever get an answer… which speaks volumes to me… either they wont disclose their source? or they dont know… or they are all just relying on their vedor telling them the truth…)

to ME…. this is utterly RDICULOUS… the entire LED EMITTER industry? across al th companies? has BIN NUMBERS… how in the ^%# the oslon black has bin umbers, and NO ONE knows ANYthing about them?

i find to be a load of crap…

===

i dont trust any vendor, as faar as i can throw them… not to me maximizing profits… by buyig the low bin numbers in VOLUME… and sellign them top dollar.

call me paranid? but i dont believe in “trust” with businessesmen, LMAO…

=

the LUCID SCIENCE project? basically took a fairly standard approach to the circuit… they used discrete components to “construct” the pulse train they wanted, and “tuned” the core components to be verified with a Oscilloscope… then the created pulse train was applied to a transistor or FET… so the output had the current to run the multtude of little LEDs…

=

the whole project, whether accomplished at close range with regular tiny LEDs… or with on bigger LD emitter… is predicated on the basic idea, that the “rise time” of the LED or the emitter, is very quick… nanoseconds… its pretty “instant”…

the video cam sensor, is basically taking 30 pictures a second, to think of it like a regular digital camera taking pictures… you “flash” fast enough Hz speed? you get bright flashes more often than any ne picture gets taken… allowing that as the amperage of the SINGLE FLASH goes up in amperage over constant run amperage ratings? the “down time” between pulses gets longer and longer, to preserve the overall dissipation values.

basically put, as the PEAKS of the pulse train go “up” in power? the valleys BETWEEN the peaks f the pulse train have to get longer and longer…

…i just wonder where the break even point is, naturally… lol

==

this all goes on a GUN though, to be lugged aroun the countryside at NIGHT for us? so… it needs to be very small, very light, and kind of bulletproof (pun not intended)

we already HAVE enough “crap” adding up in size here and there in our builds… if i can accomplish this all from within a re-programmed driver? it would be a godsend…

What forum?

http://www.nightvisionforumuk.com/

best i found for this…

we make GEN 2 and GEN 3 “performance” night vision? (hand held spotters OR gun mounted… emphasis is on gun mounted…) digitally… for the PRICE of a generic GEN I unit…

obviously, this is mainly a DIY/project related site….

Haven’t tried pulsing for higher lumens myself.
Actually I forgot it can apply to visible LEDs and possibly even work better with visible LEDs then IR.

Made an account at nightvisionforumuk so I can look around. Waiting “to be approved by an administrator”.

okay…

downoaded the datasheet…

I have been thinking abotu this, since i first heard the term PWM used around flashlight drivers, lol… we talked about it on the NV forum? but… theres only several “tech dudes” there with me, at least that talk openly about tech stuff.

so, we have TALKED abotu it, but… no ones DONE it yet.

looking at the datasheet… theres room for SOME improvement in my IR emitter, th question is how much… this isnt a “single shot” opertion a-la taking a picture… to get anything “meaningful” in the way of higher currents? those duty cycles get short, and, the actual time spent actually ON gets really short…

all right…. i suppose i could just describe what i WANT, and see if one of the “custom UI” guys here would do t… seems pretty do-able looking at the datasheet on the atmel, maybe…

i am taking 30 frames (pictures) per second… and i think thats actually 60 per second, maybe, cos of the way it delivers a picture to the screen…i suppose i want at least TWO flashes per picture taken… so if i assume 60 Hz picture-taking… thats 120 hz flash rate. lets say 200 Hz to get more flashes per frame and make the numbers rounder….

on the chart on th oslon dtasheet?

T is the Hz of the entire operation…. which here, as a minimum (more is better though) but the minimum is 200 Hz… but to even be meaningful? its looking like at least a 10% duty cycle (D=.1 on chart)

1/100 is .01 that makes 1/200 =.005 seconds for the entire cycle length… but, at 10% duty cycle? thats only ON for 1/10th of THAT tim period… and 1/10th of .005 seconds, is .0005 seconds ON, and .0045 seconds OFF

and on the chart? time spent ON, allows for too little “gain”…

===

okay, so i attack this, form the other direction… find my amp spot on the chart, and see the time on and duty cycle associated with that spot… and construct my PWM from THAT…

10 to the –3rd power… thats .001 thats 1/1000 of a second ON… and being conservative on the chart? 2.5A is a little under what it can do (safety margin and all), at 10% duty cycle.

thats 1/1000 second ON… followed by 9 times as long OFF, or 9/1000 second OFF…. time on plus time off = one period (hz)…

that yields what? 1/1000s + 9/1000s = 10/1000s which simplifies to 1/100 seconds period length for one cycle

thats only 100Hz…

=

of course… no reason to do something “half @$$ED” as in only usng HALF the amperage of flashes as the chart deems possible…

if i am reading the chart right the drop down FROM 5A, at 10% duty cycle? occurs around 10 to the –4 power… which is…

1/10,000s ON + 9/10,000s OFF = 10/10,000s period length… which reduces to 1/1000second for one cycle.

