Home » Intel Manuals » Processors » Intel QX9770 » Manual Viewer

Intel QX9770 Design Guidelines - Page 58

Output Weighting Matrix, 6.1.2 Proportional-Integral-Derivative (PID), Intel

UPC - 735858197564

Add to My Manuals
Save this manual to your list of manuals

Page 58 highlights

LGA775 Socket Heatsink Loading Figure 21. Intel® Quiet System Technology Overview 6.1.1 6.1.2 Output Weighting Matrix Intel QST provides an Output Weighting Matrix that provides a means for a single thermal sensor to affect the speed of multiple fans. An example of how the matrix could be used is if a sensor located next to the memory is sensitive to changes in both the processor heatsink fan and a 2nd fan in the system. By placing a factor in this matrix additional the Intel QST could command the processor thermal solution fan and this 2nd fan to both accelerate a small amount. At the system level these two small changes can result in a smaller change in acoustics than having a single fan respond to this sensor. Proportional-Integral-Derivative (PID) The use of Proportional-Integral-Derivative (PID) control algorithms allow the magnitude of fan response to be determined based upon the difference between current temperature readings and specific temperature targets. A major advantage of a PID Algorithm is the ability to control the fans to achieve sensor temperatures much closer to the TCONTROL. Figure 22 is an illustration of the PID fan control algorithm. As illustrated in the figure, when the actual temperature is below the target temperature, the fan will slow down. The current FSC devices have a fixed temperature vs. PWM output relationship and miss this opportunity to achieve additional acoustic benefits. As the actual temperature starts ramping up and approaches the target temperature, the algorithm will instruct the fan to speed up gradually, but will not abruptly increase the fan speed to respond to the condition. It can allow an overshoot over the target temperature for a short period of time while ramping up the fan to bring the actual temperature to the target temperature. As a result of its operation, the PID control algorithm can enable an acoustic-friendly platform. 58 Thermal and Mechanical Design Guidelines

Section	Page
1 Introduction	9
1.1 Document Goals and Scope	9
1.1.1 Importance of Thermal Management	9
1.1.2 Document Goals	9
1.1.3 Document Scope	10
1.2 References	11
1.3 Definition of Terms	11
2 Processor Thermal/Mechanical Information	13
2.1 Mechanical Requirements	13
2.1.1 Processor Package	13
2.1.2 Heatsink Attach	15
2.1.2.1 General Guidelines	15
2.1.2.2 The Pump Assembly Clip Load Requirement	15
2.1.2.3 Additional Guidelines	16
2.2 Thermal Requirements	16
2.2.1 Processor Case Temperature	16
2.2.2 Thermal Profile	17
2.2.3 TCONTROL	18
2.3 Heatsink Design Considerations	19
2.3.1 Heatsink Size	20
2.3.2 Package IHS Flatness	20
2.3.3 Thermal Interface Material	21
2.4 System Thermal Solution Considerations	21
2.4.1 Chassis Thermal Design Capabilities	21
2.4.2 Improving Chassis Thermal Performance	21
2.4.3 Summary	22
2.5 System Integration Considerations	22
3 Thermal Metrology	23
3.1 Characterizing Cooling Performance Requirements	23
3.1.1 Example	25
3.2 Processor Thermal Solution Performance Assessment	26
3.3 Local Ambient Temperature Measurement Guidelines	26
3.4 Processor Case Temperature Measurement Guidelines	29
4 Thermal Management Logic and Thermal Monitor Feature	31
4.1 Processor Power Dissipation	31
4.2 Thermal Monitor Implementation	31
4.2.1 PROCHOT# Signal	32
4.2.2 Thermal Control Circuit	32
4.2.3 Thermal Monitor 2	33
4.2.4 Operation and Configuration	34
4.2.5 On-Demand Mode	35
4.2.6 System Considerations	35
4.2.7 Operating System and Application Software Considerations	36
4.2.8 THERMTRIP# Signal	36
4.2.9 Cooling System Failure Warning	36
4.2.10 Digital Thermal Sensor	37
4.2.11 Platform Environmental Control Interface (PECI)	38
5 Intel Thermal/Mechanical Reference Design Information	39
5.1 Validation Results for the ATX Reference Design	39
5.1.1 Heatsink Performance	40
5.1.2 Acoustics	40
5.1.3 Altitude	41
5.1.4 Reference Heatsink Thermal Validation	41
5.1.5 Fan Motor Performance	42
5.1.6 Pump Motor Performance	43
5.2 Environmental Reliability Testing	44
5.2.1 Structural Reliability Testing	44
5.2.1.1 Random Vibration Test Procedure	44
5.2.1.2 Shock Test Procedure	45
5.2.2 Power Cycling	46
5.2.3 Reliability Testing	46
5.2.3.1 Tubing Material Selection	48
5.2.3.2 Reservoir Sizing	48
5.2.3.3 Reliability Test Results	50
5.2.4 Recommended BIOS/CPU/Memory Test Procedures	51
5.3 Material and Recycling Requirements	52
5.4 Safety Requirements	52
5.5 Geometric Envelope for Intel Reference ATX Thermal Mechanical Design	52
5.6 Reference Attach Mechanism	54
5.7 Socket and Voltage Regulation Cooling Strategy	55
6 Intel® Quiet System Technology (Intel® QST)	57
6.1 Intel® Quiet System Technology Algorithm	57
6.1.1 Output Weighting Matrix	58
6.1.2 Proportional-Integral-Derivative (PID)	58
6.2 Board and System Implementation of Intel® Quiet System Technology	60
6.3 Intel® QST Configuration & Tuning	62

Match case Limit results 1 per page

LGA775 Socket Heatsink Loading

Thermal and Mechanical Design Guidelines

Figure 21. Intel

Quiet System Technology Overview

6.1.1

Output Weighting Matrix

Intel QST provides an Output Weighting Matrix that provides a means for a single

thermal sensor to affect the speed of multiple fans.

An example of how the matrix

could be used is if a sensor located next to the memory is sensitive to changes in both

the processor heatsink fan and a 2

fan in the system.

By placing a factor in this

matrix additional the Intel QST could command the processor thermal solution fan and

this 2

fan to both accelerate a small amount.

At the system level these two small

changes can result in a smaller change in acoustics than having a single fan respond

to this sensor.

6.1.2

Proportional-Integral-Derivative (PID)

The use of Proportional-Integral-Derivative (PID) control algorithms allow the

magnitude of fan response to be determined based upon the difference between

current temperature readings and specific temperature targets. A major advantage of

a PID Algorithm is the ability to control the fans to achieve sensor temperatures much

closer to the T

CONTROL

Figure 22 is an illustration of the PID fan control algorithm. As illustrated in the figure,

when the actual temperature is below the target temperature, the fan will slow down.

The current FSC devices have a fixed temperature vs. PWM output relationship and

miss this opportunity to achieve additional acoustic benefits.

As the actual

temperature starts ramping up and approaches the target temperature, the algorithm

will instruct the fan to speed up gradually, but will not abruptly increase the fan speed

to respond to the condition. It can allow an overshoot over the target temperature for

a short period of time while ramping up the fan to bring the actual temperature to the

target temperature. As a result of its operation, the PID control algorithm can enable

an acoustic-friendly platform.