A DC shunt resistor is a small-value, precision resistor that is placed in series with a load so the current through that load can be determined. The voltage across the DC shunt is measured and Ohm’s law lets us find the corresponding current. The environment, physical orientation, electrical connection, and the current through the DC shunt all have a bearing on the accuracy and the longevity of a DC shunt.

It sounds simple, but I’ve seen plenty of engineers puzzled by the details. I saw it in students during my time as a professor of electrical engineering and electronics engineering technology, and I saw it in my colleagues at Sundstrand Corporation, an aerospace company, where I split my time working as an engineer. My split personality earned me the moniker “Professor” at Sundstrand.

Joe, a former cubemate, was one such puzzled engineer. He was working on a DC converter project in which the load current was to be monitored using a 50-ampere, 50mV DC shunt. He wanted the calibration laboratory to verify the resistance of the shunt and learn the calibration cycle. He informed the calibration laboratory technician the shunt would be exposed to a maximum current of 40 amperes, which is 20 percent below the shunt’s full-scale current.

The calibration technician informed Joe that DC shunts are not calibrated and therefore do not have a calibration cycle. Further, he told Joe that a current of 40 amperes is too large for a 50-ampere DC shunt. Seeing Joe’s look of confusion, he suggested that Joe “ask the professor.”

Joe had many questions for me.

You are likely to be told about the shunt materials when you talk to a distributer or manufacturer. Most DC shunts are made from manganin, a trademarked name for an alloy of a copper, manganese, and nickel. Manganin has the unusual but desirable property that its resistance does not vary significantly with temperature. Aging effects are also negligible since manganin has excellent long-term stability. That is why they do not have a calibration cycle.

Shunts are sized to produce either 50 mV, 75 mV or 100 mV when their rated full-scale current is applied.

Most shunt manufacturers offer manganin shunts with an accuracy of ± 0.25%. This is standard, but higher accuracies like ± 0.1% are available on special order. This tolerance is that of the shunt resistance at room temperature (usually 25^°C). However, because the shunt voltage is given by the product of the shunt current and the shunt resistance, the tolerance is usually specified in reference to the full-scale voltage of 50, 75, or 100 mV. While the standard shunt tolerance is ± 0.25%, at its rated (but short duration) current and measured at room temperature means we can expect full-scale accuracies like 50mV ± 0.125mV or 100mV ± 0.25mV.

Almost all manufacturers constrain the maximum continuous shunt current to be 2/3 of the shunt’s current rating, called its operating current. This means the maximum continuous current a 100A shunt can tolerate is 67A! Therefore, if the maximum, continuous current is 100A, a shunt with a current rating of 150A is required.

Joe exclaimed, “This is just crazy! Who would know this is even necessary?”

I replied calmly, “This is not the role of tribal knowledge and tribal leaders. If you pursue a conversation with a shunt application engineer or a distributor, this will be revealed in short order. They want you to have a successful experience with their product.”

Joe asked for another example. I was happy to oblige. ‘While the shunt current is 2/3 of its full-scale rating, its power dissipation is only 4/9 of its rating. Self-heating effects are made negligible by this reduced power dissipation.”

The transition from manganin to a dissimilar metal like copper or brass creates a thermocouple junction. However, the resulting junction is noted for very low thermal-electric effects. For example, from 0 to 100^°C, a copper-manganin junction generates less than 0.3uV/^°C.

The image that follows illustrates a shunt (electrical switch board type) that connects directly to the power bus bars. Bolts are run through the holes on each end of the shunt. For best cooling by natural convection, the manganin blades should be vertical.

The two small screws on the top of the shunt provide for a kelvin-measurement connection. This eliminates the voltage drops across the bus bar connections.

I told Joe, “When you need the ultimate truth, you need to go to the ultimate power.”

“I didn’t think you were a religious kind of guy,” Joe countered.

I clarified, “Joe, I went to the company calibration laboratory and talked to the manager. I was assured that shunts are not subject to periodic re-certification. His 25 years of experience has shown that aging characteristics of manganin shunts are not significant.”

However, he also indicated that if the shunt experienced an overload that took its temperature to  140^oC or beyond, its resistance can be changed permanently and the shunt must be replaced.

With low-side current sensing, the measurement is taken near ground potential. The common-mode voltage is small. There could be ground noise. In this case an instrumentation amplifier with an active guard drive is a good choice to monitor the shunt voltage. One problem that could arise is DC offset and noise developed by ground loops.

With high-side current sensing, we can get good visibility of the load operation. The issue with this method is that a very large common-mode voltage will exist. In fact, it will be essentially equal to the DC supply voltage. This can introduce measurement and safety concerns.

This a significant problem in electric vehicles. Passenger EVs use a 400V battery bus, while trucks and buses employ a 600V battery bus. However, larger battery bus voltage increases the range of an EV and reduces charging time, so new passenger vehicles are emerging that have an 800V battery bus. Safety and measurement concerns are mitigated by employing either isolated amplifiers or isolated modulators.

“So you see, shunt resistors really are simple after all,” I concluded. Slightly overwhelmed, Joe offered his thanks and left with nearly the same vigor of one of my night students shuffling their way to the parking lot at 10:30 PM.

For more from The Professor, read The Trick to Large Value, Low-Tolerance Capacitors.