2.75 vf led

[jasonsmaglites]

im building a 2d direct drive light for emergency usage, and need the lowest vf i can find. i've build one with a h bin (3.00-3.25v) but i was wondering if we've ever seen a g bin (2.75-3.00v), and where i could get one. even 2.5v would be great but as i understand it they hover around 3.6 as a rule.

 

[HarryN]

You are in luck - use a red/ orange Lux III. They can run 2 D alkaline direct drive, have Vf in the range you are after, can accept quite a range of current up to 1.5 amps, and are remarkably bright.

One other advantage - the red / orange color will preserve a lot of your night vision when you turn it off.

Heat sink it well, as this color in particular is more prone to loss of light output with temperature compared to white leds.

 

[Linger]

Photon Fanatic sells a K2 Tffc TVOD, iirc vf 3.0v
I've test run it series with a cree r2 in 2AA and 4AA configurations, and the K2 is brighter in both (arguable due to the lower bin).

 

[HarryN]

Just as a comment, CR123s have a 10 year shelf life and I am pretty sure more capacity than an alkaline D cell.

Perhaps consider just making a direct drive CR123 and a K2.

If you want to use a 2 D mag light for the host, just fill the rest with a plastic pipe full of other useful things, and run a bypass wire around the plastic pipe / storage box. A quality, primary CR123 will give you close to an hour of run time, plus spares that you can keep in the plastic pipe.

The more I think about it, the more I think I will do that with my wife's 2D light, even if I keep it as an incan.

 

[tebore]

Just as a comment, CR123s have a 10 year shelf life and I am pretty sure more capacity than an alkaline D cell.

Perhaps consider just making a direct drive CR123 and a K2.

If you want to use a 2 D mag light for the host, just fill the rest with a plastic pipe full of other useful things, and run a bypass wire around the plastic pipe / storage box. A quality, primary CR123 will give you close to an hour of run time, plus spares that you can keep in the plastic pipe.

The more I think about it, the more I think I will do that with my wife's 2D light, even if I keep it as an incan.

I'm fairly sure a D cell has over 2x the stored energy than a CR123.

CR123 Avg 1500ma * 3.2v = 4800wh
D Cell avg 10000ma * 1.5 = 15000wh

 

[HarryN]

Hi Tebore, you might be right, esp. for the NiMH cells, but I guess I was thinking that he was looking for an "emergency light", meaning - it can sit in a corner for a very long time with no interaction. Pretty much, this means primary cells, at least to me.

I looked at the Duracell web site to see if my mind was working right or not: (where else would a flashlight guy look )

For the "fresh" D cells:

http://www.duracell.com/oem/Pdf/new/MN1300_US_CT.pdf

If I assume 2 alkaline D cells in series, pulling 1 amp until it is down to only 1 volt each (so the red / orange LED barely coming on), then this would run for about 4 - 5 hours - more than I thought actually. :shrug:

So in a similar way, if I were to put 2 each CR123s in parallel, and run them with a 1 amp (total) discharge down to 2 volts, then this data sheet would say it will run just under 6 hours. Running just 1 cell by itself would only last about 1 hour. Not exactly linear is it.

http://www.duracell.com/oem/Pdf/new/Li23A_US_OS.pdf

Actually, this came out a LOT closer than I expected. Still, the cold weather performance of the CR123 would get my vote. In the winter, you can kiss that alkaline battery goodbye, but the CR123's will still fire up.

 

[LEDAdd1ct]

I was going to post the above, too; yep, a D cell has *way* more capacity.

 

[lolzertank]

I have a C (same as old G bin) Vf bin neutral Rebel 0100. So they do exist, but they're not that common. IIRC, Newbie had a G bin Luxeon I a couple years back. He mentioned it while showing how to make a boost converter more efficient. Oddly, my red orange Rebels are D Vf bin.

 

[jasonsmaglites]

i'd really like a neutral white or warm white led. orange/red just doesn't fit what im tryin to do. its not for nuclear emergency, just a power outage. d cells lyin around will suit me just fine. i meant i want a light i can throw in 2 cells, and not worry they're fully charged, and it will still run for 24 hrs. the 100ma pull of a 2d direct driven light is perfect for that. and nothing beats the bang for the buck of the d cell battery imo.

linger, tell me more about the k2. is that pushing 100 lumens per watt like an efficent q5 bin cree or u2 bin ssc p4. how is the color. as i understand, they're warm leds and can handle alot of power. am i right?

 

[Linger]

re: warm - I understand, same with me. The TVOD emiter from Photon Fanatic isn't a warm but you will find it is a very neutral tint. Khaki is my subjective opinion of it's colour (compaired vs. mc-e WD, mc-e K? R2WD, R2WC, Q35A, Q45C, Q24B)
iirc it's under 100lumen/watt, but again i've had it in series with an R2 WC and it is brighter (contra usual notion of cool tints as more efficient). (3.0v-9.0v 800ma driver from Shiningbeam, 4AA eneloops: single solder point connecting + lead from driver with + lead from each emiter, second solder point connecting - lead from driver with - lead from each emiter)

Hopefully Justin Case chimes in because he has an excellent understanding of circuit / emiter efficiency.
Yes, the K2 can reportedly handle huge current, lots of people report direct driving K2 Tffc's off 18650 or 16340's. I have not had the courage to run this for any length of time. Heat sink it properly.

Anyway, I'll set this up for you right now: K2 tffc TVOD, 2D, alkaline direct drive. I'll update tonight (unless SWMBO interupts...)

Linger

 

[HarryN]

re: warm - I understand, same with me. The TVOD emiter from Photon Fanatic isn't a warm but you will find it is a very neutral tint. Khaki is my subjective opinion of it's colour (compaired vs. mc-e WD, mc-e K? R2WD, R2WC, Q35A, Q45C, Q24B)
iirc it's under 100lumen/watt, but again i've had it in series with an R2 WC and it is brighter (contra usual notion of cool tints as more efficient). (3.0v-9.0v 800ma driver from Shiningbeam, 4AA eneloops: single solder point connecting + lead from driver with + lead from each emiter, second solder point connecting - lead from driver with - lead from each emiter)

Hopefully Justin Case chimes in because he has an excellent understanding of circuit / emiter efficiency.
Yes, the K2 can reportedly handle huge current, lots of people report direct driving K2 Tffc's off 18650 or 16340's. I have not had the courage to run this for any length of time. Heat sink it properly.

Anyway, I'll set this up for you right now: K2 tffc TVOD, 2D, alkaline direct drive. I'll update tonight (unless SWMBO interupts...)

Linger

The Vf of a K2 is actually pretty low. Keep in mind, the Lux I and III were Vf binned at 350ma, while the K2 is binned I think at 1,000 ma, so the Vf will naturally test higher for the same LED.

If you want just quick, cheap, D cell, no driver, then I would use a 3D light and just make sure there is at least 1 - 2 ohms in the circuit, which is not that hard in those lights.

 

[Linger]

Meant to post this last night, forgot.

I took a light apart to free my K2 and added some alkaline D cells.
The k2 tffc direct drive (leads touching battery ends) from 2 alkaline D batteries is very effective with full cells. (3.07v and dropping under load, apparently I don't know how to use my dmm for amp). Excellent colour and very bright. I dont' have a light box but markedly brighter then Zebralight H501 on high (97lumens) and brighter than turbo Quark AA on eneloop. I'll say ~130 lumens.
I didn't run this down, but switching to old and best before 2009 alkaline cells, the emitter lights no trouble but it is dim (2.6x v underload). However, I added a Luxdrive Micro-puck (500ma boost driver, $9 from led direct) and the K2 put out usable amount of light from the same depleted cells.

 

[Linger]

The Vf of a K2 is actually pretty low. Keep in mind, the Lux I and III were Vf binned at 350ma, while the K2 is binned I think at 1,000 ma, so the Vf will naturally test higher for the same LED.
QUOTE]

HarryN, I've re-read this twice and I'm not sure which way to read it. forgetting about lux I or III as I have little experience with them, are you saying that the K2 will functionally have a lower vf at low ma and thus be more efficient at the low end?
In other words, how would an SCC P4, a Cree R2, and a K2 tffc, if each were sitting in front of you with a 'rated' vf of 3.0v-3.25v compare?

 

[HarryN]

Hi Linger - Please bear with my long winded answer.

The trick with the Vf is that it is not a "single" number, but instead, it is a curve.

For any brand, or even individual LED, you can make a plot of current on the X axis, and Vf on the Y axis, which will virtually always be an upward sloping curve of increasing Vf with current. This is also true of conventional diodes, not just light emitting diodes.

