Test of high-speed shooting with E-M5

Started Jun 13, 2012 | Discussions thread
Senior MemberPosts: 4,899Gear list
Re: Personal attacks are easier than constructive criticism
In reply to bobn2, Jun 20, 2012

bobn2 wrote:

peevee1 wrote:

Often, heat problems in this kind of equipment are the most intractable and expensive to fix. If your thermal simulation hasn't proved accurate, and you find units failing in soak tests when you've already committed to production tooling, the costs can be horrendous. Often, the quickest fix is to squeeze in a cooler somewhere, and if it's a late addition, it probably won't be possible to put in any sound damping.

Right. Or design it with the cooler in the fist place, like almost all cutting edge electronics does. Another quick fix is to lower frequency and voltage a bit for the sake of better yields/reliability. Sure, it might hinder performance a little, like, between 9 fps and 8.15 fps, but who will notice if you don't publish frequency in the first place. And it is still better than having the device hang once in a while...

There is quite a lot of research activity in this area, simply because localised cooling of increasingly dense electronics is a big issue, and there is a lot of money for anyone solving it. My favourite is the MEMS stirling engine.

The problem with 'design it in the first place' is that the first law of design is Murphy's. Thermal simulation is OK, but until you know exactly how much each chip puts out, you don't exactly know what are the internal heat flows, and of course they are chaotic. So, one makes an informed guess, puts coolers in the wrong places, but still units fail in soak.

I think there are literally 2 chips in the camera which are heat-constrained in their performance - the sensor and the image processor (which is probably on the same chip as application processor running the menus etc). Hard to say without disassembling it if one of them is cooled by the fan or even both. When I first saw the IQ in comparison to GX1 and heard about the noise, I thought about cooling the sensor (the economics of producing their own sensor which is so much like Panasonic's seems very implausible to me). When I actually heard the noise it pretty much convinced me it is a fan, I heard thousands of them. And when I compared the speed to E-P3/E-PL3/E-PM1 with the same processor... I don't see how it can be achieved WITHOUT a fan. Maybe the fan is for general recirculation, as convection in such a dense body cannot possibly be any good. Then it could benefit the sensor and processor if designed correctly.
Anyway, I like what I see in this camera and that is what matters.

As I suggested before, putting some kind of heat redistribution device, a sink or a fan (or a MEMS Stirling engine, or a MEMS fan http://www.colorado.edu/MCEN/mems/research_cooper1.htm )


Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow