Where to buy heatsink for power amp?

Status
Not open for further replies.
To my knowledge Fischer is the only manufacturer issuing tight specs for surface planarity
and roughness....and that makes their stuff rather expensive. To produce aluminum profiles plane and with minimum surface roughness is not so easy as it meets the eye , and there is another issue of quality: shrinkage voids , these are only to be seen in special tests with sort of X rays or ultrasonic. But can lead to local thermal resistance far above the anticipated value.
Insofar most cheap heatsinks aren't worth their price. The best material for heatsinks
is solid gold. Not inexpensive either, but has the best theory of thermal conductivity!
 
To my knowledge Fischer is the only manufacturer issuing tight specs for surface planarity
and roughness....and that makes their stuff rather expensive. To produce aluminum profiles plane and with minimum surface roughness is not so easy as it meets the eye , and there is another issue of quality: shrinkage voids , these are only to be seen in special tests with sort of X rays or ultrasonic. But can lead to local thermal resistance far above the anticipated value.
Insofar most cheap heatsinks aren't worth their price. The best material for heatsinks
is solid gold. Not inexpensive either, but has the best theory of thermal conductivity!

Very interesting!!
But no...I am not going to have gold heat sink....at least till I don't have a bank account like bill gates.....ahahahahaha!!

Still I am wondering where I can buy these heat sink in the states SK538/539.
I don't mind if it is a bit more expensive as long as it doesn't cost like solid gold!! 😱😛
 
Still I am wondering where I can buy these heat sink in the states SK538/539.
I don't mind if it is a bit more expensive as long as it doesn't cost like solid gold!! 😱😛

Try this Fischer distributor

Component Trends
27142 Burbank
USA - Foothill Ranch, CA 92610

Phone: (949) 472-9050 x275
Fax: (949) 472-9010
Mobile: (949) 275-2761

asappctrends.com
CTrends - Supply Chain Specialists

your contact person:
Mr. Alex Sapp
 
I recently got about a dozen pieces from this company.

4 pieces were large-ish, 10 inches by 4 inches. The fee to cut them to order for all of those was only $5 (!!!). As mentioned above, not "finished", so the edges of the cut are very (VERY) sharp, and there were a couple uneven bits at the end of the cut (perhaps that was what was referred to in the post above). If doing this by hand, you just need a bit of patience, and some sandpaper to take the sharp edge off. (they do offer custom machining, but I have not tried it)

All perfectly flat, measured with some kind of shop tool while they were being drilled/tapped and machined for me. No issues whatsoever. Exactly as advertised, and shipped very quickly (to me, anyway).

Very happy with the result at the price!

Sharp edges and burrs can easily be addressed with files and emory cloth. Local high spots can be dressed with 120-150 grit emory cloth on a long (10") board sander. Aluminum is very soft and very easy to sand.

The prices seem too good to be true. A rough finished heatsink is no problem for me because I have a metal shop with every abrasive known to man. (I have more sandpaper in stock than your local Menards. 😀 ) I could even polish them up with jewler's rouge until they looked like chrome if I wanted. 😀

I would not be happy with a 25 cm heatsink that stood proud 2.5 cm, though. I would request a replacement. But with these prices, it's hardly worth scrounging for old heatsinks.

By the way I do make some modest heatsinks in my shop. I have various aluminum stock on hand (U-channel is perfect and you can buy 3 feet of quarter inch thick at Menards for around $10). I can make a small 3-4 degree/watt heatsink (like for a TO-220 driver) with around 25 cents worth of material.
 
that is bad. How did heatsink USA react when you told them?You should not have to mill them. Mechanically they should line up with the other chassis components.

Thermally, the curve has almost no effect.

They were cheap enough and quite suitable for what I was doing, so I did not raise the issue.

I actually just did the trouble to unpack the (currently unused) amp and separate the covers and heat sinks. Memory failed me. The curvature is 0.69mm, measured with a straight edge and feeler gauge. The 2.5mm was from memory and a 'visual approximation' when placing TWO sinks back-to-back. Apologies to HeatsinkUSA.

If you are willing and equipped to spend a little time on them, I still think they are well worth the money. The concave issue seems to be 'luck-of-the-draw", depending on which profile your sinks are cut from (they typically come in something like 8 feet lengths). As Andrew mentions, it does not affect performance. It may affect how well the enclosure goes together, depending on the design. In my case the top and bottom covers overlaps the sinks slightly, to the curvature is not visible. Finishing the cut edges and burrs are trivial with 220 and 400 grit sandpaper. I did not try painting or have having them anodized.
Photo attached.

At the time my first choice was Conrad, but they were late in responding to my inquiry and I went with HSUSA.
This was the reply from Conrad in 2010 re. shipping to Canada, if it helps:

> Delivery 1.75 kg (Aus. Post Sea Freight 10 - 12 weeks) 23.35 (22.92)
> Delivery (Aus. Post Standard Air 7-12 working days) 37.85 (37.15)
> Payment options: credit card (Visa, Mastercard),
> Paypal (in AUD to account
> sales@conradheatsinks.com )
 

Attachments

  • sIMG_4401.JPG
    sIMG_4401.JPG
    87.8 KB · Views: 598
...Photo attached.
That is a work of art, not an amplifier (chuckle)... are you hiding the real wiring under the chassis, like all good tube amps? 😉

.69 is a far cry from 2.5. In the past, I managed to fabricate a .5 mm bulge just from a bad job drilling and tapping holes... 😱 (main reason why I use real shop tools and good quality taps to make those pesky 3-mm holes now)

My heat sinks from this company were checked with the same tool used to check engine blocks for flatness before and after they were drilled. No issues were found on mine, but I guess I will have to keep an eye out in the future.
 