In order to make it possible to catalog diodes (any kind), manufacturers spec the Vf measurement at some particular current that they choose to target the product for. In conventional diodes, I am pretty sure that the manufacturers sometimes just give the same part different part numbers and rate it for different currents and Vf, to hit different targeted uses.

Power LED makers, such as Cree and Lumileds tend to spec the Vf in the general range of where they expect the customer will use the product as well. In the past, VF was very difficult to control, which is the main reason LEDs are marketed as needing current control drivers. At that time (less than 2 years ago), the construction of the Cree LED on a SiC substrate caused it to have a somewhat higher Vf than the Lumileds LED. Cree made up for this disadvantage with the better thermal and crystal quality of SiC, making life pretty much a draw.

You can imagine that each firm plays up its advantages when it has them, so Lumileds is more forthcoming about Vf specs, while Cree really does not spec it at all, at least the last time I read those spec sheets.

Sorry for this long answer, but it gets away from me.

Please refer to this excellent thread by JTR1962 on white LED testing. Starting approx post 300, there is some truly excellent technical data, curves, etc, where he measured and plotted some "example" Power LED data for output, Vf, etc.

http://www.candlepowerforums.com/vb/showthread.php?t=89607&page=11

It is in this LED section, under "threads of interst", and I think called "white led testing".

The numbers are very real - but - apply only to that specific LED and its exact bin. The difference between Cree and Lumileds for LED performance is really tight - sort of like measuring 2 Olypmic class runners on 3 different races and over 4 days. Which one has lower Vf = the bin is more important than the brand.

So, for one more "dark horse", one of the LED packages that JTR tested was by LEDEngin. LEDEngin buys a narrow bin range of premium Cree die, and puts them (4 die) into their own, quite good package. They offer the package with the die wired in series, or in parallel.

For ease of testing JTR bought one wired in series, meaning those Vf numbers are 4X what they would be if wired in parallel, plus, the current would also be divided by 4, so the Vf is even lower. (in other words, since the current is now divided into 4 die, each is only seeing 25ma)

By my reading of the data, this would indicate that the lowest Vf power LED package he has tested to date, at a load current of 100ma, is the LEDEngin product. (this is the target current posted in one of the discussions above.)

I hope this makes sense - if not, I will try to answer more clearly next time.

Take care,

Harry

 

[HarryN]

The Vf of a K2 is actually pretty low. Keep in mind, the Lux I and III were Vf binned at 350ma, while the K2 is binned I think at 1,000 ma, so the Vf will naturally test higher for the same LED.
QUOTE]

HarryN, I've re-read this twice and I'm not sure which way to read it. forgetting about lux I or III as I have little experience with them, are you saying that the K2 will functionally have a lower vf at low ma and thus be more efficient at the low end?
In other words, how would an SCC P4, a Cree R2, and a K2 tffc, if each were sitting in front of you with a 'rated' vf of 3.0v-3.25v compare?

Hi Linger - since I wrote too much in the last post, I should really answer the question. I "think" (but please double check) that the P4 and R2 Vf are measured at 350ma, and the K2 Vf is measured at 1,000 ma. If you had all three of them, side by side, with the same Vf bin "rating", then the K2 would have a lower Vf at the same current. Yes, the K2 Vf is much lower then its spec at lower currents, and will be more efficient than at higher currents.

 

[znomit]

You can imagine that each firm plays up its advantages when it has them, so Lumileds is more more forthcoming about Vf specs, while Cree really does not spec it at all, at least the last time I read those spec sheets.

Yes, cree seem to be often quoted around the 3.6V value.
The datasheet says 3.3V typical at 350mA and 3.9 maximum. You might take that to mean a range of 2.7-3.9V but I don't think we've ever seen that range!

Personally, for this app I'd use a buck/boost driver.

 

[BurmaJones]

A different direction: I've seen white leds in cell phone back lighting run at 2.76vdc. These SMDs are approx. 0.8mm x 1.2mm plus the pads. A few hundred of these emitters could be built into an array the same diameter as a D-cell. Just a thought.

 

[Linger]

Yes, the K2 Vf is much lower then its spec at lower currents, and will be more efficient than at higher currents.

Thank-you. Hopefully this helps the OP. And it explains why the K2 was performing so well. Much thanx for your background info on manufacturer's datasheets.