Resistor Tolerance and Temperature Coefficient: Avoiding Passive Component Pitfalls
Keywords: resistor tolerance, temperature coefficient, TCR, precision resistor selection, passive component pitfalls
Introduction
Resistors are the most common passive components on any PCB — and arguably the most underestimated. Engineers spend hours selecting op-amps, ADCs, and microcontrollers, then populate the BOM with whatever 1% thick-film chip resistors the distributor has in stock. The result? Circuits that drift out of spec in the field, precision analog stages that don't stay calibrated, and products that fail marginally at temperature extremes.
The two parameters that cause the vast majority of these failures are resistor tolerance (how close the actual value is to nominal at 25°C) and temperature coefficient of resistance or TCR (how much the value shifts per degree Celsius). Together, they determine whether your circuit will hold its accuracy across the operating temperature range — or quietly degrade until something stops working [1].
This article breaks down tolerance classes, TCR behavior across resistor technologies, the circuit-level impact of each, and the most common mistakes engineers make when specifying passive components. We'll also provide a practical selection decision tree to help you avoid these pitfalls in your next design.
Understanding Resistor Tolerance
What Tolerance Actually Means
Tolerance is the maximum deviation from the nominal resistance value at 25°C, expressed as a percentage. A 10 kΩ resistor with ±1% tolerance has an actual value between 9,900 Ω and 10,100 Ω — a 200 Ω spread. That might sound small, but in a precision voltage divider, it directly translates to gain error or reference error.
Tolerance Classes Compared
| Tolerance Class | Typical Technology | Cost Multiplier | Common Applications |
|---|---|---|---|
| ±5% | Thick film chip, carbon film | 1× (baseline) | Pull-ups, pull-downs, general-purpose |
| ±1% | Thick film chip (precision grade) | 1.2–1.5× | Signal conditioning, filtering, LED current set |
| ±0.5% | Thin film chip | 2–3× | Voltage dividers, feedback networks |
| ±0.1% | Thin film chip (high-precision) | 4–8× | Precision references, bridge circuits |
| ±0.01% | Metal foil / wirewound | 10–20×+ | Metrology, calibration standards |
Key insight: Tolerance only describes the value at 25°C. It says nothing about what happens when the resistor heats up — and that's where TCR enters the picture.
Temperature Coefficient of Resistance (TCR)
Defining TCR
TCR, measured in ppm/°C (parts per million per degree Celsius), describes how much a resistor's value changes per degree of temperature shift. A ±100 ppm/°C resistor that is 10,000 Ω at 25°C will shift by:
- 10,000 × 100 × 10⁻⁶ = 1 Ω per °C
Over a 70°C temperature rise (from 25°C to 95°C), that's a 70 Ω shift — equivalent to 0.7% additional error on top of the initial tolerance [2].
TCR by Resistor Technology
| Technology | Typical TCR Range | Tolerance Range | Best For |
|---|---|---|---|
| Thick film chip | ±100 to ±200 ppm/°C | ±1% to ±5% | General-purpose, cost-sensitive |
| Thin film chip | ±10 to ±25 ppm/°C | ±0.1% to ±1% | Precision analog, signal conditioning |
| Metal foil | ±0.5 to ±5 ppm/°C | ±0.01% to ±0.1% | Ultra-precision, metrology |
| Wirewound (precision) | ±5 to ±20 ppm/°C | ±0.1% to ±1% | High-power precision |
| Carbon film | ±200 to ±500 ppm/°C | ±2% to ±5% | Legacy/low-cost designs |
The gap between thick film and metal foil is dramatic — roughly two orders of magnitude in both tolerance and TCR. A design that needs ±0.01% stability simply cannot use thick film, regardless of how tight you screen the tolerance at incoming inspection.
TCR Is Not Always Symmetric
An important subtlety: TCR is often specified as a symmetric figure (e.g., ±25 ppm/°C), but the actual curve is not necessarily linear or symmetric around 25°C. Thick-film resistors, in particular, can exhibit different slopes above and below the reference temperature. A part specified at ±100 ppm/°C might shift +120 ppm/°C between 25°C and 85°C, but −80 ppm/°C between 25°C and −40°C [3]. This matters in designs that must operate across wide industrial temperature ranges.
Temperature Effects on Circuit Performance
Voltage Dividers
In a simple two-resistor divider, the output voltage depends on the ratio of the two resistors. If both resistors have the same TCR, the ratio stays relatively stable — the TCRs cancel. But if one resistor is a different technology (say, a thin-film feedback resistor paired with a thick-film bias resistor), the mismatch in TCR directly introduces temperature-dependent error.
For a divider with R1 = 10 kΩ (±25 ppm/°C) and R2 = 10 kΩ (±200 ppm/°C), the output at mid-supply shifts by:
- ΔR1 = 10,000 × 25 × 10⁻⁶ × ΔT = 0.0175 Ω/°C
- ΔR2 = 10,000 × 200 × 10⁻⁶ × ΔT = 2.0 Ω/°C
Over a 70°C rise, R2 shifts by 140 Ω while R1 shifts by only 17.5 Ω. The divider ratio drifts, and the output voltage moves — even though both resistors started at their nominal values.
