ok, ive come up with a new passive radiator design.not ideal, but practical
 

heres a picture of it, as viewed from above.

http://img.photobucket.com/albums/v2...veradiator.jpg

the orange thing is a 15mm copper pipe, the dark brown is aluminium, the black is bolts.

it will be about 40cm high and ill try keep each fin at a maximum length of 5cm.
(the fins will all be about 40cmx5cm)

(ive done a test before, clamping an aluminium plate to a copper pipe and the aluminium felt as hot as the copper pipe for about 5-10cm. that was 1.6mm alum)

there will be a whole lot of these stacked next to each other. then mounted to the side of my pc. (at least 4)

note, i will tighten the bolts quite a bit to squash the pipe a bit and give some good contact area.

assuming each pipe has 10 fins around it with 4cm on each fin exposed, 40cm high, thats 3200 square CMs.
per pipe.

yeah that's terrible

you'll have about 1% of the copper conducting with the alumnium

imagine putting the corner of a heatsink on a CPU die, how well would that work?

If you're going for cheap-plus-easy-to-build, with copper-only water contact (is that what you were trying to do?) why not just use finned "heat register" pipe?
Yes, the fins go the wrong way (built for horizontal pipe runs) so you'd either have a horizontal radiator or have to build some sort of ladder-and-rungs setup. And the register stuff is designed for a minimum 30C differential between liquid and room air, so you'd have to use a lot of it (as compared to the BTUs it's spec'ed to move from liquid to air).
Contra-wise, if you can live with water contacting aluminum (should be OK with corrosion inhibitors) then maybe the thing to do is to find an architectural aluminum extrusion that's pretty much shaped the way you need. You might start here.
I haven't done either of these things, so please take this as "thoughts" not something I know to work...

greenman: if you read my post you would see that i said i did an experiment where i clamped a pipe to an aluminium plate, and the plate felt just as hot as the pipe for about 5cm. thats without squashing the pipe at all

in this design the pipe would be squashed quite a bit, increasing contact area a lot.
so with the fin length im using, its not an issue at all.

bobkoure: pity about the heat register pipe fins being the wrong way.

im confident this will work well

i'll hopefully have it built within a week or 2

I don't know about that. There might be enough surface area, but there's not enough contact between the pipe and the fins to transfer heat efficiently. Have you considered soldering them to the pipe?

Keep in mind the Reserator is much more solidly built and it can't even keep system temps on par with stock aircooling.

it would make sense to wrap teh Al around the pipe maybe?

you could used some of the finned cylindrical transmissions coolers perhaps though they are not copper.

mold that fin structure around the pipe and you might have something that will at least conduct a minimal amount of heat.

Green is right, the picture shows a contact area of "nothing"
"Squashing the pipe or whatever aint gonna yield more than a 3 fold increase of what you have now, which would still be "nothing".

Wouldn't the old "whole heck of a lot of cheap copper pipes run in parallel" approach be even more practical?

I was going to mention those but I couldn't find any links to 'em so I forgot about it. They aren't cheap though so I doubt he'd be willing to spring for 'em. Though you could essentially have a huge Reserator for relatively cheap.