1 MC-E vs multiple XR-E which needs more heatsink?

[lightime]

If you have a single MC-E vs an array of 2, 3 or 4 XR-E Q5 running at the same mA, lets say 1000mA... which needs more heatsink the single MC-E or the team of XR-E?

Would there be any difference at other mA levels at all or would one option need consistently more heatsink?

Thanks in advance for any guidance.

 

[yellow]

I d think the quad needs more sinking, because the heat comes from just a single spot, while the four leds are spread over the mounting sheet.
... no matter what current.

 

[PaulH]

I agree with Yellow.

I have a 4 x XR-E (R2) set-up, running at around 1000mA per LED (with one of Download's multi-sinks). This seems to run on high quite a bit cooler than 1 x MC-E at 2800mA (700mA per die) - with a brightlumens heatsink.

As for efficiency, then that is going to be a slightly different matter.

 

[mds82]

which ever one has more power going to it will require more heatsinking.

4 XR-E @ 1000ma = 4000ma total
1 MC-E @ 2800ma = 2800ma totla

the 4000ma will require more heatsinking.

if its the same ma total for both, the heatsinking will be virtually the same.

 

[2xTrinity]

which ever one has more power going to it will require more heatsinking.

4 XR-E @ 1000ma = 4000ma total
1 MC-E @ 2800ma = 2800ma totla

the 4000ma will require more heatsinking.

lighttime specified the same current would be flowing through each. That is, 4 XR-E @ 700mA, or 1 MC-E @ 2800mA. Also, assuming similar Vf for each (that is, same total power consumption)

The MC-E will require better heatsinking all else equal because the power is concnetrated in one spot. This is the very reason why the MC-E is only rated for 2800mA total, NOT 4000mA (like 4 XR-Es)

 

[lightime]

This is the way I was thinking.... assuming same current through each. I thought the mc-e would require less heatsink the higher you went up in current where you'd be pushing the xre-'s harder. But say at 350 mA and 500 mA for example where they are all still pretty much well within specs what would the heatsink requirements be then...who would need more heatsink a lone mc-e or a couple (2+) xre's? Oh, and the team of xre's would be sharing a heatsink.

lighttime specified the same current would be flowing through each. That is, 4 XR-E @ 700mA, or 1 MC-E @ 2800mA. Also, assuming similar Vf for each (that is, same total power consumption)

The MC-E will require better heatsinking all else equal because the power is concnetrated in one spot. This is the very reason why the MC-E is only rated for 2800mA total, NOT 4000mA (like 4 XR-Es)

 

[phalanx]

This is the way I was thinking.... assuming same current through each. I thought the mc-e would require less heatsink the higher you went up in current where you'd be pushing the xre-'s harder. But say at 350 mA and 500 mA for example where they are all still pretty much well within specs what would the heatsink requirements be then...who would need more heatsink a lone mc-e or a couple (2+) xre's? Oh, and the team of xre's would be sharing a heatsink.

Heat transfer is partially a function of the area of the physical interface between two materials. The larger your surface contact area, the more heat (in Watts) you can transfer. If total power between multiple XREs and a MCE are the same, you will be able to transfer heat at a higher rate from the multiple XREs due to the larger overall contact area. This means your XREs will be cooler than the MCE.

 

[Marduke]

And total light output will be higher due to the lower junction temp.

 

[ergotelis]

Both need the best possible heatsinking.But because of the fact that in a small area mce is puting out a lot of heat, it needs better heatsinking. Of course,4 cree need too, but the fact that the 4 seperate dies are not that close, it helps the heat to be spread out better. If you have to choose between MCE and 4 cree, i would go with 4 cree. You won't have to care at all about donut holes and output will be a bit more.
Also,in a future update of your leds, i guess that a SST90 would not be as much efficient as 4 XP-G!

 

[lightime]

Looks like everyone pretty much agrees. Makes sense to me after reading your posts and thinking about it a bit more. Actually sounds more like common sense now.

I thought the MC-E handled a lot more power than the "poor little" 4 XRE's because In my mind I was thinking the MC-E handled 2,800 while the XRE handled "only" 1000 so the MC-E had more headroom....but I overlooked the fact like 2xTrinity mentioned that we are talking 4 x 1000 with the XRE!! So it's actually 2 strikes against the MC-E when up against a set of 4 XRE:

1) each corner handles max 700mA (2800max vs 4000max)
2) Not as much area to transfer heat with the single MC-E

Thanks for clearing all this up guys!! :twothumbs

 

[Gryloc]

Hey there again! So, yes 4 XE-E emitters are better than 1 equally matched MC-E. I just wanted to poke in to tie this to your other thread, lightime. I mentioned the MC-E solely because it would be more efficient to run MC-E at the required power to produce, say 150 lumens, than a single XR-E. This is because the current is split by 4 so the forward voltage of each die would be slightly lower.

In your above scenario, you have to consider that you are dumping more than 10W of energy into either a single MC-E or four XR-E emitters, even if they are mounted to a large heatsink (1A per emitter or die means around 14W total according to jtr1962's measurements). At these higher power levels, then the way that heat is concentrated means everything.

However, if you consider the scenario in your other thread where the tiny heatsink/heatspreader (1" dia by 0.5" deep) can barely handle more than 3W, then the effect of dumping 3W into a MC-E will have little difference compared to doing the same with 4 XR-E emitters (the effect is not as dramatic). The heatspreader will struggle with dissipating nearly 3W of heat regardless to how the heat is concentrated by the emitter(s). Either of the two types of emitters will be running hot at 3W mounted on a heatspreader that is nearly saturated with heat.

Sorry that I am tying these threads together, but I just want you to make sure you understand the difference. If you are thinking about another project idea, then ignore this post.

-Tony

 

[yellow]

while it is now clear, that the basic question has the 4 XR-E "win",
lets look at why the MC-E/P7 is used mainly now:

cost, difficulty of production, size
with just one quad, You save mounting+focusing of three additional led.

You have to find Your tradeoff!
(f.e. with a nice, thick aluminium mounting plate and a smaller housing, the quad also works nicely)

 

[saabluster]

while it is now clear, that the basic question has the 4 XR-E "win",

Sorry but I disagree. The answer to the OP's question "1 MC-E vs multiple XR-E which needs more heatsink?" is XR-E. Even if we assume there are 4 XR-Es to the one MC-E(best case senario) the MC-E still wins. If the exact amount of power is going to all the dies the MC-E dies due to the current density will be running hotter and therefore their Vf will be lower and it will be more efficient from a power standpoint. Therefore it will make ever so slightly less waste heat. That means the XR-Es needs more heatsink. However if the question had been which needs a better heatsink than the win would go to the XR-Es.

 

[lightime]

Tony, thanks for chiming in again...you have been very helpful and I appreciate it very much. I am still going with what we decided in the other thread for that application. Now I am just trying to figure out other applications where I can use this type of emitter. By the time I'm done I will probably have them all over my house, shop, and cars!!!!!!!!!!

So the other application in the car with the space constraints will still be using the single XRE we talked about, most likely with 700 or less mA. BTW, I would still be interested in your input on the other thread regarding the optimum vs easiest way (I know there will be some trade off) to drive that single XRE in the other thread.

Hey there again! So, yes 4 XE-E emitters are better than 1 equally matched MC-E. I just wanted to poke in to tie this to your other thread, lightime. I mentioned the MC-E solely because it would be more efficient to run MC-E at the required power to produce, say 150 lumens, than a single XR-E. This is because the current is split by 4 so the forward voltage of each die would be slightly lower.

In your above scenario, you have to consider that you are dumping more than 10W of energy into either a single MC-E or four XR-E emitters, even if they are mounted to a large heatsink (1A per emitter or die means around 14W total according to jtr1962's measurements). At these higher power levels, then the way that heat is concentrated means everything.

However, if you consider the scenario in your other thread where the tiny heatsink/heatspreader (1" dia by 0.5" deep) can barely handle more than 3W, then the effect of dumping 3W into a MC-E will have little difference compared to doing the same with 4 XR-E emitters (the effect is not as dramatic). The heatspreader will struggle with dissipating nearly 3W of heat regardless to how the heat is concentrated by the emitter(s). Either of the two types of emitters will be running hot at 3W mounted on a heatspreader that is nearly saturated with heat.

Sorry that I am tying these threads together, but I just want you to make sure you understand the difference. If you are thinking about another project idea, then ignore this post.

-Tony

 

[John_Galt]

This has been a question I have been pondering over for my various designs for bike lights, etc. How does one calculate how much heat dissipation you need from various sources? And is there an equation of the amount of surface area needed to dissipate a certain level of heat? Please give me more time, I have more questions, and have to switch computers momentarily...

 

[Marduke]

This has been a question I have been pondering over for my various designs for bike lights, etc. How does one calculate how much heat dissipation you need from various sources? And is there an equation of the amount of surface area needed to dissipate a certain level of heat? Please give me more time, I have more questions, and have to switch computers momentarily...

No real easy formulas since every problem is unique, but you can start with a free online thermodynamics book of your choosing.

 

[SemiMan]

Lot's of guessing and NO ONE reading the data sheets!

MC-E: 3C/Watt

XR-E: 8C/Watt

MC-E has 4 die, so that 3C/watt equates to 12C/watt per die. If you are running each die at 2 watts, then each die is going to run 8celsius higher.


In terms of basic heat transfer, 4 XRE will also have an advantage assuming they are spaced out. The 4 XRE will have greater cross-section to spread heat into the heat sink. If the heat sink is highly conductive, i.e. thick copper, then the advantage may not be high. However practically, I could see another few C/watt advantage for the 4 XRE.

Easy advantage to 4 XRE. The biggest advantage is just at the package though.

Semiman

 

[saabluster]

Lot's of guessing and NO ONE reading the data sheets!

I wasn't guessing. If you check the data sheets you will see that the Vf of the MC-E is already lower than the XR-E. Now factor in the higher die temps of the MC-E which reduces the Vf even more and you will find that it takes less energy to power the MC-E than the XR-Es as the MC-E is more energy efficient. Less energy in= less heat out. Now go back and reread the OP's question. The answer to that question is XR-E not MC-E.

 

[yellow]

the majority of ppl not necessarily must be right,
but in general it is



while besically there might be a tiny - measureable - difference speaking for the MC-E, it remains that its four dies are backed on such a small space, that the thermal plate MUST be considerably thicker to get the heat away there, while the single emitter ones are spread and nearer to the housing.
Do You build/mod You led lights? (as most other posters do?)

 

[saabluster]

the majority of ppl not necessarily must be right,
but in general it is



while besically there might be a tiny - measureable - difference speaking for the MC-E, it remains that its four dies are backed on such a small space, that the thermal plate MUST be considerably thicker to get the heat away there, while the single emitter ones are spread and nearer to the housing.

I am being rather pedantic I know but the OP's question was about which needs more. Everyone has given answers based on what would be a much better question. "Which one needs better heatsinking?" That is what should have been asked.

Do You build/mod You led lights? (as most other posters do?)

I am OMG Lumens. See my sig line.