Amplifier Gain
In non-inverting and inverting op-amp configurations, gain is set by the ratio of feedback resistors. A gain of 10 set with Rf = 9 kΩ and Rin = 1 kΩ requires the ratio Rf/Rin = 9. If both resistors are the same technology and from the same batch, TCR tracking is usually decent (±5–10 ppm/°C mismatch within a batch). But cross-batch or cross-technology pairings can introduce significant gain drift [4].
Best practice: Use resistor networks or matched pairs (often available as dual/quad packages) for critical gain-setting elements. These are manufactured on the same substrate, ensuring tight TCR tracking (typically <2 ppm/°C mismatch).
Filter Cutoff Frequency
RC low-pass and high-pass filters have a cutoff frequency determined by:
f_c = 1 / (2πRC)
If R drifts by 0.7% (from 100 ppm/°C over 70°C), the cutoff frequency shifts by the same amount. In a precision anti-aliasing filter for a data acquisition system, this can push the filter response outside the required band, allowing unwanted signals to alias into the measurement.
Common Pitfalls in Resistor Selection
Pitfall 1: Ignoring Self-Heating
Every resistor dissipates power: P = I²R. This power raises the resistor's internal temperature above ambient — a phenomenon called self-heating. The temperature rise depends on the package size, PCB thermal design, and power dissipation.
For a 0603 chip resistor, the thermal resistance is roughly 80–100°C/W. A resistor dissipating 0.1 W will self-heat by 8–10°C above ambient. With a 200 ppm/°C TCR, that adds 0.2% error — potentially more than the initial tolerance.
Example: A 1 kΩ current-sense resistor in an 0603 package carrying 100 mA dissipates 0.01 W — negligible self-heating. But the same resistor carrying 300 mA dissipates 0.09 W, raising its temperature by ~7°C. With ±200 ppm/°C TCR, the value shifts by 0.14%, adding 1.4 Ω to the nominal 1 kΩ. In a current measurement circuit, that's a direct 0.14% gain error [5].
Pitfall 2: Mismatched TCR in Critical Ratios
When two resistors must track each other precisely — such as in a differential amplifier, instrumentation amplifier, or bridge circuit — using different resistor technologies is a recipe for drift. Even using the same technology from different manufacturers can introduce TCR mismatch, since formulation and process variations affect the TCR curve.
Fix: Specify resistor networks or matched resistor pairs from the same manufacturer, same batch, and same package for all ratio-critical applications. Data sheets for these products typically specify TCR tracking as a separate parameter.
Pitfall 3: Over-Specifying (Paying for Precision You Don't Need)
Not every resistor needs ±0.1% tolerance and ±5 ppm/°C TCR. A pull-up resistor on a reset line works fine at ±5% and ±200 ppm/°C. The cost difference between a ±5% thick-film and a ±0.1% thin-film resistor can be 5–10×, and across a BOM with hundreds of resistors, that adds up fast.
Rule of thumb: Only specify tight tolerance and low TCR where the circuit actually demands it — feedback networks, voltage references, current sensing, and signal conditioning paths.
Pitfall 4: Using ±1% When TCR Dominates
A common mistake is specifying ±1% thick-film resistors thinking that's "precise enough," while ignoring the fact that over temperature, the TCR contributes far more error than the initial tolerance. A ±1% resistor with ±200 ppm/°C TCR operating over a 70°C range has:
- Initial tolerance: ±1.0%
- TCR contribution: ±1.4%
- Total worst-case error: ±2.4%
A ±0.5% thin-film resistor with ±25 ppm/°C TCR over the same range:
- Initial tolerance: ±0.5%
- TCR contribution: ±0.175%
- Total worst-case error: ±0.675%
The thin-film part costs 2–3× more but reduces total error by 3.6×. Whether that tradeoff is worth it depends on the application — but you can't make that decision without understanding both numbers.
Resistor Selection Decision Tree
Step 1: Identify the Circuit Function
Is this resistor in a signal path (gain setting, filtering, current sensing, voltage reference) or a non-critical path (pull-up, pull-down, LED current limiting, decoupling)?
- Non-critical → Use standard ±5% or ±1% thick film. Stop here.
- Signal path → Continue to Step 2.
Step 2: Determine the Total Error Budget
Calculate how much resistance error the circuit can tolerate without violating its performance spec. Include both initial tolerance and TCR over the operating temperature range.
- Budget > 2% → ±1% thick film (±100–200 ppm/°C) is sufficient.
- Budget 0.5%–2% → ±0.5% thin film (±25–50 ppm/°C).
- Budget 0.1%–0.5% → ±0.1% thin film (±10–25 ppm/°C).
