RESI Q&A - Answers at a Glance

1. What is Trimming-Free™ technology, and how does it fundamentally differ from laser or mechanical trimming?

Trimming-Free™ is a manufacturing technology developed by RESI for alloy current-sense resistors. In a conventional process, the resistive element is first manufactured close to the target resistance, and its current path is then modified by laser cutting or mechanical machining to bring the resistance into the specified range. With Trimming-Free™ technology, alloy resistivity, material thickness, effective length and cross-sectional area, weld position, and stamping dimensions are controlled together before final forming so that the finished product directly achieves the specified resistance and tolerance.

The fundamental difference is not simply the elimination of one process step, but the method used to control resistance. Conventional trimming relies on localized correction after the part has been manufactured, whereas Trimming-Free™ technology relies on consistency throughout material preparation, welding, and forming. This avoids local necking and current-density concentration caused by trim cuts, helping improve lot-to-lot consistency, pulse capability, and production efficiency. Applicable series and performance specifications are subject to the latest datasheet for each series.

2. Why can conventional trimming create hot spots, reduce current-carrying capability, and affect long-term stability?

Trimming normally narrows the current path by cutting a slot. Near the slot, the effective conductive cross-sectional area decreases and the local current density increases. Based on Joule heating, higher current density raises local power density and can therefore create a hot spot. The resulting local temperature rise may limit continuous-current capability, pulse capability, or rated power.

Trimming may also introduce a heat-affected region, residual stress, or geometric discontinuity around the cut edge. After temperature cycling and long-term loading, these localized regions may cause additional resistance drift.

3. How does Trimming-Free™ technology use alloy consistency, electron-beam welding, and precision processing to control both tolerance and TCR?

The fundamental resistance relationship is R = ρL/A, where ρ is the material resistivity, L is the effective current-path length, and A is the cross-sectional area. To achieve the target resistance without post-production trimming, the process must simultaneously control the resistivity determined by alloy composition and heat treatment, the material dimensions, and errors introduced by the weld between the copper terminals and the alloy element.

C&B Electronics' electron-beam welding process features a narrow weld, a small heat-affected zone, and concentrated energy, reducing resistance variation caused by geometric and material changes in the joint area. Precision stamping and forming control the effective length and cross-sectional area, while alloy lot consistency determines the base resistivity and temperature coefficient. Together, these factors determine initial resistance, lot distribution, and TCR. Actual production also requires four-wire measurement, process-capability analysis, and temperature-cycling verification; no single process step can guarantee final accuracy by itself.

4. Why can metal-foil resistors achieve low TCR and high long-term stability, and how does their temperature-compensation mechanism work?

A metal-foil resistor typically bonds a stabilized resistance-alloy foil to a ceramic substrate with a specialized bonding material and patterns the foil into a defined current path. As temperature changes, the intrinsic resistivity of the alloy foil changes. At the same time, differences in the coefficients of thermal expansion of the foil, bonding layer, and ceramic substrate create mechanical strain. By matching the materials and structure, the resistance change caused by this strain can partially offset the alloy's intrinsic temperature effect, resulting in a very low overall TCR.

The foil thickness and structure are relatively stable. With appropriate heat treatment and stress stabilization, long-term drift, noise, and electrostatic sensitivity can also be well controlled. However, low TCR is not identical across all resistance values, packages, and temperature ranges. Selection should therefore be based on the TCR specified for the required resistance and temperature range, rather than only on the lowest value stated for the series.

5. What do initial tolerance, TCR, and load-life stability mean, and why should resistor selection not be based on tolerance alone?

Initial tolerance is the deviation of the measured resistance from the nominal resistance under specified reference conditions. The temperature coefficient of resistance (TCR), commonly expressed in ppm/°C, describes the change in resistance caused by a change in ambient or resistor-element temperature. Load-life stability describes the permanent resistance drift after the resistor has operated for a specified time under defined temperature and power conditions.

