Heatsink - Plain old copper VS Silver plated copper. / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Heatsink - Plain old copper VS Silver plated copper.

previous 1 2

Move Thread

LAN_403

Adam Munich

Fri Sept 17 2010, 02:47AM

Registered Member #2893 Joined: Tue Jun 01 2010, 09:25PM
Location: Cali-forn. i. a.
Posts: 2242

IntraWinding wrote ...

There's also a similar problem with coating a heatsink in an emissive (black) material. You need to be sure the improvement in radiation outweighs the insulating effect of the added layer.

Quoted from the anodization paper.
"Surface emissivity limits the amount of heat transfer due to radiational cooling. With 1.0 being perfect (black body) emissivity anodized aluminum is 0.85 and unfinished is 0.05"

dmg

Fri Sept 17 2010, 03:53AM

Registered Member #2628 Joined: Fri Jan 15 2010, 12:23AM
Location:
Posts: 627

Well, I have a few small heatsinks that have that black anodized finish,
however I dont have anything bigger then those tiny ones made for TO-220 devices,

so in this case, anodization is not one of my options.

but I think Ill just sand away the layer of plating and just use the bare copper for the devices. (Ill also attach said copper plate to the large AL heatsink I mentioned earlier).

here is a picture of the plates, and the AL heatsink.

1284695573 2628 FT96649 Copperplates 001

Adam Munich

Fri Sept 17 2010, 04:06AM

Registered Member #2893 Joined: Tue Jun 01 2010, 09:25PM
Location: Cali-forn. i. a.
Posts: 2242

Why not anodize them yourself? You may be able to substitute H2SO4 with HCl if you can't find any sulfuric.

The color of the oxide really doesn't matter that much. It's the oxide itself that enhances heat transfer. Black is just a plus.

radiotech

Fri Sept 17 2010, 06:58AM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

If you sand it off you may never get as smooth a surface so any benefit may be lost due to the uneven surface which still may contain silver-copper alloy, if it is silver. If it isnt pure silver, it looks kind of bright for old silver, it may be something else you dont want to inhale the dust of, let alone what you are sanding it with. If you somehow could transfer that silver to a copper wire, say #6 guage, think of how much better it would work as a high current coil working at several hundred kHz.
Perhaps if you asked in the Chem thread someone would know how to change that plating to AgNO3, which is nifty stuff.

IntraWinding

Fri Sept 17 2010, 09:06AM

Registered Member #2261 Joined: Mon Aug 03 2009, 01:19AM
Location: London, UK
Posts: 581

I practice just use the aluminium heatsink in your photo, or a piece cut from it. It will dissipate heat better than those low surface area copper plates.

HV Enthusiast

Fri Sept 17 2010, 12:29PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

If its a coating (which is probably microns thick) its not going to do anything. Remember, heat transfer is a function of the thermal impedance of a material. If the material is as thin as a electroplated coating, its thermal impedance is going to be extremely low, regardless of what material it is. Even if it was plastic of the same thickness, its not going to affect transfer much.

As Bjorn stated, concentrate more on the thermal interface between device and heatsink. Ensure you get good contact and use a material (i.e. thermal grease or grafoil) to ensure all gaps are filled the device mates well with the heatsink.

Marko

Fri Sept 17 2010, 05:53PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Gatedbreakdown wrote ...

Well, I have a few small heatsinks that have that black anodized finish,
however I dont have anything bigger then those tiny ones made for TO-220 devices,

so in this case, anodization is not one of my options.

but I think Ill just sand away the layer of plating and just use the bare copper for the devices. (Ill also attach said copper plate to the large AL heatsink I mentioned earlier).

here is a picture of the plates, and the AL heatsink.

Hi, a few comments on this:

Firstly, I don't think radiation has any effect on the dissipation that you can't emulate by just adding a slightly bigger fan. Also, the heatsink in the picture looks more than big enough for four ISOTOP devices at any sane power dissipation level - how much power do you intend to run after all?

Do some math - junction-case thermal resistance of ISOTOP devices is usually about 0.4K/W. If your case to ambient thermal resistance is already much lower than this, beefing up the heatsink further will have very little effect on die temperature. A better CPU heatsinks have like 0.2K/W - So unless you're really pushing the limits of the devices, I'm not sure how much can you gain by using your copper heat spreader. If you still want it, don't polish the coating, it's there to preven copper oxidation and it does no harm to you.

Marko

IamSmooth

Fri Sept 17 2010, 07:16PM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

If you want to get rid of a lot of heat you can do what I have done with my inverter's mosfet heatsinks. I bought some 1/2" aluminum stock from McMaster's in NJ (any supply will do). I drilled a hole through it and threaded a nipple on each end. I tapped screw holes and mounted it to the heatsinks and ran cold water through them. They run ice cold while I'm pumping over 50A through them.

If you are interested I can upload photos. It works very well. It is the same setup as the water-cooled Celem capacitors.

Bored Chemist

Sat Sept 18 2010, 09:12AM

Registered Member #193 Joined: Fri Feb 17 2006, 07:04AM
Location: sheffield
Posts: 1022

While it's true that a blackened surface radiates more heat than a silver one you have to remember that most heat is lost by conduction to the air rather than my radiation. That's why heatsinks are finned.
Silver is only a marginally better conductor than copper and the interface layer between the two metals may well act as a barrier to heat transfer.

previous 1 2

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob