P7 O-sink interest

[yaesumofo]

HI guys I was wondering if there is any interest in an O-Sink for the P7 emitter?
I have yet to see one in person but If there is sufficient interest I will embark on research and development program and will get some sinks made.

I understand the P7 runs HOT and needs great heatsinking. I could design the sink to extend deeper into the body adding mass to the sink which would be hidden inside the body of the light.... The Mag host is a natural. I don't know what the beam will look like I suspect that an aluminum reflector may be called for if the P7 is driven at it's max.
The 3 cell mad may be right for direct drive...
Ideas thoughts? welcome...

Yaesumofo

 

[2xTrinity]

I'd be interested in one. A 2- or 3-c P7 light would be nice for sure, a great blend of runtime and output.

]I understand the P7 runs HOT and needs great heatsinking. I could design the sink to extend deeper into the body adding mass to the sink which would be hidden inside the body of the light.... The Mag host is a natural. I don't know what the beam will look like I suspect that an aluminum reflector may be called for if the P7 is driven at it's max.

Nah. Aluminum reflectors are needed because incans emit most of their input power as IR radiation, which is largely absorbed by the reflector. Even at 900lm, this will probably heat the reflector less than a stock mag bulb as ALL the light is in the visible range (unlike the mag where most isn't). With LEDs, most of the input power must be conducted away using heatsinking.

 

[Supernam]

I played with the P7 AA'd to a DHS "D" size sink. It was not attached to the body, but rather in a test configuration. However, after about 20 seconds direct driven, the sink got too hot to hold with a firm grasp. Surely if it were in the Mag body, it would dissipate heat better, but this just gives you an idea of how hot it gets.

The extended o-sink would be great, but make sure it doesn't go too long as to inhibit the use of drivers. I'd say there should be about .5" space between the bottom of the sink and the top of the switch. I think a possible design modification would be to make a heatsink similar to those that use multiple emitters. That is, one that has a large platform that fills up the inside of the head instead of just plugging into the body. This would ensure that heat is efficiently transfered directly from the emitter to the outside of the head. Most of the multi emitter sinks have a hole in the middle so they wouldn't work, also there would need to be a post for the emitter to go into the reflector.

It'd be a good idea to use thermal grease to lube the threads of the Mag head and the body to have efficient heat transfer between the two parts. I believe a finned mag head would significantly increase surface area for more effective heat dissipation.

I'm currently waiting for parts to come in to complete a Mag 4C Nimh using a DHS sink. It'll be running at 2.8A (realistically 2.4A) I'm worried that the C mag has slightly less mass than the D mag to dissipate heat. I'll definitely fin this Mag's head. I'll be locking the head in place after finding the best focal point, so I'll use Arctic Silver to "glue" the threads.

 

[Supernam]

I'd be interested in one. A 2- or 3-c P7 light would be nice for sure, a great blend of runtime and output.


Nah. Aluminum reflectors are needed because incans emit most of their input power as IR radiation, which is largely absorbed by the reflector. Even at 900lm, this will probably heat the reflector less than a stock mag bulb as ALL the light is in the visible range (unlike the mag where most isn't). With LEDs, most of the input power must be conducted away using heatsinking.

I agree that aluminum reflectors are not necessary with LED's, however it is preferred that one uses a textured reflector since the emitter's surface is so wide. I wouldn't mind using the a plastic reflector if there were some kind of orange peel to it. I'm going to try using clear spray paint on a stock Mag reflector and report my findings. I have a FiveMega MOP reflector that I'm going to enlarge the hole. It might be easier to buy the KaiDomain 15mm opening MOP reflector though.

 

[gojira54]

Hello :0)
I've been lurking this forum for a while!
I'm definitely going to P7 mod a m@g 3D direct drive. I've seen that there are gonna be heatsinks available from a couple of sources in the dealers corner in the near future but none of them look like they are out to maximise heat dissipation for a direct drive application. I think it's a great idea to make the sink as chunky as possible!
I'll be in for a few if these get made :thumbsup:

 

[ambientmind]

I think an o-sink for c and d mag lights would generate great interest. I know I'd be in for one of each, I actually need one right now for a c sized mag.

 

[DFiorentino]

Count me in.

-DF

 

[yaesumofo]

I will look into it. In this day and age of cheap Chinese flashlights coming out left right and center it is difficult to make a product which can compete. Lucky for us they don't concentrate their efforts on accessories and modification parts. What is crazy is that places like DX have complete P7 lights available for sale for under $50.00!!
The idea of modding a mag to accept a P7 is not new. Several people have already done it.
What may end up being new is the heat sink used to to the mod.
I have found that the space inside the Mag body behind the space that is created when a O-Sink is mounted is Huge!! There is plenty of room in there for more aluminum in the form of an extended sink and a driver. IMHO the driver should be mounted into the sink as well so that heat generated by the driver may be efficiently moved.

My fear is that there is simply TOO MUCH heat generated by the P7 to efficiently move it using a passive heat sinking system.
These emitters make a lot of heat. even with the mass of the head being thermally well attached to the body of the light there is a lot of heat to move.
The idea of three P7's running at full power ..I see a light which may not be usable for more than a few minutes...


My focus is on creating a sink which will allow a P7 to be mounted inside a Mag host for it to move heat efficiently enough to allow for continuous use of the light.... Maybe copper? I hate using copper because it is so yummy to machine....then then there is the issue of cost...copper is pricey especially if you choose to use the tellurium copper. I would be buying 8 foot lengths of material. Buying less is silly due to the machine shop minimums...
1 1/2 inch 8 foot length of 7075 costs $130.00
6061 is about 1/2 the cost of 7075...
the same thing in copper costs over $900.00

A copper O-sink will cost 7 or 8 times as much as an aluminum unit. I material cost alone. WOW!!

So anyway I think I will focus on 6061 for the sake of those of you who are budget minded. IN the past an o-sink cost the end user $10-$15.00 if memory serves me. I suspect that material cost have gone up since I made those as well as the cost of machine time ...everything has gone up.
Research is needed. I am unwilling to do a lot of work on a product with an extremely limited market.
Unfortunately on the surface of things it looks to me as though I may not be making these. there is good size investment required and the market for this part is not what I would call booming.
If you guys have ideas which can help make this idea become a reality please share them with me.
As I said I will do the necessary research to learn about the feasibility of mounting the P7 onto a hunk of aluminum in a Mag host.
Any info you guys feel like sharing regarding this please post.
Yaesumofo

 

[Meduza]

Of course i would be in for one

 

[Supernam]

Yaesumoso,

I am wondering if it would be a better idea to mount drivers separately than the emitter's sink. I suppose it would depend on what driver/battery combo you are using, but suppose one was using 3 or 4 nimh's with a driver for the P7. The Vbatt and Vf are so close that it shouldn't create that much heat. I believe the the emitter/sink temp would be much greater than the driver.

I'll be mounting my AMC7135's right above the switch by AA'ing them to the body (with the chips touching the body). I think that the heat from the emitter will mostly conduct outwards through the head so the temperature of the body just above the switch will be cooler than the sink.

Regarding the copper sink... I'm no physicist, but I wonder if it would actually help significantly to use a copper sink. While the heat would conduct away from the slug faster, it would still encounter the barrier of the aluminum body which would have to then conduct through body's threads onto the head's thread then though the head itself to its surface. Not to mention that there is a coating of anodizing through each of the surfaces. Certainly a copper sink would help, but the question is, by how much? And considering that the cost would be nearly 10 times more, would it be worth it?

Again, I'm no pro, but I think it is important in this type of set up to get the heat from the slug to the air as efficiently as possible. I will try to do so in my own set up by stripping anodizing from inside of the body and also AA'ing the threads. Finning and/or fluting the head would also greatly increase the surface area of BARE aluminum to the air.

 

[PhotonFanatic]

Supernam,

You seem to imply that anodized AL will not conduct heat as well as bare AL. Can you point me to any published data showing that to be the case? Thanks.

 

[Supernam]

I cannot. I am only assuming that it doesn't.

I am assuming that, although minute, anodizing has some thickness to it or at least alters the surface which may have less thermal conductivity than an aluminum to aluminum contact.

 

[Supernam]

From wikipedia (http://en.wikipedia.org/wiki/Anodising): "Anodized coatings have a much lower thermal conductivity and coefficient of linear expansion than aluminium. As a result, the coating will crack from thermal stress if exposed to temperatures above 80 °C. The coating can crack, but it will not peel.[5] The melting point of aluminium oxide is 2050 °C, much higher than pure aluminium's 658 °C.[5] (This can make welding more difficult.) In typical commercial aluminium anodization processes, the aluminium oxide is grown down into the surface and out from the surface by equal amounts. So anodizing will increase the part dimensions on each surface by half of the oxide thickness. For example a coating that is (2 μm) thick, will increase the part dimensions by (1 μm) per surface. If the part is anodized on all sides, then all linear dimensions will increase by the oxide thickness. Anodized aluminium surfaces are harder than aluminium but have low to moderate wear resistance, although this can be improved with thickness and sealing."


