On Fri, 19 May 2000, Sergio Brandano wrote:
> > >> Sorry, I do not see your point. > >> > >> My point is that if the CPU is piping hot, and you shutdown the > >> computer, no cooling is provided. This is wrong, as the fan has > >> nothing to do with the OS and it *must* spin until the temperature > >> reaches a safe level. > > >Well, the fan is there to remove the exess heat generated by a running CPU, > >if > >you shut off the CPU, there will be no more exess heat generated, and the > >existing heat will most assuredly be dissipated to less active cooling > >mechanism. Since you cannot damage the CPU, or other pieces of the hardware, > >this should cause no problem. And i think the G3 processor is not so hot > >running, that it will be so much above the safe temperature state. > > Ok Sven. You are convinced about your idea, but I am still skeptical. > In order to persuade me, I need a proof. Please perform this simple > experiment, then report the result. The experiment is as follows. > Have a nice and long trip with your car, then come back home, and > switch the engine off ensuring that the cooling fan is not spinning. > Let us know if your car starts again the next day. Bad analogy, many cars have fans that don't run after shutdown and they run fine every day (I've never owned own that was any other way). The ones that run electric fans after shutdown tend to be highly stressed small engines with thin-wall blocks in cramped engine compartments with poor air flow and small capacity high-pressure cooling systems. In such situations there is a potential for localized boiling of stagnant coolant around hot spots in the block. This can result in problems like accelerated corrosion, coolant overflow, and coolant vapor lock. Running the fans helps to maintain some coolant movement by convection and to also cool the enginecompartment to avoid fuel-line vapor lock. Notebooks, are a very different case, none are liquid cooled and the temperature levels and differences between coolant and components are not so extreme. Nor are you constrained by the boiling point of the coolant and the batteries are not going to vapor lock. If you are worried about your Pismo: 1) Leave your screen up for a few minutes after shutdown if it seems unusually hot (frankly mine has never been anywhere near as hot as it gets just idling in MacOS -- fire up a DVD and you could roast marshmellows). Convection through the keyboard is the primary cooling mechanism, the fan isn't even secondary (and if it did run after shutdown, it would have very limited effectiveness with the screen down). OR: 2) You are free to not run Linux, but you better not run Darwin either. Darwin also doesn't have any thermal management for any of the PowerBooks. The high-level abstract classes for thermal management, including fan speed control, are in the new Darwin IOKit, but no machine specific implementations -- yet. Even with the thermal management of the fan speed, all you gain is a bit of noise and perhaps power reduction for transient thermal increases. If the increased thermal load is sustained, the fan will eventually have to run full speed, which is what the PMU appearently does anyway as a fail-safe.
