fans with speed control and tach output

Where did you pull the 100mm dimension from?

I'd be embarassed to ask such a question, without at least giving a cursory glance at local sources first, for basic dimensional tendencies in the field.

This has wide-ranging implications at the earliest conceptual stages of packaging.

RL

Hi,

I am just starting to think of the enclosure design for the SMPS I am working on. There will be a maximum of 360watts of heat to remove.

I would like to have control of the fans RPM so that the fans only blow as fast as is required in order to save power. I am not sure if you can PWM the fans 12V supply to control speed using the speed sensor output, like for one of these fans:

I was thinking of using 5 of these fans to remove a total of 300watts of heat max from the enclosure, I haven't gone through the heat budget just a rough initial estimate. The side of the enclosure that the fans are mounted in will be 26inches long, and at least 5 inches in height so will be able to fit five or more 80mm to 120mm fans down the length.

Five fans would be good as there are 5 rows of heatsinks on the PCB. Some of the heatsinks do not produce heat, depending on the mode of operation so those fans could be turned off.

Maybe one ambient temperature sensor would be enough along with PWM fan control.

cheers, Jamie

With five fans, the five tach wires would likely just get in the way and add uneccessary expense. A single analog control based on air flow or hot-spot temperatures would likely be more cost-effective and allow the use of more generic impellers.

Perhaps you should establish what acceptable air volume and noise levels are before even considering the requirement for fan control. There is no point controlling the speed of fans if the resulting noise level is still unacceptable under full rated conditions. A large fan count will complicate the noise issue.

Once you've established the air movement required by targetted dissipators, internal ambient limits and permitted noise levels, you can compare apples to apples where cost vs noise and its abatement is concerned.

In a large box, any arrangement resulting in the situation where the loss of one fan in a group of five would cause a critical circuit section to fail, should be avoided.

RL

Here's a good place to start: