Adaptive Cooling Using Fan Speed Control Examples

Adaptive Cooling System Design for Electronics using Fan Speed Control

Balancing Benefits

The key to designing an effective adaptive fan speed control system for electronics applications is striking the right balance between reduced noise, energy savings, fan life, constant semiconductor junction temperatures vs. higher semiconductor junction temperatures at reduced fan speeds.

Cooling System Variables

• Fan (air flow & pressure)
• Enclosure Design (air flow path)
• Heat Sources (power dissipation of devices)
• Heatsinks (thermal resistance)
• Using thermal analysis computer modeling early in the product design is vital in determining your cooling system variables.

Once these variables are established, the equipment’s temperature rise, Delta T can be determined.

Equipment Temperature Rise, Delta T = TE – TA1

TE = Exhaust temperature of equipment at full fan speed under normal operating conditions, ie: 20°C, sea level, clean filters, typical power usage.

TA1 = Ambient room temperature under normal operating conditions ie: 20°C, sea level, clean filters, typical power usage.

The temperature rise (or slope) of a particular fixed fan speed system will remain constant regardless of the inlet air temperature. In Figure 4 below, the equipment temperature rise is 10°C

Adding Variable Fan Speed Control to the Mix

To add variable fan speed control to the mix, we need to determine what the slowest fan speed (idle speed) will be and what the Fan Speed Control Slope (sensitivity) TS will be.

Idle speed should be set so that under normal operating conditions, semiconductor junction temperatures and fan noise levels are at a satisfactory level. Using computer modeling, semiconductor junction temperatures can be calculated at various fan speeds (noise levels). Consult MTBF and noise specifications as determined for your product.

If the goal is to maintain constant critical semiconductor junction temperatures, the Fan Speed Control Slope Ts should be equal to the temperature rise of semiconductor surface or junction temperatures between fan idle speed (as determined above) and full fan speed.

Now that these fan speed control variables have been determined, one can calculate a Control Temperature.

Control Temperature Tc = The sensor (exhaust) air temperature above which fans run at full speed.

TC = TA2 + TS + (TE – TA1) / S

TA2 = The ambient room temperature below which fans will run at the idle speed. To get maximum noise control benefits under normal environmental conditions, this temperature should be close to TA1 as determined above.

TS = Fan Speed Control Slope (sensitivity)

S = Fan idle speed as a percent of full speed.

Using the 10°C equipment temperature rise example above, selecting a 6°C fan speed control slope (TS), a 53% idle speed (S) and a 20°C ambient idle speed setting (TA2), we calculate a control temperature of 45°C and a fan speed control curve like the one illustrated in Figure 5.

Alternate Fan Speed Control Method (sensing surface temperature)

The surface temperature of one or more critical semiconductors can be sensed and the fan speed control can be varied to hold this temperature near constant.

To do this properly, the fan speed control slope must be as small as possible. We have found that a 3 – 4°C slope is about as low as one can go without introducing hystoresis or “hunting”. That is, fans would tend to constantly cycle up and down in speed “hunting” for an acceptable speed.

Adaptive Cooling and Ventilation for HVAC using Fan Speed Control

• Clean Room Pressurization (pressure control)
• Attic, Whole House or Duct fans (temperature or pressure control)
• Livestock Ventilation (temperature control)
• Ceiling Fans (temperature control)
• Indoor Pools and Spas (humidity control)
• Locker Rooms and Bathrooms (humidity control)
• Industrial Equipment (temperature, humidity or pressure control)

SmartFan saves energy, reduces noise, increases fan life, compensates for changes in system configuration and provides the end user with the comfort and convenience of automatic control.

SmartFan^® Fan Failure Alarms and Temperature Alarms

Should an air mover in a cooling or ventilating application fail to operate properly, the results can be catastrophic. In electronics equipment, failure of an air mover may cause operational failure or even permanent damage. In ventilating applications, contaminants may build to dangerous levels, equipment may fail or livestock may perish.

Fan Failure Alarms

Air mover speed is sensed using pulses generated by a Hall Effect (tach) device installed in the fan. Below a predetermined speed, a fan failure alarm is triggered. This method is very reliable, however it requires a fan with a Hall Effect (tach) output. Because the Hall Effect device is part of any brushless DC fan, the output adds little cost. This feature may however add significant cost to an AC fan and may not be available on all models.

Temperature Alarms

Measuring temperature (air or surface) is fundamental. In combination with speed control it adds little cost because the sensor used to control speed can also provide a signal to the alarm circuitry. One disadvantage is a possible delay after air mover failure, before the alarm is triggered. A temperature or humidity alarm will of course also respond to failures other than the air mover. For example, operating equipment at excessive ambient temperature or with a clogged air filter could also trigger the alarm.

Alarm Pros & Cons

Alarm Outputs

Alarm outputs may be referenced to the negative terminal of a fan or can be optically isolated. In applications requiring higher switching (sinking) currents or voltages, the standard optical isolator can be replaced with a normally closed or normally open MOS relay. Alarms are available as an:

• Electrical signal to drive logic
• Visual signal (LED)
• Audible signal (piezo alarm)