The straight forward relationship for electrical power is P=I*V (I=current, V=voltage). Ultimately, all eletrical power measurements involve obtaining the voltage supplied and the current drawn by the load either directly or in a round about way. Variations of the power expression can be derived from simple algebra and Ohm’s law (V=I*R), but in it’s simplest incarnation knowing voltage and current leads to a straighforward derivation of power.

Measuring voltage is easy, just about any digital multimeter over 50 bucks can give a decent, accurate voltage measurement. Those same multimeters, can also measure AC or DC currents as well, but are not built to handle high currents. Many max out at 10 Amps, some others lower, and very few can measure currents in excess of 10 A. For the application of interest here, currents in excess of 10 Amps are not uncommon and necessitates a different approach.

There are actually two commonly used methods for extracting current from a closed loop circuit. One is to insert a shunt, with a resistance much much lower than the impendence of the load. In fact, many ammeters are simply voltmeters with a shunt resistor built internally to the meter. The shunt must be inserted in series with the power source and the load, and the resistance accurately known. By knowing the resistance and measuring the voltage drop across the shunt, the total current passing through the shunt can be calculated (good ole’ Ohm’s Law). Since all current through the shunt must also pass through the load (they are in series after all) the relationship is easily inferred that I(shunt) = I(load).

Another method for measuring current utilizes the concept of the Hall Effect. Simply put, the Hall Effect is a phenomena in which the flow of electrons through a conductor is altered by the presence of a magentic field whose field lines run orthoganol to the direction of the current flow through that conductor. The concept is actually quite easy to understand so long as it is accepted that electrons bend through a magnetic field. Electrons traveling through a conductor (let’s take a rectangular slab of metal sometimes called the Hall Element) subjected to a magnetic field will experience a force and bend preferentially in one direction as a result of that magnetic field (recall your high school/college physics and the right hand rule). This creates a charge seperation within the conductor and, in turn, causes a measurable voltage across the conductor from one side to the other.

The magnitude of the voltage across the conductor is related to the strength of the magnetic field. Ok, so far so good, so how can we use this to measure current? Well, it just so happens that flowing electrons, traveling down a wire, will generate their own magnetic fields. As such that those fields can be sensed by placing a Hall element in proximity and in the right geometry to the wires (conductors) carrying the test current.Hall Effect current measurement is very common, the most obvious to the laymen are the clamp on meters in which the meter is simply a closed loop with the current carrying wires running through the center.

In general, however, Hall Effect meters are considered to be not as accurate as the shunt method described above. There are many reasons for this. First the voltages generated within the Hall Element are typically very small, and difficult to measure; second, the effect depends on the appropriate placement of the magnetic field generators with respect to the Hall Element itself, and finally stray magenetic fields can interfere and cause extraneous noise in the measurement and effect the accuracy. This is not to say Hall Effect measurements are bad, most of these problems can be designed out of the testing system. Using Hall Effect sensors to extract current (and, by simply calculation, measuring Power) has some unique advantages as well. The metering can be done noninstrusively (for the clamp on methods, at least) and, most importantly, the measurement element or sensor is electrically isolated from the power that drives the load in question.

Shunts also come with their own set of disadvantages. The most obvious one is that the circuit must be cut open and the shunt inserted. This, in turn, introduces an otherwise foriegn impedence to the system. This is one reason why the total resistance of the shunt must be kept small compared to the load impedance. Another issue is the magnitude of the current to be measured — for high currents very low resistance shunts must be used, and the resulting voltage drop to be measured becomes very small, on the order of millivolts. In otherwords, to get good measurements you will need much better than the 20 dollar Radio Shack special.

Either method, for the purposes of our endeavor, will work well since microAmp resolution and accuracy is not necessary. Keep in mind, however, that ultimately the goal will be to measure the power drawn from all voltage inputs (simultaneously) as supplied by the PSU. Eight total lines will need to be measured which covers all the motherboard voltages, CPU auxilary, GPU and drives. This makes for an interesting predicament in terms of extracting this information easily and reproducibly.