Today I salvaged out a circuit board from an ESPE Dental Cure light. To my delight there was Kelvin precision resistor on the board. Specifically it was an Isotek PBV R033 which is a heat sink mounted 10 Watt Kelvin resistor with a value of 0.033 Ohms 1%. Here are front and back photos of the device. I first learned about these resistors a couple years ago from one of the great tutorials posted by peteroakes on building power supplies.






Now in the big scheme of things the arrival of a real 4 terminal Kelvin resistor to my bench must seem pretty trivial but this is my first. Yes in 55 + years of tearing equipment and circuit boards apart this is the first time I have had a Kelvin Resistor to test out. What follows is an exercise in simplicity but it is the type of experiment that must be performed on the occasion when something new shows up. In case you would like more information on this particular resistor here is a link to the company data sheet:


I began by producing a new custom symbol for my schematic drawing program. I always like to produce a schematic before an experiment even if it is fairly simple like this one. Here is a copy of the schematic for the test setup:


Kelvin Resistor Experiment.bmp


My plan was to power the circuit at 4 different current levels and take readings of the Current (I), the voltage of the source (Vs), and the voltage across the Kelvin Resistor (Vkr). I would then use my empirical numbers to calculate the resistance (Rk) of the Kelvin Resistor and compare the calculations to the nominal value of 0.033 Ohms 1%. The readings for the voltage of the source would not figure into the calculations but I like to have them for an overview of the system. Keep in mind that perhaps the most accurate instrument in my shop is a Fluke 177 which will be used for DC current readings. Basically none of my equipment qualifies as precision just practical for my level of endeavor. The other equipment that will be used for this experiment are a Mastech power supply which will be the current source and a Radio Shack digital meter that will read the voltage at the test terminals of the Kelvin resistor. The final digit of the Radio Shack meter will read 0.1 mV.  In the chart that follows I have listed the calculated values of (Rk) to several magnitudes of precision past the point justifiable by the precision of the instruments. This was done only to give a better feel for the proximity of the results to the nominal value of (Rk). Ohams Law V = I x R was the basis for the calculations.  Here is the test setup that I used:




The resistor resting on the glass block is a 2 ohm 25 W resistor that is being used as a load. When the circuit was fully powered at 1 Amp the dissipation required was about 2.7 watts. This resistor does not figure into our calculations either.




While not necessary for this experiment, if everything runs smoothly, the Kelvin resistor was stuck to the glass desk top to serve as a heat sink. The Kelvin resistor can dissipate 3 Watts in free air and 10 Watts when properly heat sunk. Since things do not always run smoothly on my bench I try not to take chances hence the heat sinking. Here is the data that was collected from the experiment:


Kelvin Resistor Data.jpg


The nominal value of the Kelvin Resistor is 0.033 Ohms with a 1% tolerance. This means that the actual value should lie between 0.03267 and 0.03333 Ohms. I am pleased that my calculated results also fall within that 1 % range. I will now put the new Kelvin resistor aside for use in my next power supply build.