For example, a resistor with an initial tolerance of ±0.01% may still produce a total in-circuit error far greater than ±0.01% if its TCR, self-heating, and long-term drift are significant. In engineering design, an error budget should be established from the operating-temperature range, actual power dissipation, and required service life. A conservative assessment can account separately for initial error, TCR × temperature change, aging, and mounting stress. If errors are combined statistically, the independence of the individual error terms must first be confirmed.

6. What is thermal EMF, and how does it affect low-current measurement and zero-offset accuracy?

When a current-sensing circuit contains junctions between dissimilar metals and the junctions are at different temperatures, a thermocouple-like voltage may be generated. This voltage is called thermal electromotive force, or thermal EMF. It can be approximated as Vt = S × ΔT, where S is the effective Seebeck coefficient of the material combination and ΔT is the temperature difference.

The useful signal from a current-sense resistor is V = I × R. At very low resistance or current, the sensing signal may be only a few tens of microvolts. Thermal EMF is then directly superimposed on the signal and creates a zero offset; the equivalent current error is approximately Vt/R. This error can be reduced by selecting low-thermal-EMF materials, maintaining a symmetrical structure at the current and sense terminals, avoiding heat sources on only one side, using Kelvin connections, and applying zero calibration or current-reversal measurement to remove fixed offsets. Published data for the RTCS series specifies thermal EMF below 0.05 µV/°C, making it suitable for precision measurements that are sensitive to zero-offset error.

7. What are non-inductive and low-inductance resistors, and how does parasitic inductance affect high-frequency and pulse circuits?

A real resistor includes parasitic inductance and capacitance in addition to resistance. At low frequencies, it can be approximated as a pure resistance. As frequency increases, its impedance can be approximated as Z = R + jωL, and the reactance contributed by parasitic inductance causes the impedance magnitude and phase to deviate from the nominal resistance. During fast pulses, the additional voltage V = L × di/dt can also cause overshoot, ringing, or current-measurement error.

Non-inductive and low-inductance designs reduce effective inductance by shortening the current path, using planar or solid-ceramic structures, or arranging windings so that magnetic fields from opposing current directions cancel. Selection should consider the actual frequency, edge rate, and allowable overshoot, together with impedance-versus-frequency curves or specified pulse-test conditions. 'Non-inductive' is a relative engineering term and does not mean that inductance is absolutely zero.

8. How should PZFR and EZFR be selected, and how do they differ in TCR, load-life stability, and designer cost?

PZFR is an ultra-precision foil-resistor series with a resistance range of 5 Ω to 125 kΩ, package sizes from 0603 to 2512, and load-life stability down to ±0.0025%. EZFR is an economical foil-resistor series with a resistance range of 5 Ω to 30 kΩ, package sizes from 0603 to 1206, and long-term load stability of ±0.005%. PZFR should be evaluated first when lower long-term drift or resistance values above 30 kΩ are required. EZFR may be preferred when cost and volume purchasing are higher priorities and its resistance and power ranges meet the application requirements.

9. What are the differences between PCSR and PCSK, and how do tolerance, TCR, and cost affect product selection?

The main difference is TCR: the PCSR2512 series is specified at 10 to 15 ppm/°C, while the PCSK2512 series is specified at 25 ppm/°C.

PCSR should be evaluated first for precision power supplies, instrumentation, test equipment, and BMS applications that require tight control of gain error over the full temperature range. PCSK is generally more suitable when initial error can be absorbed through system calibration, the operating-temperature range is narrower, or cost is the primary consideration.

10. What are the differences between RTCS and ARCS?

RTCS targets high-accuracy, high-current measurement. The series supports rated currents up to 20,000 A, tolerance as tight as ±0.1%, and TCR as low as ±20 ppm/°C.

ARCS targets applications where cost, size, and weight are higher priorities. The series supports rated currents up to 1,400 A, tolerances from 0.5% to 5%, and a TCR of ±100 ppm/°C.

