Executive Summary of Results
The Task International Incorporated TK-6000 chassis provides adequate cooling for the Intel D845GRG desktop motherboard using the 2.53GHz PentiumR 4 processor, three add-in cards, two CD-ROMs, one floppy drive and one hard drive.

Introduction
The data compiled in this report is for your reference to facilitate in designing ATX 2.03 compliant chassis that meet the thermal, mechanical, and electrical requirements of Intel motherboards and accompanying processors.

The importance of ATX 2.03 compliant chassis and power supplies lies in the fact that it makes system integration easier and better for everyone. The specification promotes interchangeability, which makes integration/assembly of motherboards (with chassis independent I/O shields) and chassis easier. Overall, this results in lower integration and system costs and also reduced support cost for you. As you can see, ATX 2.03 compliant chassis/power supplies benefit everyone!

Test Configuration
All peripheral bays and add-in card slots were NOT populated for these tests. A heavily loaded configuration will result in lower performance than noted in this test report. This configuration is considered to be the MINIMUM performance required of a chassis.

• *Omega Engineering, Inc. PCI-DAS-TC 16 channel Thermocouple and voltage board.
• *Alinco DM-340MV DC power supply for test chamber.
• Custom designed 1/2" Plexiglas temperature chamber that is 35" wide x 36" high x 28" deep.
• *Omega .005¡¨ diameter *Teflon coated K-type thermocouples.

Mechanical Test
The component size was evaluated against the product specification for compliance. The passing criterion is defined as not exceeding the component product specification. The ATX 2.0 board gauge was used to determine compliance with keep-out zones defined by the ATX 2.03 specification section 3.4.

Thermal Tests
The peripherals listed in section 4 were installed in the chassis and K-type thermocouples were attached at the points designated in section 11. The chassis was tested in a controlled temperature chamber held at a constant 35C. The system was run under the *Microsoft Windows 2000 environment where a test program was initialized that exercises the processor up to the 75% Thermal Design Point (TDP). An *Omega data acquisition logger then sampled the thermocouples every thirty seconds for 60 minutes providing a total of 120 data points. The system was exercised until the initial thermal gradient reached a consistent level with a slopenearing zero. During testing, the ambient temperature was monitored approximately 5cm from the front bezel of the chassis.

Chassis Description (as Tested)
The Task International Incorporated TK-6000 chassis is a midtower chassis that ships with one 420W power supply, two system exhaust fans and four intake fans. It has four 5.25" exposed drive bays, two 3.5" external drive bays and six 3.5" internal drive bays. The dimensions of the chassis are 19" L x 8" W x 17" H.

Support Software
The following software was used in this thermal evaluation:

• Max Power v6.0 - an Intel proprietary processor stress utility designed for the Pentium 4 processor
  
• *OmegaSoft v5.20 data acquisition card drivers
  
• *NetDAQ macro summary sheet - post processes acquired data and evaluates limits.
• DOS Wizard v2.03 Data Acquisition Control Program

Test Setup and Procedure
Thermocouple wires from the test points were carefully routed to avoid modification of the airflow pattern inside the system.

To achieve the 35C maximum specified ambient temperature, the system was placed inside a Plexi-glass housing. Temperature inside the plexi-glass chamber is controlled by a 2-channel process controller, which regulates a heater inside the enclosure for adding heat and two fans for cooling.

Component temperatures vary greatly depending on what application or exercise program is being run. For these tests the worst-case program was used depending on which component was being evaluated.

An *Omega PCI-DAS-TC 16 channel data acquisition card was used for data collection; readings are
monitored until all temperatures have stabilize (usually 1.5 hours). Once data was acquired, it is then post processed using an Excel macro and associated thermal limit file. The thermal margins are determined by extrapolating the data to a 35C ambient, making comparisons against the predefined thermal limits and calculating the RMS value of the last 6 adjusted stable readings collected.

Processor Testing

All tests were performed using a 2.53GHz. Pentium 4 Processor as per plan of record. This represents worstcase processor temperatures for thermal evaluation based on available components at time of testing. During processor testing, the processor heat spreader had a thermocouple attached to the top surface of the heat spreader, and the temperature was measured while the MaxPower v6.0 program was executing. This software is designed to produce the greatest possible processor heating and is intended only for processor thermal studies. The program simulates a software situation, which is more strenuous than normal real world applications. In reality, desktop applications would involve memory, disk drive, and network accesses, which force the processor to wait. The Pentium 4 heat spreader is not to exceed 69C.