nFlex input

Hello,
I am making a Lux III headlamp with an nFlex board. I plan on using 4 AA cells for input.

What I am wanting to know is, do differences in input voltage affect the brightness also, or just the efficency.

Hi Brian,

First, Welcome to CPF!

The nFlex board is a current regulated circuit. So your output will always have the same current, or LED intensity. Just make sure that the input voltage is 1.2V higher than the device you're driving.

Normally 4AA (>4.8V) will easily drive a LuxIII without any problem, when the batteries are fresh. However when the batteries are low (0.9V/cell) you won't have enough voltage to drive the LuxIII at spec (700-1000mA).

Case I (No Problem)
Fresh cells - Vin > 4.8V
LuxIII J-Vf ~= 3.5V @ 700-1000mA

Case II (Problem!)
Near Dead cells - Vin ~= 3.6V
LuxIII J-Vf ~= 3.5V @ 700-1000mA

In the second case above, would the light be dim or fail to light at all?

I think the light goes out of regulation, into direct drive, and drops in brightness.

Like Former_Mag_User said. /ubbthreads/images/graemlins/smile.gif

i will go do a test for you.. been wanting to play with my nFlex and got a great new toy... bench power supply.

-awr

OK.. done from testing... first have to say.. like the nFlex a lot... the steps of brightness are perfect.. go george.

Well.. the answer is:

'depends on the chemistry of the cells' :-D

first i set the nFlex to 700 mA and ran through some voltage levels:

7.2 = 758
7.0 = 758
6.5 = 756
6.0 = 754

5.5 = 744
5.0 = 737

4.8 = 675
4.7 = 613
4.6 = 555
4.5 = 498
4.0 = 041
3.9 = 000

bodes badly for 4xNiMH which are about 2.9V when dead.. though.. in actuality.. a four pack of NiMH really give up the ghost at about 4.4V.. let me go take one more reading... ok.. at 4.4V.. 448mA.

which means... you will get acceptable performance but constantly dimming performance if you use nicad or alkalines but a four-pack in series will get you significant amounts of light.

that said... if you put 4 LiON 14500s in parallel-series in a 4-pack... you'll have 7.2V down to 6.0V.. i ran a course simulating a 2-series pack of LiONs (like 2 17670s in series).. similar to 2x2 of AAsized LiONs.. and then i can bump it up to the 1000mA setting:

7.2 = 998
7.0 = 997
6.5 = 995
6.0 = 754
5.5 = 978.

if you ran from a pair of LiONS... you can run it full power (1A).. if you used 4x AAs.. they will start at full (750mA) power if you use alkalines.. and pretty close to full power if you use NiMH.. at 'end of life' on the NiMH, even though 448mA sounds kinda low.. that is only a 2dB drop in the current level.. it's still quite a lot of light.. it would be a useful headlamp.. especially things to consider: the VIP for example.. outputs just about 700mA on high-beam but it can't hold it very long because of the dinky battery... the nFlex with 8 levels will let you turn it down and have very long runtimes.

one nice thing... the cutoff is perfectly suited for use with 4-up NiMH.. since right at a perfect cutoff level the nFlex just shuts off.

So.. a 'no brainer' for getting your light going is that yup it will do the trick... if you use either alkalines or NiMH... the alks will start brighter and might run a little longer, but hard to say how fast they'll get dimmer than the NiMH... if you want to get fancy, or maybe make a smaller light.. 2x15400 LiON will run the light much brighter. I just calculated 5.5WH in 2 AA sized LiONs which means 75-90 minutes runtime on high... and a bazillion minutes on low..

this test was done with a TV1J emitter.. you'd absolutely want to use a J bin... if you found an H... you could probably substantially increase the power levels if using 4up AA NiMH or alkaline.

i think fred at photonfanatic still has TYAH emitters.. they are a little on the 'hid purply' tint but they are still nice.

let us know what route you take.. i'm a LiON freak so i'd use LiON... if you don't use it that much i'd recommend using alkaline or lithium primary AAs.. wow.. with 1.5V each.. 6.0 to start.. and they would not drop like alkaline.. that would be a very viable solution... and you won't have to worry about charging.. all depends on how you want to use the light...

once/month for 1/2 hr full bright and 3-6hrs of reasonable light.. primary lithium would get you the most light and they have ludicrous shelf life..

if you use LiON you need to figure out a cutout solution.... so the 'easiest' solution is 4-series AA like you wanted to use... if you use AA lithium primary cells you could get a full 1A out of the emitter on a fresh set of bats (but costly at $11 a refill).. but if you have some willpower the will last a really long time... than you can also use AA alkaline or AA NiMH... which conveniently and coincidentally the driver will cut out just at the perfect time to save the batteries.

ok... neat update:

i have a TX1H star i could try this with:

at 4.4V (NiMH cutout)... 545mA
at 4.8V (NiMH fresh).. 735mA
at 6.0V (alkaline/primary lithium fresh) 754mA

so... based on that... i think you'd be very happy with the results with using 4 series of any of those three batteries.. oops one more thing to check... cutout voltage of the nFlex (so you don't kill NiMH).. holding your hats? it's absolutely perfect.. 4.0V.. 1.0V/cell. You'd get a nice quick but paced drop of output right at the end to know better hurry and find fresh cells.

so.. if you want maximum convenience and the easiest mod.. get an H bin emitter from http://photonfanatic.com and make the light just how you were planning.

if you want maximum power... you need to add two cells if using NiMH/Alk or make it with LiONs

keep us posted.

-awr

Thanks for all the replies. This has all been very helpful.

I have three twoj's that I want to start with. I think when those are used up I will start making lux 5 headlamps.

For the moment I will use 4 cells. But, I plan to develope a 6 cell rechargable pack. I saw some rechargable lithium
AA's on ebay. Has anyone tried these? If they are worth wile, I will make lights with these, running at 7.2v.

Brian

well fella's, the headlamp is operational. The glue and the arctic silver have to dry. Tomarrow, I will test it underground. Thanks for the advise.

I will post pics when I have them (and I figure out how /ubbthreads/images/graemlins/smile.gif).

Brian.

Waiting for the pictures Brian... /ubbthreads/images/graemlins/popcorn.gif

LiON 2-cell will be awesome but you have to figure out how to cut out the power (unless the battery is protected and shuts itself off anyhow)... LiON AA is a 14500 cell... AW cells them.. i think that he's just around the corner from having a protected version.

-awr

...perhaps you should told him yesterday BEFORE he went miles underground in caves... hehehe /ubbthreads/images/graemlins/wink.gif Just kidding!

Well, she survived underground just fine. My light however, wasn't much brighter than a yukon HL(overdriven lux I with a reflector). I'm using a nxo5 optic. I think I need to switch to a reflector also.

Can you give me a link to AW's store for LiON AA's?

Brian

[ QUOTE ]
Icarus said:
Waiting for the pictures Brian... /ubbthreads/images/graemlins/popcorn.gif

[/ QUOTE ]


Lets see how this works. Pics
I know it doesn't look great. The body is made of layered teflon peices and is bolted on to an old cpu heatsink. But fear not, I'm tearing that thing apart. I've come into some nice peices of nylon. Now I can make the body from a solid chunk and hopfully improve the looks. Oh yeah, its going to be a 2 lux III headlamp now. 8 AA cells to start with. Later, maybe 4 cr123's.

Brian /ubbthreads/images/graemlins/buttrock.gif

I know the beam shot is cheesy. I'll have to work on that also. /ubbthreads/images/graemlins/icon23.gif

Brian

adding,
dont forget that there is extra stabalizer things in a N-Flex the beads and the shotkey at the input. although there is very LITTLE drop there, if you are really tight you could dump these 2 items out of the curcuit.

they are most needed for stuff like automotive voltage imputs, and less needed for the perfect dc of the batteries.

you also improve the efficiency slightly, stop some of the drop and it will reduce your Vin verses Vout voltage thing.

never done it, cant describe it right /ubbthreads/images/graemlins/smile.gif but this belongs with the above excelelent stuff from andreww.

also what andrew said about 2xli-ion, you would really want to have cell protection, or pack protection for the cells. it works great, but because of the cutoff point you can deep discharge one of the cells.

what vidpro said about the ferrite beads and diode on the input.. the beads are for removing AC feedback from the switching ckt (really messes with radios.. my first mod would take out a car stereo from about 8feet.. now it has a grounding shroud around it and it has to be within about 2' to hear anything).. the diode is to protect against reverse powering and melting the ckt.

the happenstance of a Jbin or H bin and 4-up of rechargeables shutting down right at the safe voltage is a really interesting thing... but with 6... that's way too low.. hope you are using Alklines... if not... you could at least make a a low-volt warning with a simple solution of an opamp... a red LED ... 3 resistors and a pot... you put two identical resistors from the LED+ to ground.. that'll establish a pretty close to flat comparison voltage.. high resistance like 100kohm... from the center you go to the non-inverting (+) input no the opamp... that will set up a base voltage of about 1/2 the LED voltage (if you have two series emitters.. tap off the middle between them)... you'll get a base voltage of about 1.85V.. you then put a pot (again pretty high resistance.. could be a meg)... from Vbat to ground... the wiper goes to the inverting input (-)... you set the pot so that when Vbat is at 1V per cell (maybe 1.1)... the wiper will be at 1.85V (or whatever the LED/2 is).. and then you put a resistor and LED on the opamp output... the LED will light up when the battery is lower than you set it to do so...

this can easily be configured to drive the gate of a FET and cut the power to the whole ckt... when i wire up one of these ckts.. i also put a high resistor like 500kohm from the opamp out to the non-inverting input.. this will keep the thing from cycling... but it's not needed if it doesn't shut the light off.

ideally you'd use a voltage reference vs the LED.. but even when you factor in the voltage swing of the LED at different drive levels you can get a workable low battery indicator that will help you from wrecking your batteries.

-awr

/ubbthreads/images/graemlins/thinking.gif /ubbthreads/images/graemlins/icon3.gif /ubbthreads/images/graemlins/duh2.gif