That is a work of art, not an amplifier (chuckle)... are you hiding the real wiring under the chassis, like all good tube amps? 😉

Nope, you can see all the wires in that picture. It is a simple ESP project P3A clone, so not much wiring involved.

Here are pictures before the heat sink ends were sanded and the fin tips rounded:

http://www.diyaudio.com/forums/soli...ur-solid-state-pics-here-212.html#post2805436

On the subject of flatness, I dug up a spare set of Modushop 300mm heat sinks. Putting them back-to-back they are totally flat.
 
...another HSUSA profile.

This is a pair of the smaller HSUSA profiles that was intended for a non-audio application, but never used.

This set is slightly convex...there is a ~0.3mm gap at the ends with the centers touching.
The cutting marks and burrs are also visible.

If you can work around or fix these minor issues (this is DIY-audio after all), they are still great value for money.
 

Attachments

  • sIMG_0092_cr.jpg
    sIMG_0092_cr.jpg
    105.8 KB · Views: 541
They were cheap enough and quite suitable for what I was doing, so I did not raise the issue.

I actually just did the trouble to unpack the (currently unused) amp and separate the covers and heat sinks. Memory failed me. The curvature is 0.69mm, measured with a straight edge and feeler gauge. The 2.5mm was from memory and a 'visual approximation' when placing TWO sinks back-to-back. Apologies to HeatsinkUSA.

If you are willing and equipped to spend a little time on them, I still think they are well worth the money. The concave issue seems to be 'luck-of-the-draw", depending on which profile your sinks are cut from (they typically come in something like 8 feet lengths). As Andrew mentions, it does not affect performance. It may affect how well the enclosure goes together, depending on the design. In my case the top and bottom covers overlaps the sinks slightly, to the curvature is not visible. Finishing the cut edges and burrs are trivial with 220 and 400 grit sandpaper. I did not try painting or have having them anodized.
Photo attached.

At the time my first choice was Conrad, but they were late in responding to my inquiry and I went with HSUSA.
This was the reply from Conrad in 2010 re. shipping to Canada, if it helps:

> Delivery 1.75 kg (Aus. Post Sea Freight 10 - 12 weeks) 23.35 (22.92)
> Delivery (Aus. Post Standard Air 7-12 working days) 37.85 (37.15)
> Payment options: credit card (Visa, Mastercard),
> Paypal (in AUD to account
> sales@conradheatsinks.com )

Very nice chassis.
Did you have a shop prepare front and rear panels?
 
Nope, you can see all the wires in that picture. It is a simple ESP project P3A clone, so not much wiring involved.

Here are pictures before the heat sink ends were sanded and the fin tips rounded:

http://www.diyaudio.com/forums/soli...ur-solid-state-pics-here-212.html#post2805436

On the subject of flatness, I dug up a spare set of Modushop 300mm heat sinks. Putting them back-to-back they are totally flat.
Nice job! I've been testing/listening to my P3A for the last few weeks... (somewhat inspired by Sakis) Both channels mounted on one 10x4 heat sink from HeatSink USA for now (LOL).
 
Very nice chassis.
Did you have a shop prepare front and rear panels?

No.
The front and rear panels are made from 100mm x 6mm aluminum flat bar.
I used a drill press, vertical belt sander, files (for the IEC socket hole) and sandpaper.
Being not painted or anodized the panels tend to scuff very easily, bare aluminum is very soft. They also attract fingerprints.
Some people use clear acrylic spray paint or similar for protection, I did not try that yet.

This one is the same idea, but using the Modushop heat sinks:

http://www.diyaudio.com/forums/soli...ur-solid-state-pics-here-209.html#post2795146
 
Last edited:
Afaik, curvature in extruded parts is very common... difficult to eliminate, especially in the shapes used for heatsinks - they are not symmetrical, so they cool at different rates on each side.

One can expect to finish the flat side for flatness as a matter of course, IF that makes a difference in a given application.

When parts mate, it's always a good idea to reduce surface roughness down to a micron level... which sounds exotic and difficult, but it's just a way of saying lap and polish. If it is polished there is no surface roughness. Absolute flatness is more difficult to achieve but not terribly difficult. One would use the same sort of techniques that one might employ on an automobile head for example. It's just a matter of sequential application of abrasives using finer and finer grits. The key to flatness is (for DIY, non machine methods) is to have a backer for the grit material that is already "PDF" (pretty darn flat). So that could be a thick glass (which is PDF) or a bit of machinist stock (for hand held uses). You can either move the grit on the work, or you can move the work on the block (like rubbing the back of the heatsink on a sheet of abrasive paper held to a plate glass...)

For something that is out quite a bit, a run on a flycutter bit on a milling machine saves a lot of elbow grease... then you can start with a pretty fine grit to get it dead flat.

There is also ripple from the extrusion process itself found on the surfaces of heatsinks quite often.

I think most heatsink OEMs have a spec for the mounting surface. Of course flatter costs more, because they have to post machine it.

_-_-bear
 
Last edited:
Reminds me of the time I got a great deal on a big heatsink on ebay. Didn't figure out it was in New Zealand until after I bought it. Cost double the price of the heatsink to ship it. Argh.

I've always got my eye out for junked amps and have a stash of unrepairable monsters in the basement to use as a base for the next amp project. Having a buddy at a stereo shop helps snag stuff that might get dumpstered.

I've got a Elliot P101 amp going into a unfixable Luxman Class A M-300 which is a large and heavy chassis and only put out 40W/channel in Class A mode. It's near the limit of what I'd consider generally "liftable". Your 100W/channel is going to be proportionately larger (don't want to think about 200W/channel in one chassis - forkliftable?) I'd look at doing monoblocks and adding "silent" cooling fans to help keep the size down.
 
Status
Not open for further replies.