11. How do thin-film and metal-foil resistors differ?

Thin-film resistors use a deposited metal film as the resistive layer. High-performance thin-film resistors can achieve 0.01% tolerance, 2 ppm/°C TCR, 0.01% long-term stability, and a current-noise index of -25 dB. They are widely available and relatively economical; commonly used grades offer 0.1% tolerance and 10 to 25 ppm/°C TCR.

Metal-foil resistors use an alloy foil, substrate, and stress-compensation structure. High-performance specifications can reach 0.001% tolerance, a typical TCR of 0.05 ppm/°C, 0.0025% long-term stability, and a current-noise index of -40 dB. They offer exceptional performance but are more expensive and may have longer lead times. They are suited to metrology, calibration, and high-end instrumentation requiring ultra-high performance.

12. Do shunts and SMD current-sense resistors use the same measurement principle?

Yes. Both use a known low resistance to generate a voltage drop according to V = I × R, from which an amplifier or ADC calculates the current. There is no universally defined current boundary between a 'current-sense resistor' and a 'shunt.' In engineering practice, small components soldered directly to a PCB are generally called current-sense resistors, while bolt-mounted or busbar-mounted components intended for higher currents are generally called shunts.

13. How should the resistance of a current-sense resistor be determined, and how can signal level, I²R loss, temperature rise, and signal-to-noise ratio be balanced?

The resistance can first be estimated from the full-scale input voltage of the amplifier or ADC: R = VFS/Imax. For example, if the maximum current is 100 A and the desired full-scale voltage drop is 50 mV, the initial resistance selection is 0.5 mΩ. The continuous full-load dissipation is then I²R = 5 W. The signal at the normal operating current must also be clearly greater than amplifier offset, circuit noise, thermal EMF, and ADC quantization error.

A lower resistance reduces voltage drop and power loss but also reduces the useful signal, making the system more sensitive to amplifier offset, contact resistance, and PCB trace resistance. A higher resistance produces a signal that is easier to measure but increases power dissipation, temperature rise, and system-efficiency loss. The final design must verify continuous power, peak pulse energy (∫i²R dt), ambient-temperature derating, thermal resistance, and Kelvin-sense layout, while retaining adequate design margin.

14. How should the total error budget for BMS current sensing be established?

The current measured by a BMS can be approximated as Imeas = Vsense/Rsense. On the resistor side, the error budget should include initial tolerance, TCR, resistance change caused by self-heating, long-term drift, thermal EMF, solder-joint resistance, and busbar contact resistance. The signal chain must also account for amplifier input offset and drift, gain error, common-mode rejection, ADC reference error, quantization noise, sampling synchronization, and residual zero-calibration error.

Zero-offset error and gain error should be calculated separately. Zero-offset error dominates at low current and is mainly caused by offset, thermal EMF, and noise. Gain error dominates at high current and is mainly caused by resistance deviation, TCR, amplifier gain, and the voltage reference. Kelvin sensing should be used so that the sense traces connect directly to the specified sensing points and do not share voltage drop with high-current copper paths. Finally, the complete system should be calibrated and validated at low, room, and high temperatures and at multiple current levels. The tolerance of a single resistor must not be treated as the accuracy of the complete system.

15. When is a precharge resistor required in a humanoid robot BMS, and which voltage, energy, and pulse parameters should be evaluated during selection?

When a battery energizes joint drives, inverters, or power modules with substantial DC-link capacitance, closing the main contactor directly can create an inrush current that may damage the contactor, capacitors, or connectors. A precharge resistor or active inrush-current limiter is therefore normally required. A separate precharge resistor may not be necessary if the system already uses controlled MOSFETs, a soft-start power supply, or another current-limiting architecture.

For an initial estimate, R = V/Ilimit can be used to limit the initial current. When a capacitor is charged from zero, the energy absorbed by the resistor is approximately E = ½CV². The time required to charge to a target voltage Vc is approximately t = −RC ln(1 − Vc/V). Actual selection must also consider the maximum battery voltage, DC-link capacitance tolerance, allowable precharge time, restart frequency, fault retries, ambient temperature, and the resistor's single-pulse capability curve. Rated power is a long-term average and cannot replace verification of single-pulse energy.