Others:

http://stinet.dtic.mil/oai/oai?verb=getRecord&metadataPrefix=html&identifier=ADA191755

http://www.frostytech.com/articleview.cfm?articleID=233

http://www.finishing.com/3200-3399/3207.shtml

Assumptions ended up proving true.

 

[PhotonFanatic]

Thanks and right you are.

This chart is quite helpful:

I'd love to make some silver heat sinks, but for a P7 it might be a tad prohibitive--a 1' length of 1.25" diameter coin silver rod sells for $2,374.00. :devil:

 

[cmacclel]

I agree that aluminum reflectors are not necessary with LED's, however it is preferred that one uses a textured reflector since the emitter's surface is so wide. I wouldn't mind using the a plastic reflector if there were some kind of orange peel to it. I'm going to try using clear spray paint on a stock Mag reflector and report my findings. I have a FiveMega MOP reflector that I'm going to enlarge the hole. It might be easier to buy the KaiDomain 15mm opening MOP reflector though.

Sputtering the reflector helps but you will not completely smooth the beam out. The MOP reflector will work about the same as the Stock plastic reflector sputtered. A medium stipple reflector smooths the beam out perfectly on the P7.

Mac

 

[ace0001a]

Sputtering the reflector helps but you will not completely smooth the beam out. The MOP reflector will work about the same as the Stock plastic reflector sputtered. A medium stipple reflector smooths the beam out perfectly on the P7.

Mac

Good piece of info to know Mac, but I would've thought an Orange Peel reflector would still offer a smoother beam than the stock one. I've tried the heavily sputtered reflector I got from Malkoff Devices and it did totally smooth out the beam, but how much output was compromised I don't know...it was still really bright. KD seems to be out of their MOP reflectors at the moment and I've been told they're working a reflector that will work well for the P7. The Sandwich Shoppe has some Modamag Camless reflectors in stock and they offer hole enlargement...but for a quantity flashaholic myself, $30 is alot to drop for a single reflector...and considering I'd want more than one. So that leaves sputtering your own reflector to be the most economical.

As for a P7 O-sink, I believe H22A is going to have a P7 DHS heatsink available next month. The one I did, I just modified and used a standard DHS heatsink and it worked just fine...the heat transfers well to the Mag body and to the touch, only gets warm and not hot.

 

[ambientmind]

I will look into it. In this day and age of cheap Chinese flashlights coming out left right and center it is difficult to make a product which can compete. Lucky for us they don't concentrate their efforts on accessories and modification parts. What is crazy is that places like DX have complete P7 lights available for sale for under $50.00!!
The idea of modding a mag to accept a P7 is not new. Several people have already done it.
What may end up being new is the heat sink used to to the mod.
I have found that the space inside the Mag body behind the space that is created when a O-Sink is mounted is Huge!! There is plenty of room in there for more aluminum in the form of an extended sink and a driver. IMHO the driver should be mounted into the sink as well so that heat generated by the driver may be efficiently moved.

My fear is that there is simply TOO MUCH heat generated by the P7 to efficiently move it using a passive heat sinking system.
These emitters make a lot of heat. even with the mass of the head being thermally well attached to the body of the light there is a lot of heat to move.
The idea of three P7's running at full power ..I see a light which may not be usable for more than a few minutes...


My focus is on creating a sink which will allow a P7 to be mounted inside a Mag host for it to move heat efficiently enough to allow for continuous use of the light.... Maybe copper? I hate using copper because it is so yummy to machine....then then there is the issue of cost...copper is pricey especially if you choose to use the tellurium copper. I would be buying 8 foot lengths of material. Buying less is silly due to the machine shop minimums...
1 1/2 inch 8 foot length of 7075 costs $130.00
6061 is about 1/2 the cost of 7075...
the same thing in copper costs over $900.00

A copper O-sink will cost 7 or 8 times as much as an aluminum unit. I material cost alone. WOW!!

