hollow heat sink source

Hi all,

I am looking for a source of small quantities of hollow aluminum heat sink extrusions, similar to what is shown at this page:

formatting link

Has anyone come across anything like this? Basically, a ~3 inch wide by 1-3 inch high hollow rectangle/square with fins/separators inside. I need this for a low-budget prototype I am building.

everything I could easily find were open heat sinks. I do need the closed design.

Reply to
runcyclexcski
Loading thread data ...

Why not buy from them? They are a good supplier, providing much of our extrusions at competitive prices... Need only a sample? Contact them, and they may be willing to send you a small leftover end piece for the cost of shipping.

Reply to
PeterD

Peter - sounds like you know the company. I wrote them for a quote, and I can't wait to hear back, that's why I was hoping that some online store sells smth similar.

I am usually disappointed when one can not purchase products from a company's web site - one has to email for a quote, wait till one hears back. Plus, no prices were published.

Reply to
runcyclexcski

I"m not sure where you are, I use their local rep for New England customers (they are in NH...) The only problem I ever had was when their sales rep for my area died last year, that was awkward...

Prices are very fluid in this stuff, based on the market for aluminum mostly. They are oriented towards larger sales (they don't sell to end users for example) the stuff I get comes on a 60ft semi trailer...

Unfortunately I don't have anythign like you are looking for.

For anyone else...

We do have end cuttings (typically about 2 to 2.5 inches long of some other extrusions. They are free for the asking but you must pay postage (typically about $4.90 or so to US addresses. Right now these just go into the scrap box.

Reply to
PeterD

Thanks for letting me know about the end cuttings. Would you happen to have longer pieces, like, 12 in? I thought more about my, and I should be able to get away with "open" finned extrusions, since it's so hard to get exactly what I want.

I googled and googled, and I only found large-scale extrusion companies. I am in the Bay Area. Should I drive around and look into dumpsters in the Silicon Valley? :)

Reply to
runcyclexcski

Have you tried a hobby shop? Some of these folks have access to fun materials like that. You might be on the bigger end of what hobby extrusions look like but give it a try.

Reply to
a7yvm109gf5d1

Do you do custom extrusions, or machining? If so, please post a link or email me. We need stuff like this from time to time.

jjlarkin atsymbol highlandtecnology alltherest

John

Reply to
John Larkin

If you get desperate, consider bonding a flat plate to a regular parallel-fin heatsink with similar fin spacing to get the hollow section you need. If more desperate still, machine away some of the fins in a densely-spaced heatsink and proceed as above. Extrusion companies often have a minimum order that is in the range of several hundred pounds. The heatsink companies like Thermalloy will sample you if what you want is one of their standard products.

Paul Mathews

Reply to
Paul Mathews

perate, consider bonding a flat plate to a regular

this is what I was thinking.

Can you recommend a heat-conducting bonding material? One can't solder Aluminum, at least trivially.

Reply to
runcyclexcski

I was thinking to make my own heat sink by bonding plates together.

But I found out that heat conductivity of bonding materials is *much* poorer compared to solid metal. E.g. this adhesive claims to be 4.5W/ mK, while solid Aluminum is 221 W/mK. That's 50 times worse compared to a solid heat sink.

Are there better bonding adhesives?

Reply to
runcyclexcski

Allright, I guess I should have looked up Al Oxide, which is 30 W/mK - only ~7 times better than the adhesive I found.

Reply to
runcyclexcski

Our finished product is about five inches, so there would never be a piece longer than that... The extrusions we get are dimensioned to do this size, with a bit left over for errors, etc. The suppliers then add on an half an inch or so so that their pieces are not short.

Dumpster diving is not such a bad idea, as is salvaging 'em from old equipment too. When I do a one-up project, and nothing I have on hand is suitable, I usually just trash some old piece of junk...

Reply to
PeterD

No custom extrusions, but we cut, drill, tap and sometimes mill them as needed. Mostly cut, drill, tap, radius the holes, clean the edges, followed with a washing.

Reply to
PeterD

How about this beast:

ftp://66.117.156.8/Amp.jpg

The shiny things are solid copper, nickel plated heat spreaders. The transistors are un-insulated and the whole heatsink is signal-hot, bolted to the chassis through the delrin blocks at the top of the pic.

17 KW peak power output per axis into an MRI imaging gradient coil. Since it's imaging, we get to sell three boxes per system.

Some of our users want more output, so we may do a water-cooled version next. That should be interesting.

John

Reply to
John Larkin

Please note actual thermal resistance depends on:

1) cross sectional area (inversely proportional) 2) thickness of layer (linearly proportional)

Meaning: with a large cross sectional area and a thin bonding layer, thermal resistance is low. Get everything very flat, with the Al pieces in good contact, clamp if necessary. Use fillets of adhesive instead of a bonding layer. Whatever. Will NOT be a problem.

Also, adhesives filled with high thermal conductivity materials have higher thermal conductivity than non-filled materials. However, it won't matter if you follow above.

Paul Mathews

Reply to
Paul Mathews

As far as adhesives go, 4.5 W/mK is about as good as it gets. Just clean the parts well with alcohol and clamp them together while the adhesive is setting. Then heat transfer is through just a thin layer of adhesive material, and it will barely make a difference for the overall thermal resistance.

For example, if you had 1/4 square inch of total contact area and

0.002" thick adhesive, the adhesive would add to the thermal resistance:

(1 m K / 4.5 W) * (0.002/39.37)m / (0.25/39.37^2)m^2 = 0.07 K/W or 0.07 C/W

Regards,

Mark

Reply to
redbelly

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.