See zen.htm for pics. Note this sink was designed to avoid any bending of sheet/plate/bar - only drilling/cutting/filing to suit tools available. The metal came from local stockist - no problems there - worth paying them to cut (proper guillotine - nice clean cut) to the most used dimension. The sinks for the 4 Zen gain stages used virtually a whole 8*4 ft sheet of 3mm cut into 250mm strips - the height of the sink fins. Cut to length by trusty jig saw (wood blade works best, metal 'hacksaw' blade is too fine for soft Al, takes all day to cut it). Same with the 12.5mm plate. Purchasing the metal was the most expensive bit of the project.

The sink is fabricated from 3mm Al sheet, and 12.5mm Al plate. The plate carries the MOSFETs and gets the heat away quickly. Seems to work a charm. SIL-PAD K10 or SIL-PAD SP2000 plastic insulating washers, (no grease required) from Bergquist, under the MOSFETs. These claim and seem to achieve (despite claims to the contrary on the web) 0.2deg/watt with a TO3 case. I confirmed this figure. That combined with the MOSFETs 0.5deg/watt from junc to case keeps the sinking requirements in hand - though looking at my sink you wouldn't know it! Hopefully they will last (UPDATE: May 2013: They have! ) - I am mindful of Nelson's comment on MOSFET life expectancy and junction temps.

Clamped either side of the main plate are alternately 9mm thick (25mm wide) Al spacer bars, and 3mm sheets (the fins). It's all clamped up by (2 rows of 3 - see pics) long M8 steel studs, and all thermally mating surfaces smoothed with 320 grade Al oxide paper wrapped around a =flat= bit of scrap Al bar, then smeared carefully (not too thick) with standard white thermal grease, and clamped up as tight as I dare. Good scientific approach eh?

All other surfaces except the spacers were painted with Blackboard black. I considered black spray - some experiments are required to see if the spray was too thick (i.e. it starts =insulating= the fin), but it would be a lot quicker. The black stuff went on better watered down, with 2 coats, and the Al rubbed with wire wool and degreased first.

Well, although this sink seems to work very well, I don't recommend it simply on the grounds of the metal work involved. I guess if you have access to machines it would be better - I used a jig-saw, large/small file and a small but adequate pillar drill. A certain mount of cutting avoided by getting the metal stockist to pre-cut to main dimensions. As you see, the sink is the amp, i.e. it is not enclosed, therefore additional finishing (filing!!) was required to make it presentable and safe. The pics show a wooden cradle I made to hold the two units adjacent. The other channel will have a smart Al sheet version of this as designed by my wife. There - how to get buy-in for these crazy projects!

Performance figures: 0.17deg/watt with no fan, 0.11 with the fan. A lot lower if the fan was moving a little faster - I really have it running slowly. The tops of the main plates run warm to the touch, no more - a slightly warm breeze comes off the whole thing when looking down on it. Each semi circular sink weighs about 11Kg. Phew!