I mentioned this briefly in an article thread but in hindsight it is more relevant here. When testing the cooling performance of cases, would a thermal image show a useful "picture" of the hots spots in a case under CPU and/or GPU load? My thought is that a large piece of acrylic, with foam around the edges to seal against the side of the case, could act as a window during the load test(s). Taking a thermal photo would show, to dramatic affect, the thermal pattern in the case. Example: http://www.behardware.com/articles/787-6/report-nvidia-geforce-gtx-480-470.html For cases with side fans supplied as standard, a bit of creative cutting and remounting of the fan could work. If that sounds like too much work, maybe something other than acrylic would work? Secondly, whilst I mentioned it in the article thread with my tongue firmly in my cheek, the videos of Silverstone demonstrating the positive air pressure in the FT01 actually isn't that silly. They used a smoke machine, with a piece of acrylic replacing the side panel, which physically demonstrated the actual airflow through the case. I don't know if is actually practical for Bit-Tech but hey, maybe it could be? http://www.youtube.com/watch?v=DLOg9yI3rjs http://www.youtube.com/watch?v=Qe-2ZqmSGug I think that either of these demonstrative methods would really increase the draw of the case tests for the performance community.
In the case labs test, they use a borrowed £50,000 thermal imaging camera for viewing the hotspots etc. but for every single test, I don't think its a real worthwhile thing to do... Anyone who knows the slightest thing about heat can stop where and when there is going to be a heat issue in a case.
£50k Oh well, it was a thought... What about the smoke machine airflow test, that wouldn't cost much. It would show the case airflow as well as demonstrate how much air supplied fans can shift, something that is harder to do in words than it is in video. Bit-Tech could also re-use the smoke machine when testing fans.
Ahhhh, but there is method in my madness... Tell the boss you are doing a smoke test then do the test outside in the summer sun! Do a deal with the landlord to let you do it in the beer garden.
Think Behardware copied us on that one tbh as we did it years ago in Custom PC! Having done the last case lab test with James, I can freely admit that delving any deeper into testing takes a LOT of time - something we don't have for individual case reviews. The results and images from using the thermal imaging camera were awesome but as already mentioned, we can only loan this on rare occasions. Even then, the acrylic side panel (we actually used sheet acetate) has one limitation - if the case has a side panel fan, you either spend hours cutting new panels and fitting the plethora of different sized side panel fans, or you take the images without it. And yes, using a smoke generator in our lab would be good for one thing - an afternoon in the pub, clutching our P45s.
No worries, a little bird told me you were considering a case group test. I am still undecided on a new case so I thought I would suggest some things that could help me decide which sexy looking, high airflow and silent case I should buy without having to house it in the garden or risk a slapdown from the bank manager.
Would if not be more pratical to take vigorous temp measurements? Cpu, core logic, mobo vrm, hdd, gpu. gpu card, case ambient? Perhaps with one very hot high end tests system for the toprage cases and a moderate pc one for the rest? No one seems to do really comprehensive case testing, and I don't think you can tell how well the case cools by simply looking at where the holes are.
Not sure if you've looked at our current tests but we do most of that already? Certainly for the CPU and graphics card. There's no easy way of measuring VRM temperatures - probes are very hit and miss and you'd have to make sure that A) the probe wasn't touching the VRM heatsink and B) the probe was in exactly the same position for each test - not easy when you're swapping hardware between scores of cases. In any event, many VRM and chipset heatsinks are designed to make use of residual air from the CPU cooler, rather than general airflow inside the case. I think it's fair to say though that if the CPU runs comparatively cool in a particular case, that the other hot spots on the motherboard are probably well catered for as well.
Not You Again! Fair point about VRM temps actually, it's always seemed to me that as some VRM's on mobo's and particularly graphic cards get so damn hot this must limit overclocking or worse, long time reliability, but I do appreciate standardizing tests is really tough. About your standard test kit; why a passive graphic card? As most cards direct heat out [or around about anyway] wouldn't this effect the results by a lot? Ta
We're updating the test methodology and kit for an upcoming group test at the moment, one of the components being a more up-to-date actively cooled graphics card with a direct exhaust cooler.