Optimizing case airflow and cooling Feel free to ask questions after reading. Automatic heat / load control of case fans
Interesting Lloyd. I am still awaiting the Phanteks from Overclockers, it is due shortly so I expect to be picking your brains in the future.
One additional thing I can suggest for actually visualising airflow is a a clear enclosure (to compartmentalise the smoke) and a perspex cover tacked where the case lid would normally be, and smoke run through the PC - and videoing it at a high frame rate. While temp sensors are key, the smoke method can also help you identify trouble spots. HVAC test smoke is a good budget homebrew test candidate - the enclosure you'll probably have to build yourself, I didn't find any pre-fab candidates.
@Kronos, Pick away, but there's not much to pick at. @seapanda, Nice in theory, but every case, every build in same case, every cable move, different fan speeds, etc. all cause changes in airflow pattern. Years ago I was in R&D for automotive engine induction/exhaust airflow, engine and other component cooling airflow in racing. It is amazing how very small changes can cause radical changes in airflow pattern/performance.
Smoke only works if you can see what it's doing. So we could use smoke in a clear case to setup optimum cooling in the clear case. But what is working in the clear case will most likely change if we add HDDs, change GPU, etc. Which is a clear case of what happens in a clear case when things change .. this makes a clear case for why each system must be optimized independently. This is a clear case showing how the clear case will change airflow patterns when components are changed. It's like tuning an engine to get the most power. Every engine performs a little differently even though they may be the same design and size.
OK... I can see we're on slightly different wavelengths. I mean a clear enclosure-baffle for the actual build you're testing, which in turn has a clear temporary lid applied to it to allow the case to take in and direct smoke properly. You didn't honestly think I was suggesting people build a clear case to build in so that they could then transfer their build to their real rig after they had run smoke tests in that case, did you?
I was joking with much of my last post. Have you tested airflow with smoke? As often as not it is impossible to tell what the flow is doing because the smoke mixes with clear air and becomes unidentifiable. I've had better results monitoring temps in different positions. If airflow is in a smooth pattern smoke trails can be followed. But if there is much turbulence it mixes so fast we loose tracking ability without some very expensive cameras and smoke equipment .. Way out of my price range.