So anyway I think I will focus on 6061 for the sake of those of you who are budget minded. IN the past an o-sink cost the end user $10-$15.00 if memory serves me. I suspect that material cost have gone up since I made those as well as the cost of machine time ...everything has gone up.
Research is needed. I am unwilling to do a lot of work on a product with an extremely limited market.
Unfortunately on the surface of things it looks to me as though I may not be making these. there is good size investment required and the market for this part is not what I would call booming.
If you guys have ideas which can help make this idea become a reality please share them with me.
As I said I will do the necessary research to learn about the feasibility of mounting the P7 onto a hunk of aluminum in a Mag host.
Any info you guys feel like sharing regarding this please post.
Yaesumofo

the only thing i can think of to make this simple and cost effective is to make a heatsink similar to H22A's but instead of milling out the inside of the bottom, leave it solid aluminum to add mass. Also, you could probably make it about 1/2" longer to add even more mass and surface area to move the heat into the body of the light. I was also thinking you could coat the aluminum heatsink with a thin layer of copper, if thats even possible. That might further help the transfer of heat between the led to heatsink, and heatsink to body. But that would totally depend on cost and if its something that can even be done. That step could obviously be skipped. Just my $.02.

 

[CM]

..I understand the P7 runs HOT and needs great heatsinking. I could design the sink to extend deeper into the body adding mass to the sink...

I'm really glad someone is paying attention to the heat issue. I've seen so many posts from people who want to put the P7 into a small 2x123 light, drive it at spec, and expect it to perform.

Whatever you do, please make sure you leave plenty of room for a regulator. I'm not talking about those cheap DX or KD linear regulator, but proper switcher(s). For example, using three Downboys in parallel.

It's a pity that the LED's anode is not isolated. If it was, I would probably remove the anodizing from the inside wall of the M@g so as to have a better metal to metal contact. With that and an extended heat sink, one can drive the LED to spec and have the heat properly dissipated onto the body.

 

[2xTrinity]

Regarding the copper sink... I'm no physicist, but I wonder if it would actually help significantly to use a copper sink. While the heat would conduct away from the slug faster, it would still encounter the barrier of the aluminum body which would have to then conduct through body's threads onto the head's thread then though the head itself to its surface. Not to mention that there is a coating of anodizing through each of the surfaces. Certainly a copper sink would help, but the question is, by how much? And considering that the cost would be nearly 10 times more, would it be worth it?

In the case of the interior surfaces, It might be beneficial to sand away the anodizing, then polish the sanded bare-aluminum surface, then glue the heatsink directly to the aluminum with epoxy. One less interface to worry about compared to gluing the heatsink to the anodize.

As far as the exterior surface, in that case, the anodizing could actually help as anodized aluminum is an almost ideal radiator, while bare aluminum is pretty terrible (good reflector = bad radiator). At the temps the flashlight will probably reach, radiation could well be a significant method of heat dissipation.

Another variable I haven't seen mentioned so far is that copper has about 50% more heat capacity than the same volume of aluminum (See Specific Heat Capacity), meaning it can "soak up" more energy before increasing in temperature. Taking that into consideration, brass may be another viable option as it has almost the same heat capacity as 100% copper. The greater heat capacity of the brass would almost certainly be cancelled out by the poorer conductivity, though.

It's a pity that the LED's anode is not isolated. If it was, I would probably remove the anodizing from the inside wall of the M@g so as to have a better metal to metal contact. With that and an extended heat sink, one can drive the LED to spec and have the heat properly dissipated onto the body.

Just insert your batteries so as to have a positive ground. Negative ground is just a convention, not a law. A custom rechargeable pack might be a good idea to prevent someone frying your light by trying to put the batteries in the "right way" though.

 

[darkzero]

Sputtering the reflector helps but you will not completely smooth the beam out. The MOP reflector will work about the same as the Stock plastic reflector sputtered. A medium stipple reflector smooths the beam out perfectly on the P7.

Mac

If you sputter it enough it will smooth out the P7 pretty good. I have some left overs that I sputtered too much which I did not like how the beam was with LuxIIIs. But they work good for the P7.

I've sputtered so many Mag reflectors in the past that I've learned how to sputter them frim light to heavy.

Old threads/pics here:
http://www.candlepowerforums.com/vb/showthread.php?p=492893#post492893
http://www.candlepowerforums.com/vb/showthread.php?t=43577&highlight=sputtered