Intel Xeon 5500 Series, Design Manual

Page: 44 / 98

Quality and Reliability Requirements
44 Thermal/Mechanical Design Guide
8) Thermal
Performance
Using 1U heatsink and 1U airflow from
Table 5-1 :
1) TTV @ 95W (Profile B), Note 1.
Using 2U heatsink and 2U airflow from
Table 5-1 :
2) TTV @ 95W (Profile A), Note 1.
3) TTV @ 80W.
4) TTV @ 60W.
Using Tower heatsink and Tower airflow
from Table 5-1 :
5) TTV @ 130W, Note 1.
6) TTV @ 95W (Profile A).
7) TTV @ 80W.
8) TTV @ 60W.
As verified in wind tunnel:
1) mean
Ψ
CA
+ 3s + offset not to exceed
Table 5-1 value for 95W in 1U.
2-4) mean
Ψ
CA
+ 3s + offset not to exceed
Table 5-1 value for 2U.
5-8) mean Ψ
CA
+ 3s + offset not to exceed
Table 5-1 value for Tower.
5 heatsinks
X 8 tests by
supplier.
Note 1: 30
heatsinks X
3 tests by
Intel.
9) Thermal Cycling Required for heatpipe designs.
Temperature range at pipe in heatsink
assembly: -25C to +100C for 500 cycles.
Cycle time is 30 minutes per full cycle,
divided into half cycle in hot zone and half
in cold zone, with minimum 1min soak at
each temperature extreme for each cycle.
See
Figure 6-1 for example profile.
As verified in wind tunnel:
• Mean
Ψ
CA
+ 3s + offset not to exceed
value in Table 5-1 .
• Pressure drop not to exceed value in
Table 5-1 .
15
10) Heat Pipe Burst Continuously raise oven temperature and
record the burst/leak temperatures of fully
assembled heatsinks
No failures at minimum of 300C @ 20
minutes
32 pipes
11) Heatsink Mass Design Target < 500 g All samples < 550 g 30
12) Heatsink Load Design Targets:
0.062" board = 38.7 ± 7.2 lbf (Fmin =
31.5 lbf).
0.100" board = 51.4 ± 7.9 lbf (Fmax =
59.3 lbf).
No samples < 30 lbf on 0.062" board.
5 highest load samples (from 0.062" test)
< 60 lbf on 0.100" board
30
Table 6-1. Heatsink Test Conditions and Qualification Criteria (Sheet 2 of 2)
Assessment Test Condition Qualification Criteria
Min
Sample
Size