Can you combine resistors to make a value you cannot buy?
Yes: combine them in series (R1 + R2) or in parallel (R1 × R2 ÷ (R1 + R2)) and pick the pair that lands closest. For 12.345 kΩ from E24, 13 kΩ ∥ 240 kΩ gives 12.332 kΩ, −0.105 % off, where the nearest single resistor (12 kΩ) is −2.79 % off. With ±1 % parts the pair can be at most ±1.1 % from the target.
Resistor combination formulas
The calculator ranks by error the pairs of standard values that land within 5 % of the target. Series parts go down to 1/1000 of the target and parallel parts up to 1000 times it, since a more extreme trim moves the result by less than the tolerance of a ±0.1 % part. For a target from 1 Ω to 10 MΩ both parts also stay in that range.
Series: R = R1 + R2 Parallel: R = R1 × R2 ÷ (R1 + R2)
- R1, R2
- the two standard resistors, in Ω
- R
- the combined value
The nominal error against the target, and the worst case once each part may be off by ±t:
e = (R − Rtarget) ÷ Rtarget worst case = (1 + e)(1 ± t) − 1
- e
- nominal error of the pair, as a fraction
- t
- part tolerance, as a fraction (0.01 for ±1 %)
The pair is never worse than ±t, because scaling both parts by (1 + t) scales the series sum and the parallel value by exactly (1 + t). If the two parts’ errors are independent, the root-sum-of-squares spread is smaller:
spread = t × √(R1² + R2²) ÷ (R1 + R2)
That is 0.707 t for two equal resistors and close to t when one resistor is much larger than the other, in either connection. Among pairs with the same error, the calculator lists the more evenly matched one first.
Worked examples
Example 1: 12.345 kΩ from E24
The nearest single E24 value, 12 kΩ, is −2.79 % off. The best pair is a parallel one, 13 kΩ ∥ 240 kΩ: 13 × 240 ÷ 253 kΩ = 12.332 kΩ, −0.105 %. The best series pair, 360 Ω + 12 kΩ = 12.36 kΩ, is nearly as close (+0.122 %). With ±1 % parts the parallel pair stays within ±1.1 % of 12.345 kΩ.
Example 2: 2.5 kΩ from E12
E12 jumps from 2.2 kΩ to 2.7 kΩ, so the nearest single value is 2.7 kΩ, which is +8 % off. In series, 1 kΩ + 1.5 kΩ is exactly 2.5 kΩ. With ±5 % parts the pair stays within ±5 %, against ±13.4 % for the single 2.7 kΩ.
Example 3: 1.234 kΩ from E96 with 0.1 % parts
The nearest E96 value, 1.24 kΩ, is +0.486 % off, several times the 0.1 % tolerance. 536 Ω + 698 Ω makes 1.234 kΩ exactly. The two parts are close in value, so if their errors are independent the result spreads only about ±0.0713 % (0.713 of one part’s ±0.1 %).
E96 values from two E24 resistors
When only E24 parts are stocked, these resistor combinations land closest to some common E96 values. Errors are nominal, before part tolerance.
| Target (E96) | Best E24 pair | Result | Error | Nearest single E24 |
|---|
| 1.05 kΩ | 430 Ω + 620 Ω | 1.05 kΩ | exact | 1.1 kΩ (+4.76 %) |
| 1.13 kΩ | 510 Ω + 620 Ω | 1.13 kΩ | exact | 1.1 kΩ (−2.65 %) |
| 1.43 kΩ | 680 Ω + 750 Ω | 1.43 kΩ | exact | 1.5 kΩ (+4.9 %) |
| 2.49 kΩ | 91 Ω + 2.4 kΩ | 2.491 kΩ | +0.0402 % | 2.4 kΩ (−3.61 %) |
| 3.48 kΩ | 180 Ω + 3.3 kΩ | 3.48 kΩ | exact | 3.6 kΩ (+3.45 %) |
| 4.99 kΩ | 5.1 kΩ ∥ 240 kΩ | 4.99388 kΩ | +0.0778 % | 5.1 kΩ (+2.2 %) |
| 6.04 kΩ | 6.2 kΩ ∥ 240 kΩ | 6.04387 kΩ | +0.064 % | 6.2 kΩ (+2.65 %) |
| 8.06 kΩ | 560 Ω + 7.5 kΩ | 8.06 kΩ | exact | 8.2 kΩ (+1.74 %) |
Round values from two E12 resistors
Values that fall between E12 steps, made from the coarsest common series.
| Target | Best E12 pair | Result | Error | Nearest single E12 |
|---|
| 2.5 kΩ | 1 kΩ + 1.5 kΩ | 2.5 kΩ | exact | 2.7 kΩ (+8 %) |
| 4 kΩ | 1.8 kΩ + 2.2 kΩ | 4 kΩ | exact | 3.9 kΩ (−2.5 %) |
| 5 kΩ | 10 kΩ ∥ 10 kΩ | 5 kΩ | exact | 4.7 kΩ (−6 %) |
| 6 kΩ | 12 kΩ ∥ 12 kΩ | 6 kΩ | exact | 5.6 kΩ (−6.67 %) |
| 25 kΩ | 10 kΩ + 15 kΩ | 25 kΩ | exact | 27 kΩ (+8 %) |
| 40 kΩ | 18 kΩ + 22 kΩ | 40 kΩ | exact | 39 kΩ (−2.5 %) |
| 50 kΩ | 100 kΩ ∥ 100 kΩ | 50 kΩ | exact | 47 kΩ (−6 %) |
| 60 kΩ | 120 kΩ ∥ 120 kΩ | 60 kΩ | exact | 56 kΩ (−6.67 %) |
Using a pair on a real board
Power divides with the values: in series each resistor dissipates in proportion to its resistance, in parallel in inverse proportion, so check each part’s rating rather than the total. A pair costs an extra placement and board area, so first check whether a single E96 or E192 value is close enough; 1 % chip resistors are commonly sold in both E24 and E96 values. Parts from the same reel can err in the same direction, which pushes the result toward the worst case rather than the statistical spread. For very high values, board leakage in parallel with the pair matters more than the last fraction of a percent; for very low values, solder joints and trace resistance add to a series pair.
Glossary
- E-series
- The IEC 60063 preferred values: E12 has 12 values per decade, E24 has 24, E96 has 96.
- Nominal error
- How far the pair’s marked value is from the target, before part tolerance.
- Tolerance
- The most a part may differ from its marked value, for example ±1 %.
- Worst case
- Nominal error plus every part at its tolerance limit in the same direction.
- Statistical spread
- The likely spread when part errors are independent: the root-sum-of-squares of each part’s weighted error.
- Parallel (∥)
- R1 ∥ R2 = R1 × R2 ÷ (R1 + R2), always less than the smaller resistor.
Frequently asked questions
How do I make a resistor value that is not standard?
Put two standard resistors in series (their values add) or in parallel (product over sum). For 12.345 kΩ the best E24 pair is 13 kΩ ∥ 240 kΩ = 12.332 kΩ (−0.105 %), against −2.79 % for the nearest single value.
How do I make 5 kΩ from standard resistors?
10 kΩ ∥ 10 kΩ gives exactly 5 kΩ. Two equal resistors in parallel always give half the value, and two in series give double.
Should I combine resistors in series or in parallel?
Whichever lands closer; the calculator tries both. In series a small resistor trims a large one upward, and the larger part dissipates more power. In parallel a large resistor trims a smaller one downward, and the smaller part takes more of the current.
What is the tolerance of two resistors in series or parallel?
No worse than the parts: two ±1 % resistors make a ±1 % result in either connection, because scaling both by 1.01 scales the result by 1.01. If their errors are independent the spread is smaller, down to 0.707 of one part’s for two equal values.
Is it worth using two resistors instead of one?
When the nearest single value misses by more than the tolerance, yes. For 2.5 kΩ with ±5 % parts, the nearest E12 value (2.7 kΩ) can be ±13.4 % off, while 1 kΩ + 1.5 kΩ is within ±5 %. When a 1 % E96 part already exists near the value, one resistor is simpler and cheaper to place.
What is the resistor combination formula?
In series the values add: R = R1 + R2. In parallel it is product over sum: R = R1 × R2 ÷ (R1 + R2), always less than the smaller resistor.
How many values are in the E24 and E96 series?
IEC 60063 defines 24 values per decade for E24 and 96 for E96 (plus E3, E6, E12, E48 and E192). Each step is about 10^(1/n), so E24 values are about 10 % apart and E96 values about 2.4 % apart.
Design the whole board, not one number
ProtoFlow is free AI PCB design software: describe a circuit and get a schematic with real parts, PCB layout with an autorouter, SPICE simulation and Gerber files. Native KiCad, Altium and EAGLE import and export. Free on macOS, Windows and Linux.