Multi P7 heatsink

Hi All

Looking for a spot of advice. Im looking to build a 3x P7 mod in a m&g head- the batteries will be contained in a seperate waterproof canister and will be connected via umbilical. As it is intended for underwater use, i will have the ocean as a pretty decent heatsink as well as the aluminum head.

Heres the question- how close together could you put the P7's? Ive looked at the specs, and it seems that 3 can fit on a H22A D size heatsink (was thinking id sand off the ridge and flip it round so the LEDs were on the flat side). i could add more aluminum for extra mass, and i would use paste between contact surfaces for more efficient heat transfer. In addition, since its a dive light, id have the Sea as a constantly moving, high heat capacity medium on the outside.

Any thoughts on if this is possible? would be looking to run it with 3 Li-ion at 2A. Intended driver is the 'poorman multilux' solution found on the site.

Cheers for any advice, much appreciated!

download, the same guy that created the poormans multi lux thread sells a multi sink I think you could you.

If I was you, I would put as much reflector as you could.

Thanks, ill look into that.

I wasnt planning on using a reflector. I currently use a P7 with a fresnel lens, and with some slight defocusing can get a real nice circular hotspot.
Currently, its about 8" hotspot at 10 feet. If i concentrate the 3xP7 as close as possible, and defocus a little more, it should give a good circular hotspot with not noticeable die shapes. If i get it to focus a 2400 lumen output to 12" at 10feet, ill be very happy.

Thanks again

I think it would be easier to use a PES with the lip machined off,available here,http://theledguy.chainreactionweb.com/product_info.php?cPath=48_53&products_id=1162

Or one of download's adjustable sinks available here,http://candlepowerforums.com/vb/showthread.php?p=2060148#post2060148

Or maybe one of LED Zepplin's quad P7 sink if they're available,but I dont think they fit a standard Mag D head,https://www.candlepowerforums.com/threads/205635

It would probably be easier to buy one of these instead or having to buy a standard D sink and then modify it.

-Michael

The problem is that i need to have the emmitters as close as possible to the centre point, and about 6 cm away from the front of the head. Only this combination, or close to it, will give me the beam i am thinking about when using the fresnel. Anything closer to the lens, and a wider spacing of the LEDs will give a pretty noticeable artifact to the beam. All of those heatsinks sit on the flat area in the m&g head- i need something that is the same diameter as the tube, so that the LEDs can be moved away from the lens by backing the tube out of the head almost completely.

Thanks for the links though, much appreciated

I was going to do something similar using the heat sink from here . But in the end I brought a mini lathe and made my own sink.
What put me off the 3x P7s was lack of a tight beam so I ended up going with 4 x Q5's with aspheric lenses instead.

Which Fresnel lens are you using? 1 per LED or 3?

What design are you using for your can?

Hmm. With the LEDs sitting so far back wouldn't you lose a lot of side spill light from them resulting in poor out put?

Im also guessing heat may be a problem. You would want to make sure the heat sink had a very good thermal path to the body.

Hi

I bought some of the 50mm diameter 32fl fresnel lenses a month or so ago and love them. Really tightens up the beam of the P7, and with a little defocusing you get a nice round beam. The hotspot is about 8" diameter from 10 feet. im looking at using the same lens, and by clustering the P7s with a little further defocussing im hoping to achieve the same effect- it works great ofr the P7 (effectively a tight cluster of P4s). The problem is id need to defocus about 6 cm or so away from the lens, which is almost backed out of the head- about an inch lower than the flat part of the head where the other sinks sit. Hence my current dilemma

My canister is an 8 inch long piece of 2.5 inch PVC sch40. Caps are made out of laminated 1/8 PVC flat stock. Bottom is PVC glued in and the top is secured with clips. painted black, looks pretty good.

I could have made it a lot smaller, but i kinda like the size. Plus ill need it if i upgrade to 3 P7s- dont want the extra expoense hassle of Li-ions, so sticking with the Nimh Cs i have already

Cheers

Got a link for the Fresnel?
Any beam shots of you P7?

The P7 is on the left and colour balance was set to the LED. The right hand side is my original 35W halogen, with a 12 degree spot. Exposure is metered so as to not completely blow out the hotspots of either- in reality, there is a lot more light coming from both of these. This shot was to get a decent comparison of the hotspots. The bottom image is of a store bought 9watt LED dive light for comparison.





edit- photo embeded

The fresnel was from Rolyn Optical. They have a $75 minimum, which gets you 5 lenses. Shipping in CONUS was a further $20. I actually have one spare im trying to sell in the marketplace.

http://www.cpfmarketplace.com/mp/showthread.php?t=182703

Thanks.
Looking at the pics it looks like the P7 has a larger hot spot but less spill than the halo. Is this correct?

Yet you say the hot spot is 6in at 10ft which is clearly a lot tighter than 12 deg.

What sort of lens are you using on the front of your mag?
I ended up using 2 x 1.8mm lenses from DX. I have had them down to 60meters in a pot and 37 meters in a lake but I have no idea much pressure they can take.

Packhorse said:

Thanks.
Looking at the pics it looks like the P7 has a larger hot spot but less spill than the halo. Is this correct?

Yet you say the hot spot is 6in at 10ft which is clearly a lot tighter than 12 deg.

What sort of lens are you using on the front of your mag?
I ended up using 2 x 1.8mm lenses from DX. I have had them down to 60meters in a pot and 37 meters in a lake but I have no idea much pressure they can take.

Click to expand...

Sorry, this should have been 8" at 10 feet. The hotspot from the P7 is about the same size as the canister, with little spill.

