Key takeaways
- A braided shunt is a flexible conductor first and a resistor second. Its copper resistance moves with temperature.
- A precision meter needs a separate manganin element, or a temperature compensation model built around the braid.
- Four-wire sensing only works if the sense points are fixed and separate from the current terminals.
- Specify resistance at 20 C, the temperature coefficient and the sense point location, not just the cross-section.
What a braided shunt is
A braided shunt is a short, flexible copper conductor used where a rigid bar would put stress on a terminal, a cell or a measuring element. In a meter it often links the incoming terminal to the current path that the sensing circuit measures. In a battery pack it links cells, modules or a pack terminal to the next stage of the circuit.
The name covers two different jobs. In the first, the braid is a current path and the actual measurement happens in a separate shunt or a current sensor. In the second, the braid is the measuring element: the voltage drop across a known length of braid is used to infer current.
Those two jobs need different specifications. A braid that only carries current can be sized on ampacity, voltage drop and flex life. A braid that is also the resistor must be specified on resistance tolerance, temperature coefficient and the stability of the sense points.
Resistance comes from geometry, not from the braid pattern
The resistance of a braided shunt follows the same rule as any conductor: resistivity times length divided by effective cross-section. The braid pattern changes the effective cross-section because the strands follow a longer helical path and because the braid is not fully dense.
For an illustrative example, take a copper braid 100 mm long with a nominal 50 mm2 cross-section and an effective fill of about 0.75. Using a copper resistivity of about 1.72 x 10^-8 ohm metres, the resistance works out at roughly 46 micro-ohms. At 100 A that is a drop of about 4.6 mV and a dissipation of about 0.46 W.
The numbers are illustrative, but the proportions are useful. A short braid at a high current produces a small voltage drop and a small amount of heat. The heat is the problem, because it changes the resistance that the meter or the battery management system is trying to measure.
The braid angle and the number of strands also affect how the current shares between strands. A braid that is tightly compressed at the terminals has a different current distribution from one that is free to spread. The terminal joint, not the braid body, is often the largest single resistance in the assembly.
Thermal drift: why copper makes a poor precision resistor
Copper has a temperature coefficient of resistance of about 0.0039 per kelvin. A 40 K rise above 20 C therefore raises the resistance by about 16 per cent. A meter that assumes a fixed shunt resistance will read 16 per cent low at that temperature unless it compensates.
Manganin, the usual precision shunt alloy, has a temperature coefficient around 10 parts per million per kelvin, which is roughly 400 times lower than copper. That is why a class 1 or class 2 energy meter normally uses a manganin shunt or a compensated current sensor, not a copper braid, as the measuring element.
Self-heating adds a second loop. Current through the braid raises its temperature, which raises its resistance, which raises the dissipation for the same current. The loop settles at a new temperature that depends on the mounting, the surrounding metal and the air flow. A braid bolted to a large copper bar runs cooler than the same braid hanging in still air.
The practical answer is to measure the braid temperature or to characterise the assembly over its operating range. If the meter has a temperature sensor and a calibration table, a copper braid can be used as a measuring element within a defined accuracy band. Without that compensation, the braid should carry current and a separate element should do the measuring.
Copper is a conductor, manganin is a resistor
If the accuracy target is tighter than a few per cent over temperature, do not use the copper braid as the reference. Use it for the current path and put the measurement in an alloy designed to stay still.
Four-wire sensing and where to put the sense points
A shunt is measured with four wires: two carry the current and two sense the voltage across the resistive section. The sense wires must attach at points that are inside the current terminals, so that the resistance of the terminals and the joint is excluded from the measurement.
On a braid this is harder than on a solid bar because the strands move. A sense wire soldered to the outside of the braid may pick up the voltage of only the outer strands. The usual fix is a small crimped ferrule or a welded tab that captures the full braid section at each sense point, with the two points a fixed distance apart.
The sense leads should be twisted or routed together and kept away from switching currents. In a battery pack, the sense leads run near high-current cables and can pick up noise. A small filter at the measurement input is normal, but it must not slow the response so much that a short-circuit event is missed.
Calibration should be done at the assembly level, not on the braid alone. The terminal joints, the bolt torque and the mounting all change the resistance. A calibration certificate for a bare braid does not prove the behaviour of the finished shunt.
Braided shunts in battery packs
Battery packs use braids for two reasons: they absorb tolerance and movement between cells, and they give a short, low-inductance path. A pack that expands and contracts with temperature and state of charge will fatigue a rigid bar. A braid with enough free length takes that movement without loading the cell terminals.
The measurement job in a pack is usually done by a dedicated shunt in the pack disconnect unit or by a hall sensor. The braid may be in series with that shunt, and its resistance adds to the total pack resistance. That is fine for protection and state estimation, but it must be included in the model.
