Peltier Cooler Testing
Further investigation was needed into the behaviour of the Peltier Cooler
arrangement with an adjustment of provided power. In order to stabilize the
temperature of the 0-200V voltage supply, two aluminium slabs would be shaped and
used to provide some shielding from the ambient temperature interferences. The
purpose of this procedure was to observe if the aluminium blocks had a significant
effect in providing some temperature stabilization. The Peltier would be situated on the
outer side of the bottom aluminium block, with the voltage control circuit carefully
sandwiched between the two blocks. Therefore, to deduce whether or not the blocks
were effective in providing some form of stabilization for the voltage tuning control, the
thermal effect of the aluminium block had to be compared with what was assumed to
be less effective shielding. This was an aluminium box, which was used in the first tests
carried out on the Peltier cooler, where a relatively large temperature sweep was
1) The masses of the following apparatus were taken:
Two large aluminium blocks (which would be shaped for the purposes of
containing the voltage supply in the final system).
The box that was used to house the Peltier cooler (in the preliminary stages of
testing where the saturation occurred).
The mass balance used has an uncertainty of 50g. Each of the aluminium blocks
therefore had masses of 5.5(5) kg. The aluminium box has a mass of 0.5(5) kg.
2) The resistance of the 25W resistor was measured as 4.7Ω, so this provided a
basis measurement for the maximum current that’d be allowed through the
resistor before exceeding its power rating. It was deduced that approximately
2.2A was the maximum permissible current.
Arrangement and Experiments
The initial arrangement consisted of a 25W resistor mounted on the bottom of
an aluminium box; having been mounted on it with thermal paste and electrically
insulating tape. A PT100 thermistor was used to provide a measurement of temperature
between 20 and 30 degrees Celsius. The probe tip was mounted at the bottom of the
box. Connected to an appropriate +15V power supply which then read out to a DAQ,
the thermistor was then used to provide a temperature measurement. A voltage supply
was connected to the resistor, and set to provide a voltage of approximately 4.7V. The
current it provided was adjusted over a range of 2A; the currents fed in so as to power
the resistor being in the region of 0.5A, 1A, 1.5A and 2A. Whilst the voltage supply was
turned off, the current it provided was adjusted. A stop watch was then used to, from
the second the current was turned on; record the voltage that was read out on the DAQ
(which numerically corresponded to the temperature above 20 degrees Celsius). At first,
when the temperature increase was most rapid, the readings were taken every 5-10
second and, afterwards, longer intervals were used. After waiting for some stabilization
to occur, or at least having waited for the temperature profile of the thermistor probe
to take on a smooth shape in which the growth had gradually began to slow down, the
current was turned off. Temperature measurements were again taken using the DAQ
Multi-meter, at appropriate time intervals.
PT100 temperature sensor
to voltage supply
4.7 Ω + 5% resistor
to Peltier cooler
Fig. 1: The set-up for the first tests on the change of temperature; with the probe of the
PT100 fastened to the inner base of the aluminium box with thermal paste and
electrically insulating tape.
The run was repeated for an approximate current of 2A fed into the thermistor
probe; but this time with the lid placed on top of the aluminium box. Again, readings
were taken as well as recorded time intervals, until saturation; or at least some
indication of the beginning of temperature stabilization, was attained.
Comparisons of this presumably weak form of environmental shielding, with the
actual intended set-up, were needed to get a feel of whether or not stabilization was
aided by use of the blocks.
To loosely mimic the intended set-up, a 25W resistor was placed below one of
the two horizontal aluminium blocks, with a thermistor probe sandwiched in between
the two. For the purposes of making sure the probe of the PT100 thermistor was not
damaged in this experiment, a small aluminium caging was used to surround the PT100
probe. For 1A and 2A supplied to the resistor, the temperature read-out on the DAQ
was again taken in accordance with appropriate time intervals.
aluminium blocks PT100 temperature sensor
4.7 Ω +5% resistor
to Peltier cooler
probe tip encased in aluminium
Fig. 2: The arrangement for examining the behaviour of temperature in between the
aluminium blocks, which will be shaped to correspond to that required to encase the
Results and Analysis
The plots for an approximate current of 2A fed into the 25W resistor are as
shown. To inherit some understanding of the timescales needed for certain temperature
changes and the variation of the profiles of the curves, the regions displaying
exponentially decaying behaviour were extracted.
Fig. 3: Temperature dependence on time, for no lid on the aluminium box. A delay of
approximately 10 seconds before the current begins to decay; occurs from the moment
the current (and hence power to the resistor) is switched off. Estimating from the plot, it
takes approximately 1300 seconds for the temperature to decrease by 7 degrees
Celsius, having begun to saturate at this point.
Fig. 4: The region of the temperature profile shown in Figure 3 that’s displaying
exponentially decaying behaviour. The time constant is ~260 seconds.
Fig. 5: Time dependence of temperature for the experiment repeated for the lid on top
of the aluminium box. A delay of approximately 40 seconds before the current begins to
decay; occurs from the moment the current is switched off. Over a time span of
approximately 2500 seconds, the temperature has saturated to approximately 25
Fig. 6: The region of exponential decay for the lid on top of the aluminium box. The time
constant is ~640 seconds; a factor of two and a half times larger than the test in which
there was no lid on the identical box. This illustrates the importance in completely
surrounding the temperature sensor in order for it to retain the heat it senses. The
amplitude of this exponential profile is 6.5 °C.
Fig. 7: Time dependence of temperature when the PT100 probe tip was situated in
between two aluminium blocks. A delay of approximately 100 seconds occurred after
the current had been turned off; before the temperature read-out indicated a decrease.
The temperature varies over a range of only 1 degree Celsius over an approximate time
span of 3000 seconds.
Fig. 8: The region of exponential decay is extensively elongated for the arrangement in
which aluminium blocks surround the temperature sensor. The time constant is ~5700
seconds, which is a factor of 9 larger than that for the arrangement with the lid on top
of the aluminum box. This illustrates the effectiveness of the aluminium blocks in
stabilizing the temperature between them over relatively long time scales. The
amplitude is 2 °C, conveying a temperature span of more than three times less than for
the lid on top of the aluminium box. Indeed, the extent of the temperature over the
experimental time scales was limited by the blocks.
The tests carried out show that the blocks are very effective in stabilizing the
temperature in between them, by comparison with the first arrangement in which the
thickness of environmental shielding was very limited indeed. Over a very long time
(over 3500 seconds); the probe measured temperature changes that were just under 1
With the aluminium plates not in their final shapes, only approximations can be
made concerning the effect that the applied current, along with the shielding (from one
test arrangement to the next), had on the temperature variations obtained,
accompanied with the time scales on which these occurred. As long as the masses of he
blocks are not massively reduced in the final arrangement for the voltage supply, we
need to wait ~3 x10³ seconds to get a temperature increase of approximately 1 degree
Celsius, for a heater having a power in the region of 25W.