Temperature Monitor Report Team: David Smith (email@example.com) Submitted 11/23/99 2:20PM Tuesday class Goal: To build a temperature sensor with visible LED gauge and flashing alarm. Block Diagram: Thermistor Voltage Rising Comparators Final red loses divider voltage light LED’s LED is resistance as changes triggers to visually connected to temperature resistance to comparators represent the a 555 timer rises voltage one by one rising to flash when voltage active Circuit Diagram: +5V Thermistor +5V 10,000 100 +5V +5V +5V +5V +5V ~5358 1M 100 1M 100 1M LM339 IRF513 LM339 IRF513 LM339 IRF513 +5V +5V +5V ~4570 ~4848 ~5160 +5V +5V 1M 1,000 +5V 100 1 8 330 F 555 Timer 2 7 LM339 IRF513 3 6 +5V 4 5 1,000 +5V ~5520 Description: The voltage divider, in the upper left corner of the circuit diagram, is used to translate the resistance of the thermistor to voltage. The 5-volt DC current passes through the thermistor and its compensating resistor, and then through a series resistor to ground. This creates a voltage divider. The node above the thermistor is 5V, the node below the series resistor is 0V, and the node in the center is somewhere in between, depending upon the resistance of the thermistor. The thermistor is on the upper side of the voltage divider and not the bottom so that voltage will rise as temperature rises, and as the thermistor’s resistance drops. The voltage from the voltage divider is then distributed to four comparators. The comparators then compare that voltage with a set voltage created by using multi-turn potentiometers as voltage dividers. If the voltage from the thermistor is greater than the set voltage, the comparator switches. When the comparator switches, the current stops flowing from the pull-up resistor through the comparator to ground, and flows to the MOFSET. When the MOFSET receives power from the comparator, it switches and opens a path from the remainder of the circuit to ground. In the first three comparator- MOFSET blocks, the remainder of the circuit is just an LED, which then lights. The fourth comparator-MOFSET block is connected to a 555 timer. This timer is configured to blink the fourth LED at a rate of about two blinks per second. This, of course, only occurs when the fourth MOFSET has been switched, since the MOFSET is it’s only access to ground. Design Approach: The first item to decide was the temperature range to have the experiment sensitive to. 60-90F was decided upon because it would be the easiest range to generate in the lab for demonstration purposes. Then, the resistance value of the compensating resistor for the thermistor was calculated. The actual result was 10,700, but due to a lack of potentiometers, the impracticality of a decade resistance box, and the relatively low need for accuracy, this was rounded to 10K. The resistance value of the lower resistor in the voltage divider was then calculated. Because the temperature range was from 60-90F, it was safe to assume 75 as a good average temperature. In order to get the greatest precision, this average temperature would have to create a voltage of 2.5V on the voltage divider. Simply figuring out the resistance of the thermistor and compensating resistor would complete the calculation. Setting the bottom resistor equal to that value gave the desired voltage at that temperature. The next calculation was the reference voltage for each of the comparators. Since the desired result was for the first comparator to switch at 60F, and then the next three at 10 intervals, I had to calculate the voltage coming off the voltage divider at those temperatures, and then set the reference voltages to match. I used the equation, T=100-70*(Runknown-R100)/(R30-R100). I plugged in T and solved backwards for the unknown resistance. Then I used the voltage division equations to calculate the voltage coming off the voltage divider at each switching temperature. Then, I needed to calculate the settings for each LED’s potentiometer. I used the ratio of the calculated voltage over 5 and set it equal to the desired resistance over 10,000 (the max of the potentiometer). The pull-up resistors were set at 1M because the circuit they are on is only used for voltage, and any lost current is wasted. I put a 100 resistor in series with every LED to take a little bit of the load off the LED and MOFSET without dimming the LED too badly. As for the two resistors and the capacitor on the timer, it was more or less trial and error. As I only knew the concept of the purpose of those components, and not the method for calculating their exact values, I picked one set to start with, and then worked from there until I had the blink rate and duty cycle I desired (or thereabouts). Test Approach: I had the majority of the test supplies I needed in my dorm room. I used an old AT computer power supply to provide good constant unlimited 5VDC power to operate the circuit, and I have access to a multimeter. I used the multimeter to test the actual voltages and resistances of the circuit (I didn’t bother with the capacitor). As each subsection was assembled, it was tested. Due to the lack of an accurate thermometer, the accuracy of the temperature readings may be slightly less than perfect, but the circuit did function exactly as expected, under estimated temperatures. As the voltage passed each turn-on voltage, each comparator did switch and each LED properly lit. Data: Temp. 60 70 80 90 Resistance of 6366 5694 5022 4349 thermistor/comp. Divider Vltg. 2.285V 2.424V 2.58V 2.76V LED Pot. 4570 4848 5160 5520 Setting Resistance of compensated thermistor at 30F: 8383 Resistance of compensated thermistor at 100F: 3677 T=100-70*(R-3677)/(8383-3677) Voltage/5V=Resistance/10,000 Conclusions: The experiment went better than I had originally expected. Most parts functioned on the first try. Those that didn’t weren’t hard to fix. I did not have adequate equipment to tell exactly if the LED’s were switching on at the right time, but they did work as they should conceptually, and they followed estimates very closely. Discrepancies could have originated from the use of a rounded resistor for compensating, error or change in the potentiometers, the fact that the voltage differences being measured were very small, or error in the parts. What caused me the most trouble was calculating the correct potentiometer values for each LED, and figuring out the correct values for the timer. If I were to do this again, I would probably get a more accurate thermometer for verifying accuracy of the circuit, I would calculate the exact values for the timer, and I would strive for more accuracy in the “cut corners” due to lack of parts.