Canon 550D cold finger

[MarkS]

A common procedure for cooling a Canon DSLR is to insert a copper cold finger between the back of the sensor chip and the circuit board directly connected to it.

Here is the rmoved 550D sensor assembly:



Looking from the side, the gap can be seen:



The gap is tiny - I can't find my feeler gauges but I estimate the gap to be around 0.3mm  This might not be sufficient to conduct away the heat.  I'll need dig out my thermal conductivity equations and do some maths.

This might explain why CentralDS do a "Frankenstein" conversion of the similar 600D whereby the cold finger goes around the whole sensor and circuit board assembly:
http://www.centralds.net/cam/?product=cds-600dx-cooled-eos-600d-2

So I might have to go back to the drawing board

Mark

[JonH]

0.3mm doesn't sound like much to work with at all, but getting some copper shim that size will be quite easy.
But to make the most of it the surface finish of the copper will need to be flat lapped to give optimum contact and conductivity.

[MarkS]

0.3mm doesn't sound like much to work with at all, but getting some copper shim that size will be quite easy.
But to make the most of it the surface finish of the copper will need to be flat lapped to give optimum contact and conductivity.

When I did the 350D cooling mod I was able to use a 2mm thick copper finger, so 0.3mm will only result in one seventh of the heat transfer.

[mickw]

I seem to remember an alloy that has a greater heat transfer than copper - Can't remember what though 

[Mike]

I seem to remember an alloy that has a greater heat transfer than copper - Can't remember what though 

Only Silver has a higher thermal conductivity than Copper.

[MarkS]

Only Silver has a higher thermal conductivity than Copper.

Cool!  How much is a sheet of 0.3mm thick silver?

[Mike]

Not as much as you might think:

http://www.cooksongold.com/Sheet/Sterling-Silver-Sheet-0.30mm-Fully-Annealed-Soft-prcode-CSA-030

[MarkS]

According to
http://en.wikipedia.org/wiki/List_of_thermal_conductivities
we have the following thermal conductivities:
Copper    385
Silver      427
Carbon Nanotubes  3180
Graphene              4840
Diamond is potential much higher.

Where can I buy 0.3mm sheets of diamond?

Mark             

[JonH]

According to
http://en.wikipedia.org/wiki/List_of_thermal_conductivities
we have the following thermal conductivities:
Copper    385
Silver      427
Carbon Nanotubes  3180
Graphene              4840
Diamond is potential much higher.

Where can I buy 0.3mm sheets of diamond?

Mark             

Actually that would also be much cheeper than you might imagine...
Still several hundred at least though.
Diamond machine tools use very thin wafers of diamond bonded to a substrate. There is a limitation on the size though, it is about a 1" diameter i think.
I haven't a clue how they create the wafers though, i did ask a supplier few months ago but they didn't have a clue and just buy them in. Apparently it is a VERY closely guarded secret, or perhaps he was BS'ing...
Anyway, a company like nixon diamonds would prob be able to supply you a wafer.

[MarkS]

Actually that would also be much cheeper than you might imagine...
Still several hundred at least though.
Diamond machine tools use very thin wafers of diamond bonded to a substrate. There is a limitation on the size though, it is about a 1" diameter i think.
I haven't a clue how they create the wafers though, i did ask a supplier few months ago but they didn't have a clue and just buy them in. Apparently it is a VERY closely guarded secret, or perhaps he was BS'ing...
Anyway, a company like nixon diamonds would prob be able to supply you a wafer.

Yes, I saw it on the Royal Institution Christmas Lectures just a few weeks ago (The Modern Alchemist - the final episode I think).  The lecturer produced a wafer of diamond - it looked just like glass - produced by chemical vapour deposition.  Just holding it in their hand they easily cut through a block of ice because it was so incredibly thermally conductive.

So I did know such a thing existed

He also burnt some diamond in pure oxygen for a laugh.

Mark

[JohnP]

Mark,

We have some copper sheet in work that is 0.5mm thick & has had silver deposited on it for improved conduction. I could probably get you a small piece if any good...?

John

[Tony G]

Hi John,

Can you get me some pieces............about 7853 12" square please. 

Tony G

[The Thing]

How about something like this? A thin active heat pipe.

[MarkS]

That is cool (pun intended)!

Thermal conductivity up in the diamond zone.  But it is way too fat  
I got a feeler gauge from Halfords tonight and the gap is just slightly more than 0.3mm

We have some copper sheet in work that is 0.5mm thick & has had silver deposited on it for improved conduction. I could probably get you a small piece if any good...?

Thanks for the offer but 0.5mm is just too thick.  I'm probably better off getting 0.3mm sheet rather than grinding 0.5mm down to 0.3mm.

[The Thing]

I got a feeler gauge from Halfords tonight and the gap is just slightly more than 0.3mm

Have you thought about extending the flexible copper circuit cables and shimming the back board away from the CCD? Would mean rehousing the camera - but you are never going use it off a scope anyway...