Planning a 3 or 4-LED (44 watt!) quad core build.. need reflector/driver advice

Hmm..... now that I am back in lighting and I have some time to spare I am thinking of building something crazy

I have a 4D Mag sitting around doing nothing, and I'm thinking of re-using the well-understood, safe 6xSubC NiMH powertrain that has served the ROP well for so many years. Easily good to 40W that the ROP requires, no problems here. As a bonus, this powertrain delivers the same voltage as 2 x lithium ion, making design much simpler (7.4V). We'll pull max 6A from a high-rate 3300mAH NiMH pack - slightly below 2C discharge even. The powertrain part I have settled, it's the reflectors/drivers I need advice for.


As for the front end, I'm thinking of using the quad-core LEDs, either 3-4 of them. I need a recommendation for a P7 reflector.

Plan 1 is to use 3 SSC P7's in series, but I need to find a good boost driver to get me to around 11 volts up from 7.4. Preferably one with a simple UI. 3 modes would be nice, 1 mode would be perfectly ok (ie. max blast!) and I don't want to have to rtfm to find out how to change modes on the light.

Plan 2 is not being able to find a driver, and going direct drive 2S2P. :devil: yes that smile is because I am perfectly aware of the thermal issues of driving 44 watts of LED in a mag head. I am only debating with myself, should I go nuts and go for a 2S2P setup or a far more tame (but more practical) 2S1P setup with just 2 LEDs in series? ie. do I want to build a flamethrower, or do I want to build a light that is actually useable without USL-style trigger locks! I'm already thinking I will have to store this one unloaded without its 6 NiMH pack lest someone accidentally get hold of it and cause thermal burns

Call me crazy, egg me on, give me advice: all welcome, bring it on!

I believe user Der Wichel has what you're looking for: https://www.candlepowerforums.com/threads/229098

3 SSC P7/MC-E heatsink+3reflectors+driver

Awesome, thanks!

Unfortunately it looks like I can't use his driver, which makes a 3S setup impractical. Direct drive has to be at least 2S LEDs (3.7v x2)

http://www.cpfmarketplace.com/mp/showthread.php?t=217586

Hmm, I'm also wondering about plan 3 - six XP-Gs using two of the individually addressable 20mm boards. 2S3P direct drive?

2S2P MCEs x3, you may not need a driver:

https://www.candlepowerforums.com/threads/289680

I can't give any advice, I'm the guy they test stuff on to see if it's idiot proof, but I want to ask.. Why do you want to use multi-die emitters in the age of the SSTs and XP-Gs? Cost?

You could do a triple SST-90 or a 'triple triple' as they call the 9x XP-G with three Carclo optic-builds.. :naughty:

ma_sha1 said:

2S2P MCEs x3, you may not need a driver:

https://www.candlepowerforums.com/threads/289680

Click to expand...

Aha..... a thought..

jellydonut said:

I can't give any advice, I'm the guy they test stuff on to see if it's idiot proof, but I want to ask.. Why do you want to use multi-die emitters in the age of the SSTs and XP-Gs? Cost?

You could do a triple SST-90 or a 'triple triple' as they call the 9x XP-G with three Carclo optic-builds.. :naughty:

Click to expand...

Cost and because I'm familiar with them - they're easy to solder especially the P7. I have very limited faith in my modding skills I can botch a KL1 emitter swap

The SST series are a nightmare to solder and the SSR-50 star commands a staggering 50% price premium over a SST-50 emitter. Of course the emitter is next to impossible to solder with my limited capabilities. Worse, with their super high amperage, they will need thick lines which are even harder to solder. I definitely prefer stars as they are easier to manage.

The 9X XP-G sounds great, with one small problem: driver. I need a boost driver capable of accepting 8.4V input, boosting to 11.1V to drive each group of 3 in series. Not only that, since the driver has to do 3 groups, it has to provide 4.5A!

DW's SST 5/9A driver is a buck driver and thus doesn't suit my needs.

I was thinking 6x XP-G in individual addressable boards, so I can wire them 2S3P for direct drive. Very similar to the MC-E topology described above. 3x MC-E delivering ~2000 emitter lumens is actually pretty close to 6x XPG at 1A (also about 2100lu) so it might be a race between the two of them.

If I wire 9 XP-G in this topology I will have one LED left over if I run 2S4P (8 LEDs). If I hook it up straight to the 8.4V source on its own, kaput

KevinL said:

Awesome, thanks!

Unfortunately it looks like I can't use his driver, which makes a 3S setup impractical. Direct drive has to be at least 2S LEDs (3.7v x2)

http://www.cpfmarketplace.com/mp/showthread.php?t=217586

Click to expand...

Maybe you could try something like the workaround user download came up with?

https://www.candlepowerforums.com/threads/201392

aEx155 said:

Maybe you could try something like the workaround user download came up with?

https://www.candlepowerforums.com/threads/201392

Click to expand...

Thanks! The multi-MC-E DD setup looks interesting.

ok guys, appreciate all the help. I think after a little bit of inspiration and LiFePO4, I have figured out how I want to do things.

I credit ma_sha1 for the LiFePO4 idea....
https://www.candlepowerforums.com/threads/295357

Except instead of 2, I am going to stack 4 of them into my 4D Mag to deliver a nominal 12.8V. This also then GREATLY simplifies my LED wiring because with a 12V pack I can now use strings of 3 LEDs. My plan is a little different - I know the SSTs are the hottest thing in the land right now, but I plan to use those triple XP-G 20mm modules wired in series.

And I plan to use 3 of them. 3x3 = 9. :devil: wire three parallel strings of 3 series LEDs (3S3P), 1-1.5A per string. 20mm Carclo triple XPG optics on top. It won't be that much of a thrower but I guess it'll blind everybody.

At 450 lumens/1.5A, I could theoretically be hitting 4K lumens. on freshly charged 'hot' cells with a 1 ohm resistor (plus losses that I'm sure are present in the Mag).

At 350 lumens/1A, an easy 3K lumens still. And a 3A draw will be easy on the Mag and the cells. This particular Mag used to be the Original ROP (4D/6-SubC) so I know it does 4.2A no problem.