FAN's and Noise
Okay, I have reviewed what everyone has sent so far, and most seem to be trying to change the power supply fans by dropping the airflow rates using external inline controllers, that is never proper in my book for the long life of the power supply or reliability. But that is not where the noise is coming from, 75% of the system noise on these units comes from the 3 primary fans for the unit itself. Simple test, fire up the unit without the 3 fan tray in the back, discover for yourself.
I'm working on a solution to use 3 fans of a much larger size, but without the noise level associated with the factory installed fans. Once I have the solution, I will share. Most of us have plenty of room in the rear of the rack to have an air movement device added, and this switch is not that deep. No where near the depth of a full server, so most everyone will have plenty of room.
I just need to know interest, as once I design and produce this for myself, I need to know if anyone else would be willing to pay for the same solution??? I can do the AutoCAD and find a supplier to print the part, but need to know what kind of volume (which I know will be very low) to estimate to the supplier. Look guys, I'm not looking to make a profit here, I just want to share, and I don't want to misrepresent myself to the supplier.
Please understand, dropping airflow across silicon chip heat sinks is a very BAD idea, you say "it only increased two or three degrees C", but in my field of failure analysis, that's several years of life shaved off the silicon. So the key to noise pollution control would be to enlarge the fan tunnel, use larger fans (that are whisper quiet) and move the same volume (if not more) of air. The cooler you can keep the silicon, the LONGER it will live! Just go look up HALT failure analysis, that should enlighten those that want to learn. Just my 2 cents.
Okay, I have reviewed what everyone has sent so far, and most seem to be trying to change the power supply fans by dropping the airflow rates using external inline controllers, that is never proper in my book for the long life of the power supply or reliability. But that is not where the noise is coming from, 75% of the system noise on these units comes from the 3 primary fans for the unit itself. Simple test, fire up the unit without the 3 fan tray in the back, discover for yourself.
I'm working on a solution to use 3 fans of a much larger size, but without the noise level associated with the factory installed fans. Once I have the solution, I will share. Most of us have plenty of room in the rear of the rack to have an air movement device added, and this switch is not that deep. No where near the depth of a full server, so most everyone will have plenty of room.
I just need to know interest, as once I design and produce this for myself, I need to know if anyone else would be willing to pay for the same solution??? I can do the AutoCAD and find a supplier to print the part, but need to know what kind of volume (which I know will be very low) to estimate to the supplier. Look guys, I'm not looking to make a profit here, I just want to share, and I don't want to misrepresent myself to the supplier.
Please understand, dropping airflow across silicon chip heat sinks is a very BAD idea, you say "it only increased two or three degrees C", but in my field of failure analysis, that's several years of life shaved off the silicon. So the key to noise pollution control would be to enlarge the fan tunnel, use larger fans (that are whisper quiet) and move the same volume (if not more) of air. The cooler you can keep the silicon, the LONGER it will live! Just go look up HALT failure analysis, that should enlighten those that want to learn. Just my 2 cents.