===

my intuition tells me (which could be WAYYYY off, but still) that if a camera is taking at most 60 pictures per second, and there is a FLASH occuring 1000 tes per second,it should be fairly impossible to get a “dark picture” otu of it… if i am doing the math right, 1000 flashes per second, over top of a 60 per second snapshot takig… there is a little over 16 flashes “per picture”…. which should “saturate” the image, i should think.

i vaguely wonder, how much more “power” i get out of a 5A flash, in a 1A device…. as you go UP above constant current rating, you lose efficiency… but still…

this is seeming like “safe” overdriving, sort of…

===

so… if smeone said my calculations were correct according to the graph… i would be asking a builder here to program a 4 or 5 amp driver… with a 1kz pulse train out the controller… with a 10% duty cycle ? eh?

sedders,

You are definitely on to something here. A while back I built a IR headlamp for Rolaids. The headlamp was built exactly the same way I build them with visible led’s. I didn’t take into account that the headlamp had PWM in the lower modes only. The high mode was direct drive so no PWM. Well the frequency was such that it created pulsing when viewed through the ccd camera.

I had a video of the effects of the headlamp but have misplaced it. I have contacted Roland, to see if he still has it, if so I will post it up. I will also try to get a scope and measure the PWM. I do know that the max current to the LED was 1.5A. Which is what i have found to be about max overdriving current(constant not pulsed) for the oslon black 4715s, but only when mounted on the higher quality Berquist stars and mounted to a decent heatsink. The thing I don’t know is how much increase in radiant intensity was achieved by overdriving. And only seat of the pants engineering. I compare the oslon 4715s to the cree xpe2, which I have done extensive testing. I am simply applying what I learned with the xpe2 to the oslon. I am merely geussing they behave the same to overdriving. I do not have a meter to measure light in the IR wavelengths…yet. My only way of measuring is by perceived brightness when viewing through the ccd and viewing screen, which is highly subjective. The major drawback to overdriving the oslon is the thermal slug and anode are electrically connected. An isolated anode would allow the Oslon to be mounted on a Noctigon or SinkPad and allow larger amounts of overdriving. Your pulsed /duty cycle idea may allow the led to run cooler at higher average current.
The other consideration with overdriving the Oslon is wavelength shift due to increased temperature. I didn’t see anything on the datasheet about that, but i did find this

http://www.ledengin.com/files/products/LZ1/LZ1-00R400.pdf

look at page 10 and 11 i would expect the oslon to behave similiarly.

I totally agree with the binning of oslon. Mouser sells specific bin visible emitters but do not specify what you’re getting with the oslon…don’t know why???

I will be watching what you do with this, it will be very interesting to see how the PWM spec’d by osram affects the ccd/viewer.

Just some thoughts, sorry i couldn’t add more specifics just generalities.

send some PM’s to the programming guys to see if they can hook you up. Stacking chips to 4A shouldn’t be a problem.

i have several of a “budget zoomie” that i put different LED emitters in (visible) then finally found my one old oslon, i thouht had died… it was just the driver i cooked, so, i tossed it in there…

in visible, i never noticed any flashing effect… but, when viewed on the bckup screen hooked up to the CAM… theres a slight “flashy” effect on the LO setting. None on HI, naturally.

i dont like it, but, its not something so BAD it drives you crazy… you can just notice it on the LO… sort of reminds me of the “old security monitors” back in the day, the way some of them had a slight flicker you barely noticed.

havent had my o-scope out in ages, dont even know whee my probes ARE, let alone if i find the probes, which one is the “good one”, ha ha

mind you, this is the “craptastic” driver that came with the budget light, lol… it does “all” of almost 500ma to the oslon on HI, with the proscribed 3-D batteries… i run it on 4-c batteries to get the juice up to a whopping 700-ish milliamps oh HI, hee hee

i guess if i could nail the frequency of the cheap light on LO, i would know i needed highgher frequency PWM for this, (duh!)

maybe if i look at the lucid science project, i can find the fequency he used for the little round LEDs in that project…

Not surprising, many of the cheaper lights have bad pwm. I use the nanj 101-AK-A1(set to three modes) in all my t20’s with oslons and no problem with pwm affecting the ccd camera / viewing screen in any mode. Also no problem with dr. Jones drivers either.

Roland just texted me and we don’t have that video anymore, sorry.