CREE MC-E @ 1000ma (250ma per die)?

[SlowE60]

Hi all,

Newbie here - I'm trying to build a custom flashlight but I am limited to a single 18500 battery. I am wondering what the brightness levels would be if I were to run a CREE MC-E @ 1000ma. Has anyone tried this before? Is this the best solution?

How far am I away from 300lumens?

TIA!

 

[SlowE60]

CREE MC-E @ 1000ma (250ma x 4) ok?

Hi newbie here -

I am wondering if it's ok to run a CREE MC-E @ 1000ma (250ma x 4) as I am limited to a 18500 battery (3.7v). What would the brightness be? Does anyone have any experience with this? How close to 300 lumens would I be? This would be a grade K or higher MC-E

TIA!

 

[Ryanrpm]

Re: CREE MC-E @ 1000ma (250ma x 4) ok?

First of all....:welcome:

Second of all, slow down!

You generated 2 threads essentially saying the same thing....and only 17 minutes apart. It takes time for members to see the questions and then respond.

An MC-E at 1a is IMO severely underpowering it!! :naughty:

It should operate fine...and I also beleive that it will be over 300 lumens at that current. I could be wrong though.

 

[DM51]

Welcome to CPF, SlowE60.

Please don't post the same thing twice. I'm merging your 2 threads.

 

[bigchelis]

Hi all,

Newbie here - I'm trying to build a custom flashlight but I am limited to a single 18500 battery. I am wondering what the brightness levels would be if I were to run a CREE MC-E @ 1000ma. Has anyone tried this before? Is this the best solution?

How far am I away from 300lumens?

TIA!

If you use the DX 1000mA drivers I am sure it would work. DX has MC-E P60 drop-ins for only $23 delivered. They make 340ish out the front on high and 150lumens out the front on low. Check the Sticky 6in Lumens thread.

I think you could just get any P60 compatible light and then drop in the MC-E with those 2modes, but the DX also has strobe.:sick2:
Jose

 

[Justin Case]

I built a 4C Mag mod using a Seoul P7 DxxxI-bin driven at 1000mA (250mA per core). The MC-E in 4P is very similar. If you have an M-bin MC-E, you can expect about 250-300 lumens out the front. For a K-bin MC-E, maybe about 220-260 lumens out the front.

IMO, this is a great setup. Very cool running, low stress on the batteries, low stress on the LED.

 

[Lighthouse one]

Are you sure you're ready to build a custom light already. There are many available in the $40 range already.

To answer your question....a single 18650 will be able to drive an led at 3,000 ma or a bit more. Many of the 4 die lights run at 2800ma and push over 500 lumens out the front. This gives 700 ma to each die. The reflector design has a lot to do with the beam- as most 4 die lights have tremendous spill, but not as much throw.

Stay awhile and read some more before you decide.

 

[kramer5150]

I am doing something similar with a 1400 mah driver:twothumbs. I was going to use an XR-E-Q5... but I no longer think thats optimal

I dont think its a bad idea at all under-driving the MC-E at 1-1.4A. An XR-E driven at 1000mah will get warm quickly if you dont provide it with an adequate thermal path. An MC-E will hardly break a sweat.

 

[SlowE60]

Hi guys,

Thanks for your response! While I'm a total newbie to flashlights, I do have access for a 3-axis CNC machines and CAD so I think I can handle this project with a bit of trial & error!

Since a few of the responses was that at 1000ma (250x4) the MC-E would be underdriven - should I bump the ma a bit? Maybe 300 x 4? Would an 18500 battery be capable of supprting this?

Lighthouse one - how long would a 18650 be able to drive the MC-E at 3000ma? I thought the battery was only rated at 2400ma (so less than 1 hour)?

Thanks!!

 

[kramer5150]

Hi guys,

Thanks for your response! While I'm a total newbie to flashlights, I do have access for a 3-axis CNC machines and CAD so I think I can handle this project with a bit of trial & error!

Since a few of the responses was that at 1000ma (250x4) the MC-E would be underdriven - should I bump the ma a bit? Maybe 300 x 4? Would an 18500 battery be capable of supprting this?

Lighthouse one - how long would a 18650 be able to drive the MC-E at 3000ma? I thought the battery was only rated at 2400ma (so less than 1 hour)?

Thanks!!

that all depends on how robust the thermal design is. If you design with an emphasis on thermal conduction, emissive radiation and cooling, there should be no reason the emitter can't handle more current.

The fore-mentioned DX-MCE gets warm (but not hot) running at 150 Lumen low when using it in a surefire 6P. Running it at 345L on HIGH gets very HOT quickly.

IMHO 4x250mah is not under-driving the LED... if its being driven in a poor thermal conducting host. In a sufficient cooling host you can afford to drive it harder.

An 18650 should be able to handle 300x4 (1200 total). Note that the DC-DC circuit might need to draw more than that.

Oops.... Im not lighthouse one... sorry I can't help myself.

 

[polkiuj]

SlowE60

DX_MC-E High_____________385__turn-on,__________1_18650 LI____________Solarforce L2 Host AR coated glass,
DX_MC-E High_____________345__warm____________1_18650 LI____________Solarforce L2 Host AR coated glass,

From http://www.candlepowerforums.com/vb/showthread.php?t=229135

Tested with IS by MrGman. I'm unsure of the driving amp of this unit. Perhaps you can ask MrGman. Quoted figures are out the front.