The hotspot from the Halogen has been confusing me a little too. The 12degree from the 35w is what you see in the picture- the exposure was set to just show the hotspot. However, it seems that the 'real' hotspot of the 35w is less than 12degrees (probably a little smaller than the P7) since there is clearly a brighter inner spot and the less bright outer spot. Bear in mind though that at 'normal' exposures, its very difficult to see the difference in the inner and outer zones of the halogen hotspot, so i think the whole ring is being used to calculate the beam angle. Measuring the hotspot in the halogen, you get ~11 degrees if you include the spill, and ~7 degrees if you only look at the inner bright spot, so i think the 12 degree measurement is including the 'spill'.


For the P7, the 8inches at 10 feet gives me a beam angle of about 8 degrees. Not too shabby.

Im using the Acrylic fresnel lenses cut to a 50mm circle, and since it 1.5mm thick, it fits perfectly in the front groove in the maglite head. Then a 2.5mm UCL glass lens on top of that, sitting flush against the top of the head. A 3mm o ring seal the bezle against the glass, with a 2mm oring at the base of the threads of the head. Had this down to 130 feet about a dozen times when used in the original halogen head. Dont plan on going much below this, so ive been happy with it

Cheers

Packhorse said:

What put me off the 3x P7s was lack of a tight beam so I ended up going with 4 x Q5's with aspheric lenses instead.

Click to expand...

Any pics of the 4x Q5? Estimated overall beam angle?

Cheers

Its still a work in progress.
But I have used a Aspheric and Q5 in another light.


This is the project so far.
I am short 2 Q5's so that is the hold up. It has a KAI 2.8amp driver and the LEDs will be in series parallel resulting in 1.4amps to each Q5




This shows beam patterns. from Mag P7 Q5 asperic and Q5 with 20mm reflector. Q5s driven at 1 amp.
I intend to run all 4 beams in parallel resulting in a beam angle the same as above but 4 times more intense.

Beam is 30 cm at 2.4 meters no spill. resulting in a true 7 deg beam. At least thats what I calculate it at.
2400mm X 2 XPi= 15,000mm diameter circle
15000/360deg = 42.666 mm per degree
300mm /42.666= 7 deg.

Is that correct?

I have run this 3 x P7 with 1400A four 10 min and i can hole this in my hand . I will wait to i have modifield the Mac head to 1 Big cooler befour i try the 2800 A it will be a Dive ligth to ( sorry for my englisk )

Download's multisync may vest suit your heatsinking needs as you can adjust the height of the sync and you can place the emitters as close together as you like - no center pillar to sand.
https://www.candlepowerforums.com/threads/168062

Packhorse said:

Beam is 30 cm at 2.4 meters no spill. resulting in a true 7 deg beam. At least thats what I calculate it at.
2400mm X 2 XPi= 15,000mm diameter circle
15000/360deg = 42.666 mm per degree
300mm /42.666= 7 deg.

Is that correct?

Click to expand...

yeah, i think thats right. It actually makes me realise my numbers were way off. A combination of posting at work and being understaffed. My apologies. I re-did the numbers, and measured the hotspot a bit better with a real ruler rather than guestimating from the soze of the canister, and got a hotspot diameter of 25 cm from 200 cm distance (these units are far easier, just because i moved to the US doesnt mean i need to forsake real measurements!). That gives me a beam angle of around 7 degree also. This isnt the tightest i can get it, but its the tightest useable 'round' beam i can get. im going to re-do these images with overlays once i get some time at the weekend

Your Q5 with the aspheric- is that a single Q5? Does the actual size of the pattern change when you add a 2nd and 3rd Q5. Id expect that the spill increases also, given that a lot more light is coming from 'off centre' of the aspheric and doesnt contribute to central hotspot because of refraction. Thats one of the reasons the LEDs need to be real close.

Also, if that is 2 Q5s, any chance of a fast exposure shot, showing the die pattern more clearly? In the P7, with a fast shutter, you can see the individual dies and the bond wires. Actually, come to think of it, running it at 1 amp, you can begin to see the effect with the naked eye, but thats because light output is so low.

The white acetal ring- is that your magnetic switch collar? i was messing around with a piezo switch, but might actually end up going back to a regular toggle switch. No-one i know has had a switch boot tear on them and cause a leak, so it might be safe enough.

nelson49 said:

I have run this 3 x P7 with 1400A four 10 min and i can hole this in my hand . I will wait to i have modifield the Mac head to 1 Big cooler befour i try the 2800 A it will be a Dive ligth to ( sorry for my englisk )

Click to expand...

looks good. how big is that heatsink? does it fit in the maglite body, or does it rest on the little 'shelf' on the inside of the head?

How are you going to power the P7s? sticking with the SLA, or using Nimhs? Or even Li-ion? Id be interested in hearing about your switch mechanism. Anything fancy, or just a regular toggle?

Good luck for the rest of the build

packhorse- just noticed the finning on the head. Nice!

I so want a mini-lathe...

Yeah that is just 1 aspheric Q5! Cant wait to get 4 going. As long as I can aim them all correctly there is no reason the beam will increase in size more than 2 cm from any length. Less than 500mm you may see each LED but beyond that it should look like one.
The spill from it is minimal. Its well focused and shows all the detail of the die including the wires when shone against a smooth surface. But when its used underwater its just a nice square of light. To be honest round light beams are over rated ! LOL

Yeah the front white ring will be the holder for the magnet and the rear one will be epoxied to the body as a stopper. You can see the grooves in the heatsink where the reed will sit. Still need a suitable magnet. I will probably still use a reed on the can too so I can turn them both off while in transit.

The finning on the head aint al the precise. Was just fooling around. Was probably only my 2nd job on a lathe ever.