Optimum D-size Mag for P7 on Alks?

[Scattergun]

I´m looking to build a D-size Mag, with a P7 preferably on DD. Wich bin and how many batteries are the optimum if going for a resistorless design?

I´m thinking along the lines of a 3,5-3,75V forward voltage and a 4D Mag... is this possible or should I go with a 3,25-3,5V Uf and three cells?

And if not possible.. should I limit current to say 2 Amps using 7135-circuits and go with a 4D?

Whats the optimum Mag using alks to power a P7??

 

[mash.m]

i prefer to use one d-size lion cell. you get the most power and runtime. if you wish to be on the shure side, then use a pwm driver with current limitation.

markus

 

[Linger]

I notice you posted 'P7 on alkaline' in the title.

I'm not sure what 'optimum' means to you. 700 lumens or 70 hours run-time. I have read that a higher vf will produce less output and longer run-times, while lower vf will be brighter for less run-time.

 

[Scattergun]

I guess I´d like to go for as bright as possible, without having the cells destroyed. I was just wondering about wich bin the LED should be given a certain number of cells. And if it was possible to run two 1050mA 7135-circuits on alkaline cells.

 

[jasonsmaglites]

its a tough science picking the perfect match. i just build a high vf 3d that is only pulling .2amps whereas my low vf 3d pulls 1.3amps. im changing the high vf led to a 4d and according to my multimeter i'll then get .7 amps. that seems backwards of logic that a 4d should get less current to the led than a 3, given they're all tenergy d cells, but its all in where and when you buy the led. your odds are definitely best with a 4d given that you're running alkalines on a p7, and i'd shoot for about 1.5 amps or else you're gonna sacrifice runtime on those alkalines for not alot more brightness. maybe find a mid vf like 3.5 exactally.

*just for your info my low vf is a 3.00-3.25 from cpf vendor
high vf is probably 3.5-3.75 from dx (ugly blue tint)

 

[Linger]

I'd like to clarify that jason's experience is exactly in accordance with what I posted - the lower vf led is actually brighter (and with shorter burn time) lower vf draws more current as the v batt is closer.

re: if it's possible to run two 7135's...

Possible, I'm not sure they'd run in regulation. Futhermore, they'd just be added resistance that you don't need. If you're just going for bright, direct drive is all you need.

3 alkalines will sag pretty quickly. 4D will obviously give you more output / longer run time. You're overvolting the P7 a bit when the alkalines are fresh (6.0v). However, some CPF'ers run their P7 DD off two rc123's (6.2v) and the 123's hold a higher voltage a lot longer. It means that you might not make that 50,000h life expectancy. But for the few years that you run the torch, you'd probably be happier with the 4D set-up, and long before the P7 starts showing age you'll have swapped it out for something else.

You'll need to ensure excellent heatsinking. You don't have a driver to reduce output when it gets too hot. Infact, the led will draw more current as it over-heats. This is bad, and you want to avoid it. Keep it cool.

 

[jasonsmaglites]

the 7135 simple resistors will help you keep a constant current until the batteries start direct driving the led (at the point the 1.4 amp 7135 is actually more than the almost dead alkalines can give they will no longer "resist" any current). find a chip with four of them, as they are 350 ma each you can have what i mentioned earlier driving it at 1.4 amps. yes the resistors will waste some of the initial energy of fresh alkalines. the two benefits is that by limiting the draw of those alkalines they will not sag and drive to the point that they become inefficent, thereby saving you runtime. not to mention that but 2.0 amps vs 1.4 with the p7 wont be that much brighter to your eyes. (ok a little) but it will strike a nice balance and give you a long lasting, cool running led. good luck whatever you decide.

 

[Scattergun]

I guess I´ll go for an DSVNI-bin and a 4-cell setup... If the 7135-circuits don´t add that much of a resistance I´ll probably connect two 1050mA in parallell. Or will this be to much for the Alks?? Should I limit myself to one 1400mA circuit?

 

[MorePower]

I guess I´ll go for an DSVNI-bin and a 4-cell setup... If the 7135-circuits don´t add that much of a resistance I´ll probably connect two 1050mA in parallell. Or will this be to much for the Alks?? Should I limit myself to one 1400mA circuit?

Let's do some figuring with the following assumptions:
4D cells at end of discharge are 0.9V each = 3.6V total Vin
7135 based circuits need 0.2V over Vf of the LED to regulate current
I bin LED with a Vf of 3.40V to make things easy (I bin is 3.25 to 3.5V)

At 2 amp constant current draw, a D cell will run close to 1:45 to 0.9V. That'll give you nearly 1:45 of non-dimming runtime, followed by a long period of direct drive with slowly dropping output.

At 1.4A constant current draw, a D cell will run 4:30 hours to 0.9V, followed by a long period of direct drive with slowly dropping output.

It's your choice as to how you want the light to work. Brighter, with shorter regulated output, or slightly dimmer, but with a considerably longer period of flat output.