- Budget < 0.1% → Metal foil (±0.01%–0.1%, ±1–5 ppm/°C) or resistor network.
Step 3: Check Power Dissipation and Self-Heating
Calculate P = I²R or P = V²/R. Estimate self-heating based on package thermal resistance. Verify that self-heating-induced TCR error stays within budget. If not, either:
- Upsize the package (e.g., 0603 → 0805 → 1206) to reduce thermal resistance
- Choose a lower-TCR technology
- Reduce the current (if the circuit allows)
Step 4: Check TCR Matching Requirements
If the resistor is part of a ratio-critical pair or network, verify that TCR tracking is adequate. Use matched resistor networks (dual/quad packages) for best tracking.
Step 5: Verify Voltage Coefficient (for High-Voltage Circuits)
At high voltages (>50 V across the resistor), thick-film resistors exhibit a voltage coefficient of resistance (VCR) — typically 0.1–1 ppm/V. This is separate from TCR but causes similar drift. Thin-film and metal-foil resistors have much lower VCR (<0.1 ppm/V) [6]. If your circuit involves high-voltage sensing or divider applications, add VCR to your error budget.
FAQ
What is the difference between resistor tolerance and TCR?
Tolerance describes the maximum deviation from the nominal resistance value at 25°C (room temperature). It is a static, one-time specification. TCR (temperature coefficient of resistance) describes how much the resistance value changes per degree Celsius of temperature shift, expressed in ppm/°C. Tolerance tells you where you start; TCR tells you how far you drift. Both must be considered together to determine total error over the operating temperature range.
How do I calculate total resistor error over temperature?
Total worst-case error = |initial tolerance| + |TCR × ΔT × 10⁻⁶|, where ΔT is the temperature change from 25°C to your operating extreme in °C, and TCR is in ppm/°C. For example, a ±1% resistor (10,000 ppm) with ±100 ppm/°C TCR over a 70°C range: 1% + (100 × 70 / 10,000) = 1% + 0.7% = 1.7% total error. Note that this is worst-case; RSS (root-sum-square) statistical analysis may yield a lower practical estimate.
When should I use metal foil resistors instead of thin film?
Metal foil resistors (e.g., Vishay Bulk Metal Foil) are justified when you need ultra-precision stability — typically TCR below ±5 ppm/°C and tolerance below ±0.05%. Applications include metrology-grade instruments, calibration references, ultra-high-precision ADC/DAC feedback networks, and circuits that must maintain accuracy over very wide temperature ranges (−55°C to +125°C or beyond). They cost 10–20× more than thin film, so reserve them for applications where nothing else meets the spec.
Does resistor package size affect TCR?
Package size itself doesn't directly change the TCR specification (which is a material property), but it strongly affects self-heating. A smaller package has higher thermal resistance to the PCB, meaning the resistor junction runs hotter for the same power dissipation. Since the resistor's actual temperature determines the TCR-induced shift, a smaller package effectively increases the temperature-driven error. Going from 0603 to 0805 can reduce self-heating by 30–40%, and 1206 by 50–60% compared to 0603.
What is TCR tracking and why does it matter?
TCR tracking refers to how closely two or more resistors' TCR curves match each other. In ratio-critical circuits (voltage dividers, differential amplifiers, bridge sensors), what matters is the ratio between resistors, not their absolute values. If both resistors drift by the same amount, the ratio stays constant. TCR tracking is specified for resistor networks manufactured on the same substrate — typically <2 ppm/°C mismatch for thin-film networks, compared to ±25 ppm/°C individual TCR. Using matched networks is the most cost-effective way to achieve ratio stability.
Can I use thick-film resistors for precision circuits if I calibrate them?
Calibration can remove the initial tolerance error, but it cannot compensate for TCR drift in the field. If your circuit operates in a controlled temperature environment (e.g., inside a temperature-controlled enclosure), thick-film resistors may be acceptable after calibration. However, in any application where temperature varies — which includes most real-world products — the TCR-induced drift of thick-film resistors will cause the calibrated accuracy to degrade. For field-deployed precision circuits, use thin-film or better, and calibrate to remove residual tolerance error.
External References
- Vishay - Basics of Linear Fixed Resistors — Comprehensive primer on resistor tolerance, TCR, and selection criteria.
- KOA Speer - Temperature Coefficient of Resistance — Technical note explaining TCR measurement and specification.
- Texas Instruments - Precision Labs: Resistors — Video series on resistor behavior in precision analog circuits.
- Yageo - Thick Film Chip Resistor Datasheet — Typical specifications for general-purpose thick-film resistors.
- Vishay Foil Resistors - Bulk Metal Foil Technology — Application notes and specifications for ultra-precision metal foil resistors.
References
[6] Analog Devices, "MT-064: Op Amp Settling Time," Tutorial, 2023. (Referenced for voltage coefficient discussion in precision resistor selection.)