16. How can temperature rise in passive cell-balancing resistors be reduced?

In passive balancing, the balancing current is often set to approximately 100 mA. After selecting an appropriate resistance for the cell voltage and target current, choose a sufficiently large package and an adequate power rating. Multiple resistors may also be connected in parallel to distribute dissipation and reduce the power and temperature rise of each resistor.

17. How does the CB series current-sensing module differ from a discrete shunt-and-amplifier solution?

With a discrete solution, the system manufacturer must design and integrate the shunt, analog front end, ADC, isolation, power supply, temperature compensation, calibration, and communication interface. The CB series combines a low-TCR shunt with matched sensing, processing, and communication circuits in one module. It uses a 24-bit ADC, an M0 core, and CAN 2.0A/B communication, and integrates temperature compensation, current calibration, and electrostatic-discharge protection. This reduces system-level analog design and production-calibration effort.

18. How should a high-energy pulse resistor be selected, and why must pulse waveform, energy, peak voltage, and repetition rate be considered in addition to average power?

The energy absorbed by a pulse resistor over a short interval can be expressed as E = ∫i²R dt. For capacitor discharge, the maximum stored energy can be initially estimated as E = ½CV². The resulting pulse energy must then be compared with the resistor's specified energy capability, rated power, and overload capability to determine whether it can safely absorb the pulse.

For repetitive pulses, the allowable interval must be evaluated together with ambient temperature, the derating curve, average power, and thermal recovery. The resistor solution should then be verified under the actual pulse sequence.

At a minimum, selection requires the peak voltage, pulse width, energy per pulse, repetition rate, number of pulses, and ambient temperature. Published data for the APLR series specifies a resistance range of 20 Ω to 10 kΩ and a rated power of 6 W. Selected models provide single-pulse energy capability up to 1,000 J and pulse-voltage ratings from 400 V to 5,000 V.

19. Why is maximum working voltage alone insufficient for high-voltage resistor selection, and why must the voltage coefficient of resistance also be considered?

In addition to nominal resistance and power, a high-voltage resistor is limited by its maximum working voltage. The allowable applied voltage is generally the lower of the datasheet voltage limit and √(P × R), with additional derating for ambient temperature. The voltage coefficient of resistance (VCR), which describes the change in resistance with applied voltage, must also be considered. In a precision high-voltage divider, a significant VCR can create ratio error even when TCR is very low.

Published specifications for the HVLR series cover 1 kΩ to 1 GΩ, 2.5 kV to 48 kV, and 0.7 W to 15 W, with tolerance as tight as ±0.1% and TCR down to ±100 ppm/°C. The series uses epoxy encapsulation and 100% high-voltage testing. The 48 kV rating is the maximum available within the series.

20. What structural advantages do TPAN/TPAL planar power resistors offer, and how do PCB or heatsink conditions affect usable power?

TPAN/TPAL resistors use a thick-film resistive layer on an alumina substrate and a metal flange as the primary heat-dissipation path. The planar current path helps reduce parasitic inductance, while TO-220 and TO-263 packages facilitate heat transfer to a heatsink or a large PCB copper area. These resistors are suitable for power conversion, precharge, snubber, load, and pulse-handling applications.

TPAN0220/0263 models are rated at 50 W, while TPAL0220/0263 models are rated at 35 W. Their resistance range is 0.5 Ω to 10 kΩ, with tolerance as tight as ±0.5% and TCR down to ±100 ppm/°C. These power ratings depend on the specified flange temperature, heatsink, copper area, and mounting conditions. Without adequate heat dissipation, allowable power decreases significantly. The design should use thermal-resistance calculations to verify resistive-element or hot-spot temperature and should control the interface material, mounting pressure, and reflow-soldering conditions. Pulse capability must be verified separately against the applicable pulse curve.