Week Seven

After several weeks of using the heat gun for methods of testing metals and the thermoelectric cell, the best course of action to proceed with was to actually apply the cell to the computer heat source.  The heat gun tests were believed to be inefficient throughout the entire process, so although those tests were necessary and did provide helpful results, the direct application of the cell on to the heat source is believed to provide the bet results.

Therefore, in Week 7 of the project, the cell was applied to the computer.  Based on the computer architecture analysis completed in prior weeks, the optimum location to apply the cell was decided to be on top of the heat sink, which is on top of the processor.  The heat sink of the laptop is not very efficient, and the surface of it reaches high temperatures when the computer processor is given a maximum workload.  The surface reaches temperatures that could burn the skin if applied for several seconds, so it should provide enough heat for the cell.

The computer was loaded with large video files and a Maple 16 script with a continuous loop so that the computer processor and RAM could be given a maximum workload and therefore reach high temperatures.

The computer keyboard was removed, and part of a metal casing under the keyboard was cut so that the heat sink could be reached.


Computer base with keyboard removed and area to cut marked in blue marker.


The functioning computer with metal base cut for access to processor heat sink.

Now that the computer was modified for access to its hottest heat source, the cell could be applied.  During Week 7, the group completed the first application of the cell to the computer.  The thermoelectric cell was applied to the heat sink using Dynex thermal paste.

Dynex thermal paste

The Dynex paste is comprised of silver and is believed to be a better conductor of heat than air, and therefore these tests were predicted to have better results than heat gun tests.  Then, in order to create a makeshift heat sink to keep the other side of the cell cold to create the desired temperature gradient, a sheet of copper was placed on the other side of the cell, with thermal paste. 

For testing, the computer was turned on and given a maximum workload, so that it reached a high temperature.  The cell was then connected to the multimeter to measure the generated voltage.

Week 7 Test Results:

Tests with thermal paste in the manner described above resulted in generated voltages of only 0.1 to 0.2 volts. 

Tests were also conducted without the use of thermal paste, simply the cell on top of the heat source.  This resulted in voltages of 0.05 to 0.08 volts. 

To the group’s surprise, the test results were much poorer than expected.  The maximum voltage acquired was about 0.2 volts, and this is obviously not enough potential difference to power the 5.0 Volt computer fan, or even a 1.5 Volt external LED light. 

Initial cell testing on the computer with thermal paste did not go as well as expected, however unexpected results should be expected with all engineering projects of discovery. 

The day after testing, the group searched for answers to explain why the tests did not go well.  Several factors were deduced:
·         The application surface was uneven, and the metals of the surface should be sanded down to create a flat, even surface.
·         The Dynex thermal paste was determined to be inefficient.  After learning about its “Percent Metal Composition,” the amount of silver metal in the paste was very low, and other thermal pastes are available with higher percent metal compositions.
·         The temperature gradient between the hot and cold sides of the cell was not good enough.  A better heat sink is needed for the cold side.

The group then found a better paste to use, Arctic Silver 5, which was believed to have a percent metal composition at least one order of magnitude higher than the Dynex paste.  This paste was purchased, as well as a silicone based heat sink compound, for application of the heat sink.

Also mentioned above, a better heat sink was needed to dissipate the heat from the cold side of a cell.  Therefore, a heat sink was harvested from an old desktop computer to be used on the cell.



Old gateway desktop computer opened.


Aluminum heat sink taken from desktop computer.

Next, the surface of the laptop were the cell was applied was sanded down to create an even surface.  A piece of copper of the laptop’s heat sink was the problem, because it rose above the metal around it.


Laptop heat sink, cell application site, sanded down. View 1.


Laptop heat sink, cell application site, sanded down. View 2.

Finally, more of the metal base under the keyboard was cut so that the computer fan could be seen, and accessed if it is to be used in the project.


Laptop with heat sink and fan accessible.

The next step of the project will be completed immediately at the start of Week 8.  The cell will be applied to the computer once again, however this time with the improvements of:

  • A more even surface of application
  • A better thermal paste for a more direct and efficient transfer of heat
  • A real heat sink for an improved temperature gradient across the cell














No comments:

Post a Comment