Question: What happens when you exceed the power rating of a resistor?
This is a question that I asked myself while I was replacing a resistor in a battery charger. I calculated that the current would exceed the rating of the resistor, but I never saw smoke. So what really happens when you exceed the power rating of a resistor?
I used the scientific method on this project. I have my question, and now I need to form a hypothesis.
Hypothesis: If the power rating is exceeded then the resistor’s value will fall out of tolerance.
The next step was to experiment. I mapped out the variables that I would encounter: room temp, voltage, current, resistance, power, power rating. I ignored room temperature. It was 70 degrees F, and it didn’t change. Everything else was controlled. The voltage should remain constant. Everything else in this experiment will change depending on what happens.
Formulas: V=IR, I=V/R, R=I/V, P=IV, P=(V^2)/R
I chose to use .25W, 47 ohm, 5% tolerance resistors for this experiment. My power supply (which I built in an earlier project) could only supply 1.5A of current, so I had to pick something that would easily be overstressed without stressing the power supply. A 47 ohm resistor can be stressed to over 8 watts at just 20 volts at less than .5 Amps. This fits the bill for this experiment.
Before each experiment I measured the resistance of each resistor and gave it a number. All fourteen resistors were within the 5% tolerance. The range of resistances tested was from 46.8 ohms to 47.3 ohms.
The first experiment would be to run current through the resistors for 1 minute. After each stress test, the resistances would be measured to see if they still fall within the 5% tolerance. The voltages that I tested the first resistors with were 3.4V, 5V, 6.85V, 10V, and 15V.
After one minute, the only resistors to fail completely were the 2 that were stressed at 15 volts. They were subjected to 353mA and were forced to dissipate 5.3 watts. They smoked and burned out. I captured one of these on video and graphed the values at one-second intervals. Click here to view the PDF of values and a graph.
Observations: .25W resistors can withstand 1 minute at 9 times their rating. Resistor #4 was subjected to 10V. The current was 226mA and power was 2.26W. It was slightly charred, but still remained in tolerance. As resistors degrade by exceeding the power rating, the resistance ultimately increases.
The second test involved four resistors. This time the resistors would have to endure 2.5 minutes. The voltages that I tested these with were 3.4V, 5V, 6.85V, and 10V.
The second test proved that resistors can withstand a great deal of overpower. Resistor #9 was subjected to four times its power rating, and it still stayed within tolerance, and it did not discolor. Resistor #10 was subjected to 10V just as resistor #4. The current was 220mA and power was 2.2W. It was charred, but still remained in tolerance although the resistance did decrease slightly from 47 ohms to 46.9 ohms.
Observations: As the resistor first begins to degrade, the resistance falls slightly.
A third test was conducted. This last test subjected the resistors to five minutes of over current. The same voltages were used as in test two.
The same results were seen in test three that were seen in test two. Resistor #13, even though it was at 1.01W, still remained unchanged in both color and value. Resistor #14 was subjected to 10V. The current was 220mA and power was 2.2W. The resistor was charred and changed its resistance from 47.1ohms to 46.9ohms.
Conclusion: It is never wise to exceed a power rating. Resistors dissipate power as heat, and that heat can be intense enough to start a fire.
The resistance of a resistor seems to drop before the resistor completely fails. After a lot of smoke, and possibly a flash as the resistor burns, the resistance increases well beyond the tolerance. Depending on how long the resistor was subjected to high power, the resistance went from 47 ohms to as high as 1.5kOhms.
Choose resistors that match their given power rating to that of the circuit’s maximum power. Even if they will last some time while operating above their rating, it is not safe to hope that they do not fail.
Click here for a PDF copy of my notebook for this experiment.
3-25-2010: I finally found a link to a similar project. http://www.allaboutcircuits.com/vol_6/chpt_2/7.html
The only difference is that they had a thermometer to take the temperature of the resistor.