Design for Thermal Issues

Assignment 1
Modelling heat at the board level

Covers Learning outcomes 1 & 2
Units 1–8
Notional workload 25 hours
Weighting 50%
Comments For information on submitting assignments, please refer to the AMI home page under Students, Assignments
Submission date Midnight at the end of study week 8
Please refer to the module planning chart for the date

Introduction

As described in the Module Descriptor, this module is assessed by two assignments, the first of which focuses at the component/board level; the second assignment will extend your use of simulation to evaluating an enclosure with multiple boards.

As part of this first assignment, you are asked:

To supplement your report on this work, we will be looking both for a critical evaluation of the different approaches used, and for an analysis of how the thermal characteristics of the original design might be expected to affect the reliability and overall function of the product.

The application

An organisation sub-contracted to CISCO systems has developed a new fibre-optic/copper router, part of which is a 12-layer PCB assembly. As head of the thermal analysis team for this organisation, you are required to undertake an evaluation of this assembly in order to:

This link describes the thermal environment for the router PCB assembly and the components used, including a preliminary layout for the assembly.

Key information

The router board that you analyse in this assignment is a key component in Assignment 2. Even at this stage you need to bear in mind the overall application, and think about how you might incorporate your model into a model of the complete assembly.

 

The task

The elements of the task are:

  1. Having read the assembly description in detail, use your knowledge of assembly and package structures to estimate the steady-state temperatures that will be reached by the most significant dissipating elements. You are encouraged to use web resources to identify typical values for package parameters and other missing information, but should, as a matter of general good practice, record and acknowledge your sources in your report.
  2. Having made first-pass estimates using a high degree of simplification, refine your estimates using appropriate calculation aids such as spreadsheets.
  3. Model the assembly using FLO/PCB and perform thermal simulations, using the same generic components as before. Check that your preliminary estimates correctly identified the hottest component(s).
  4. Build up using FLOPACK a detailed model of the most critical component and import it onto the layout in order to refine your FLO/PCB analysis.
  5. Modify the layout, within the given constraints, to achieve an optimal thermal solution. Then investigate and advise on the options for making further improvements, for example by adding heat sinks to appropriate components.
    Note: You are expected to demonstrate that you have used your knowledge of the underpinning theory to select appropriate candidate solutions for successive models, and have not simply jumped to a conclusion without explicit justification.
  6. Examine the impact on the thermal performance of the assembly of moving the power supply in order to improve the mechanical stability of the unit.
  7. Estimate the rate at which the temperatures within the assembly achieve stability after switch-on.
  8. Based on your knowledge of semiconductors and the other kinds of components typically used in power supplies, research appropriate advice for your team on:

The first five of these need to be tackled in sequence, elements 6 and 7 can be tackled in either order, and you are recommended to start work on element 8 before completing the simulations.

Notes

Throughout the task, don’t forget to apply common sense! Is your result within the limits of what you might expect? This is especially true in element 1, where we expect a reasoned engineering estimate based on your knowledge of the thermal environment and your study of the module, and some explanation of (or at least comment on) any unexpected results.

For element 2 of the task, we are expecting you to try to refine your estimates by building more detailed spreadsheet models. For example, the 144-lead device that in element 1 would be modelled as just one or two thermal resistances, in element 2 might be modelled by calculating appropriate values for the more detailed set of interfaces indicated by Figure 1.

Figure 1: Schematic thermal model of a 144-lead device

chematic thermal model of a 144-lead device

 

For element 7 of the task, you will find that FLO/PCB does not support transient analysis (for which you would need to use FLOTHERM). However, simple calculation based on the thermal mass of the assembly and the heat input will scope the thermal time constants of the board and give you sufficient information on which to base your comments in the second part of element 8.

In element 8, we expect you to demonstrate some understanding of failure rate issues as they affect the whole range of likely components (not just semiconductors), under both steady state and power-cycling conditions.

A word to the wise . . .

In general, there are no single correct answers to simulation questions, so the key to getting high marks is to submit reports that show the thought processes through which you have gone, and display your understanding of the topics, as regards both the simulation itself and the practical issues involved in thermal management.

 

Requirement

Your submission should be a report to the design management team within your organisation covering the following elements:

You can assume that the members of your audience are aware of the components used and of the functional features of the design, but have only a partial understanding of thermal matters.

As a guide, you are unlikely to include all the points we expect to find if your main report has fewer than 2,000 words. There is no maximum word count as such, but excessive length may be penalised.

In order to keep your script concise and well-argued, and its structure clear, you may find it helpful to provide relevant supporting material in appendices to the main report. This applies particularly to detailed calculations and information collected during the analysis and simulation elements of the task.

Caution

In writing your report, you are expected to include some evidence of having studied all the relevant course material and perhaps supplemented this with wider reading/web browsing. In order to make reference to this in an appropriate and consistent way, we recommend that:

Given the transitory nature of some web sites, we also recommend that you retain electronic copies of any material cited!

 

As always, you are strongly recommended to re-examine your draft report and conclusions to check that you have covered all the elements required in the report.

Marking scheme

The maximum marks available for each element of the report are as shown in the table below:

Element
Maximum marks
Modelling the assembly using a range of approaches
30%
Critical evaluation of modelling approaches
10%
Optimising thermal performance
15%
Analysis and recommendations concerning moving the power supply
10%
Analysis of reliability and performance impact of thermal characteristics
15%
Analysis of reliability and performance impact of duty cycle reduction
10%
Quality of presentation (including introduction and conclusions)
10%

For information on grades, please refer to the AMI home page under Students, Assignments, Marks and Grades.

[back to top]