 

[Scattergun]

Won´t the 1400mA circuit get mighty hot by reducing more current into heat??

 

[MorePower]

Won´t the 1400mA circuit get mighty hot by reducing more current into heat??

Don't think of it as reducing current, think of it as reducing voltage at a set current of 350mA.

What follows are approximations, but should be good enough for government work, as the saying goes.

You'd generate less heat with 4 chips (1.4A out) than with 6 chips (2A out), because each 7135 needs to initially "waste" (Vin-Vf)*0.35A of power.

From this we can see that at the start of discharge, each chip must deal with excess power of approximately:

(6V - 3.4V) * 0.35A = 0.91W

4 chips on a board heat up due to 3.6 watts dissipated, while a 6 chip solution heats up even more from dissipating 5.4 watts.

 

[kz1000s1]

Or will this be to much for the Alks??

Actually anything with a P7 is too much for alkalines if you want to see the LED's performance potential. Everyone keeps asking these questions and the answer is simple.

Don't bother trying to run a high performance light on low performance batteries.

There's no reason to for most people unless you're building something like wquiles did that was going to be used in Africa.
http://www.candlepowerforums.com/vb/showthread.php?t=234172
If you want bright, you will need rechargeable batteries of some kind, NiMh, Lithium-Ion or IMR. Alkalines will also end up costing more in the long run.

Now go build it and have some fun! :twothumbs

 

[Scattergun]

I have built several Mag-P7´s on rechargeable batteries, and I know that is the best option if going for maximum output. In this case rechargeables isn´t an option, and I need to go with Alks, and I want as much out of them as possible.
Saying I need to go with LiIon or NiMh is like answering a totally different question....

Another thing, the P7 isn´t a bad option for low current solutions, it´s a very efficient LED when underdriven. I just want to know wich solution is the best for Alks without stressing them too much.

 

[Linger]

+1 MorePower, thank-you for working this out step by step.

"because each 7135 needs to initially "waste" (Vin-Vf)*0.35A "
So to conlcude the OP's question, the most efficient solution is a P7 with a higher VF, as close to (Vin) during the majority of the cycle.
//Vin is a dynamic number that keeps reducing as cells are depleted. But with 4D cells and a 4x7135 amc chip (0.35amp draw), the alkalines will maintain voltage above 1.0 for around 6 hours
http://www.candlepowerforums.com/vb/showthread.php?t=64660
depending which brand is chosen.
(4D @ 1.0v = 4.0vf) Given the choices I've seen in the marketplace laterly, J bin is the higher VF available.)
***edit on the link there's a chart with D discharged at 0.5amp, which is roughly the 4 D cells sharing the 1.4amp load

 

[MorePower]

+1 MorePower, thank-you for working this out step by step.

"because each 7135 needs to initially "waste" (Vin-Vf)*0.35A "
So to conlcude the OP's question, the most efficient solution is a P7 with a higher VF, as close to (Vin) during the majority of the cycle.
//Vin is a dynamic number that keeps reducing as cells are depleted. But with 4D cells and a 4x7135 amc chip (0.35amp draw), the alkalines will maintain voltage above 1.0 for around 6 hours
http://www.candlepowerforums.com/vb/showthread.php?t=64660
depending which brand is chosen.
(4D @ 1.0v = 4.0vf) Given the choices I've seen in the marketplace laterly, J bin is the higher VF available.)

None of the tests in the thread linked were run at 1.4A. At 1A, the best cell didn't last 300 minutes (5 hours) to 1.0V. New cells (those tests are 3+ years old, I believe) will perform better, and my 4:30 to 1.0V came directly from some recent test data which was available to me.

Minimizing [Vin-Vf] will waste the least amount of power as heat; personally, though, I'd run a lower Vf LED and maintain flat output as long as possible.

 

[Linger]

None of the tests in the thread linked were run at 1.4A.

No, of course. OP had set on 4 cells so I don't need 1.4amp / 1D cell.
4 x D / 4 x amc7135 (1400ma) = 1 D cell per 7135 chip / 350ma output

There's a chart with D discharged at 0.5amp, my mistake for not specifying. Yes it's an old test so perhaps field conditions could be read a bit on the positive side of that.

Minimizing [Vin-Vf] will waste the least amount of power as heat; personally, though, I'd run a lower Vf LED and maintain flat output as long as possible.

Of course, I was just returning to OP's comment re: most efficient P7 on D alkalines (and OP clarified 'efficient' as more output over longer run time)

 

[MorePower]

No, of course. OP had set on 4 cells so I don't need 1.4amp / 1D cell.
4 x D / 4 x amc7135 (1400ma) = 1 D cell per 7135 chip / 350ma output

But that's not how AMC7135 chips work. They LDO regulators, which simply drop voltage. All 4 D cells would need to be in series, and would discharge at 1.4A if 4x AMC7135 chips were used.

 

[Linger]

You're totally correct, thank-you. I had to re-read that three times to figure out how I made a mistake (4D in series to equal 6.0v, also means each is discharging at 1.4amp)