Quiet components = quiet power supply?
Commonly, one imagines this point as follows: There are active, semi-passive or passive power supplies. For the active and semi-passive ones, you install a fan that is as quiet as possible but still sufficient, and the great new power supply is ready. Far from it, because it can take days and even weeks just to optimize the operating noise. This is because there are far more sound sources and interactions within the power supply enclosure than you might think.
That’s exactly why you shouldn’t leave out the other noise sources in the power supply, because the quieter a fan works, for example, the more likely you are to hear other noises. Even at this stage, you can still make circuit-specific changes to a certain extent and also test other components.
We were able to talk to the engineer from be quiet! for two days and also see how complex an optimisation in an anechoic chamber really is. Since be quiet! uses a completely newly developed fan for the Straight Power E10, the fan control must also be individually adjusted for each individual wattage class and the individual load levels.
How high does the starting voltage have to be for the fan to start safely? And how must the curve be designed so that all loads and overloads can be safely cooled even under difficult conditions? The paver then changes the circuit in a few minutes if necessary and testing starts all over again – and this continues until the optimum is reached.
Incidentally, the test room at FSP has the very unpleasant property for visitors of not only swallowing sound but also producing high temperatures. In the end, the display showed less than 200 rpm for the safe start-up of the fans. This is such a good value that it almost makes semi-passive solutions absurd, because you simply don’t hear such fans. We were allowed to try this out in a sweaty self-experiment with a very sinking feeling of being trapped in practice.
You just hear everything in such situations – not only the fans, but also coils and capacitors, for example. In other words, exactly what our readers always like to call an “electric” sound. They are not completely wrong – only that we did not hear anything with this model.
Incidentally, the behaviour of the frequently used air baffles can also be tested quite well acoustically and thermally. If a power supply has been planned with the appropriate effort and the interactions of the assembled board including its components as well as of the case and fan itself have been and are taken into account, then such subsequent tricks are absolutely superfluous and only testify to a later hastily executed cosmetic correction of an otherwise probably rather inadequate and cost-saving development work.
4 Antworten
Kommentar
Lade neue Kommentare
Urgestein
Veteran
1
Veteran
Alle Kommentare lesen unter igor´sLAB Community →