A braid that is also used as the measuring element needs a stable sense point and a temperature input. The battery management system already measures temperature, so the compensation is possible. The risk is that the braid temperature differs from the temperature sensor location, especially during a high-current event.
Fusing is a separate question. A braid is not a fuse, and a small braid that overheats under fault current is not a controlled protective device. If the pack needs a fuse, specify a fuse.
Conductor duty compared with measuring duty
The table sets out the specification differences between a braid that only carries current and a braid that is part of the measurement.
| Requirement | Braid as current path | Braid as measuring element |
|---|---|---|
| Material | Copper, tinned or bare | Copper with a defined temperature coefficient |
| Resistance tolerance | Loose, often plus or minus 10 per cent | Tight, with a calibration at 20 C |
| Temperature coefficient | Not usually declared | Must be declared and compensated |
| Sense points | Not required | Fixed ferrules inside the current terminals |
| Calibration | Not normally needed | Assembly level, with the terminals fitted |
| Thermal design | Keep below the insulation rating | Keep the temperature rise small and known |
| Typical use | Pack interconnect, door bond, meter link | Precision current measurement |
| Main risk | Overheating at a poor joint | Drift and a wrong current reading |
A single assembly can do both jobs, but only if the resistance and its temperature behaviour are specified and calibrated. The table is a starting point, not a substitute for a measurement plan.
What to specify and test
Start with the current. State the continuous current, the short-time current and the duration, and the ambient temperature inside the enclosure. A braid sized only on the continuous current will run hot during a fault or a fast charge.
State the resistance at 20 C, the measurement method and the length between sense points. If the braid is a current path only, a resistance per metre is enough. If it is a measuring element, ask for a calibrated resistance and a temperature coefficient.
For the joints, state the terminal material, the plating, the bolt size and the torque. Ask for a pull test or a cross-section check on crimped ends. A braid that pulls out of its ferrule is an open circuit waiting to happen.
For the finished assembly, ask for a four-wire resistance measurement, an insulation resistance check if there is a sleeve, and a thermal cycle between the lowest and highest service temperatures. The cycle should include a resistance measurement before and after, so that a joint that loosens shows up.
- Continuous, short-time and fault current
- Ambient temperature and mounting arrangement
- Resistance at 20 C and the sense point distance
- Temperature coefficient, if used for measurement
- Terminal material, plating and torque
- Crimp pull test or weld inspection
- Thermal cycle with resistance before and after
Assembly and routing notes
Keep the braid as short as the movement allows, but not so short that it is under tension. A braid that is stretched between two terminals has no margin for tolerance and will load the joints as the assembly heats and cools.
Support the braid so that it does not rest on a sharp edge or vibrate against a housing. A sleeve helps with abrasion and with dielectric clearance, but it also traps heat. If the braid runs warm, check the temperature with the sleeve fitted, not on a bare sample.
Finally, record the torque and the resistance at build. A braided shunt is a simple part, but its performance depends on the joint and on the calibration. The records are what let you tell a drift from a loose bolt.
Frequently asked questions
Can I use a copper braid as a meter shunt?
Only with temperature compensation and a calibration at the assembly level. Copper resistance changes by about 0.39 per cent per kelvin, so an uncompensated braid will drift well outside a class 1 or class 2 meter accuracy over its temperature range. A manganin shunt is the usual choice for the measuring element.
Why does my shunt resistance read higher than the calculation?
The calculated value usually covers the braid body only. The terminals, the crimp or weld, the bolt joint and the sense point location all add resistance. Measure the finished assembly with a four-wire method and compare like with like.
Where should the sense wires attach on a braid?
At fixed points inside the current terminals, using a ferrule or tab that captures the full braid section. Attaching to the outer strands only gives an unstable reading. Keep the two sense points at a known distance and calibrate that distance.
Does tinning a braided shunt change its resistance?
The tin layer is thin and its effect on the braid body is small. The bigger effects are the joint preparation and the terminal plating. If the shunt is a measuring element, calibrate the finished tinned assembly rather than assuming the bare copper value.
Sources and further reading
- IEC 62053-21, Electricity metering equipment, particular requirements for static meters for AC active energy
- IEC 60068-2-14, Environmental testing, change of temperature
- ASTM B172, Standard Specification for Rope-Lay-Stranded Copper Conductors Having Bunch-Stranded Members
Qumek Engineering Desk
Engineering team
The engineering team behind Qumek drawings, DFM notes, tolerances and inspection standards. Every guide is reviewed against the drawings and processes used on live programmes.
Have a shunt that drifts or reads high?
Send the assembly drawing, the current profile and the accuracy target. The engineering desk will review the sense points, the joint and the thermal path.
Send a drawing