 

[Gunner12]

A good 18650 battery should be able to drive at 2C without much problem. 2C is almost like 2x the capacity rating, so a 1000 mAh battery at 2C would be driving at 2000 mA. At 2C the battery should be discharged in half an hour. A 18500 shouldn't have problems with the LED at 1400 mA (350 mA per die).

 

[Justin Case]

I would suggest sticking with the original plan. 250ma per core will give you a bright, cool-running light. Only a couple of years ago, flashlight owners would have drooled over 250-300 lumens out the front running on 1x18650. Now people are being seduced by these 700 lumen monsters.

If you went with 350ma per core, you'll probably get about 100 lumens more OTF. I have several MC-Es and P7s driven from 250ma to 613ma per core. The difference really is not that striking in my observations using these multi-die LEDs in various SureFire TurboHeads and Maglite heads. IMO, the real key is how the light is distributed, i.e., the type and quality of the beam pattern. In fact, a prosaic Seoul P4 can outthrow almost every one of the MC-E and P7 LED lamps I have. The only time I had a P7 have a higher hot spot lux than a P4 was when I went direct drive and ran a P7 at 2.8A in a smooth reflector. But the P4 was driven only at 1A and ran much cooler and longer.

 

[yellow]

1A to led --> can be made with a single die also (when thermal path is good)
the less floody beam of the single will give the impression of a much brighter light, in the beam at least

imho using a quad emitter @ 1A makes no sense at all, give it 2A max
... considerably brighter while still not cooking

 

[Justin Case]

Driving a 4-die LED at 1000ma makes lots of sense. You have a lower Vf vs a single die LED for the same forward current of 1000ma, and thus that much less heat and that much longer run time. If you are using a boost driver, the lower Vf helps to ensure that you run in regulation. If you are running a buck driver, the voltage overhead to reach regulation is also less.

The down sides are that the 4-die emitters cost a lot more than the single-die and they don't focus as tightly.

 

[Benson]

Driving a 4-die LED at 1000ma makes lots of sense. You have a lower Vf vs a single die LED for the same forward current of 1000ma, and thus that much less heat and that much longer run time.

And you get more light out!

 

[kosPap]

hmmm seems it is a good time to ask soemthing about this too...
how about using a 1000mA driver that is capable of driving two LEDs in series...
`
wiring the MC-E in 2sereis2parallel should it be geting 2x1000 = 500mA per die?

 

[Justin Case]

More heat generation, shorter run time. In 2S2P, the voltage drop on the MC-E is around 6.6V for 500ma per core. You have to deliver about 6.6W to the LED, meaning even more power generation by the driver.

I'm running this configuration in an AW Turbo Tower for a SureFire TurboHead: K-bin MC-E in 2S2P using an SOB1000 buck driver.

However, if you plan on using 1x18500 (or any 1xLi-ion), you will need a boost driver that can handle the required input current to reach regulation. If we assume 90% driver efficiency (being slightly generous on the part of the driver) and a Vbatt of 3.7V,

Ibatt = 6.6V*1A/(0.9*3.7V) = 2A

It looks like the TaskLED Fatman and maxFlex drivers can handle this.

For my application in an AW Turbo Tower, I couldn't use these larger driver boards. The AW tower is made to accept 14mm drivers, and small drivers like that just aren't going to like 2A on the input.

I went with a minimum battery source of 2xAW16340 in a 6P body (also could go with 4xSF123A in a 12ZM body, for example) and used the SOB1000 buck driver. Using the same calculations as above except assuming Vbatt of 7.4V, we get Ibatt = 1A.

 

[yellow]

single die at full vs. quad die at low current:
I dont swallow this, has anyone with equipment measured that?

sure: less current = less heat
but its 4 dies running and their heat adding up together

there might be a bit of positive for the quad - tiny bit less power transferred to heat - but in reality I would wonder if a difference can even be measured
(in a sphere, where the wide beam of the quad won't make much of a difference to the single die beam)

 

[Justin Case]

single die at full vs. quad die at low current:
I dont swallow this, has anyone with equipment measured that?

sure: less current = less heat
but its 4 dies running and their heat adding up together

You are welcome to believe that Vf at 250ma is the same as Vf at 1000ma, until someone actually provides measurements. But I don't need to see the sun set at night to know that it is setting.

Regardless, see this thread for some Vf data. In this other thread, George at TaskLED measured Vf at about 3.4V-3.5V at 700ma per core. I think it is quite clear that the P7's Vf at 250ma per core will be significantly lower than that.

The total power for the multiple dies have been accounted for. A P7 has 4 dies running in parallel. The voltage drop is Vf, estimated at around 3.1V. The forward current is 250ma per core, or 1A total. That is 3.1W.

An SSC P4 at 1A forward current is probably 3.5V for Vf, or 3.5W total.

The real point is that with a significantly lower Vf, you can run a boost driver with 2xNiMH and have confidence that you'll run in regulation and you can run a buck driver with 1xLi-ion with the same confidence. With higher forward currents, that is simply not the case.

And you get the added benefits of longer run time and greater light output for the same forward current vs a single die like an SSC P4.