Inverting amplifier

Op-amp inverting amplifier

Diagram: Inverting amplifier
Ω
Ω
%
%

SI prefixes accepted (4k7 / 1M / 10m / 220). Upper-case M = mega, lower-case m = milli

Gain
—×
Gain
—dB
History
Press Enter in any field to keep a snapshot here.

Formula

Gain (×): G = Rf / Rs
Gain (dB): G[dB] = 20 × log₁₀(Rf / Rs)
The output is inverted with respect to the input

Design notes

The inverting amplifier is everywhere: amplifying sensor signals, scaling to an ADC's full scale, building active filters. Its appeal is that the gain depends on nothing but a ratio of two resistors.

Design notes:
• The input impedance is Rs — pick it with the source in mind
• The more gain you ask for, the less bandwidth you get (the gain-bandwidth product is fixed)
• Where gain error matters, use 0.1% resistors

For a non-inverting amplifier the gain is 1 + Rf/Rs instead.

When you need this

For setting the gain of an inverting op-amp stage. The circuit is simple — gain is a resistor ratio — but you still need to check the gain error once tolerance is included, and whether the op-amp can actually deliver that gain over your bandwidth.

Gain is set purely by the ratio

The gain is −Rf/Rs. The op-amp itself barely enters into it, which is exactly why the topology is popular: resistor accuracy becomes gain accuracy.

The output is inverted relative to the input. This calculator reports the magnitude, so remember the sign. On a single supply, bias the non-inverting input to set the output midpoint.

How tolerance propagates

Because gain is Rf/Rs, the worst case is the two resistors erring in opposite directions. Two ±1% parts give up to about ±2% gain error — twice the effect it has in a divider.

Higher gains spread the two values further apart. Gain of 100 with Rs = 1kΩ means Rf = 100kΩ, which is the region where input bias current offset starts to matter.

What to check

  • Gain-bandwidth product. Gain times bandwidth is fixed. A 1MHz GBW op-amp at a gain of 100 leaves only 10kHz of usable bandwidth. Choose a part where required bandwidth × gain is at most a fifth of GBW.
  • Input impedance is Rs. The inverting input is a virtual earth, so the stage presents Rs to the source. For a high-impedance source, raise Rs or use a non-inverting stage or a buffer.
  • Add a balancing resistor. Conventionally, put Rs in parallel with Rf between the non-inverting input and the reference to cancel input bias current offset. With CMOS-input op-amps the bias current is tiny and the resistor's own noise can be the bigger problem, so it is sometimes omitted.
  • A small capacitor across Rf stops oscillation. Input capacitance working against Rs adds phase shift. A few to a few tens of picofarads across Rf stabilises it, at the cost of bandwidth.

Common mistakes

  • Using it single-supply referenced to ground. The output inverts, so a positive input drives the output below ground and it saturates. On a single supply, bias the non-inverting input to something like Vcc/2.
  • Making the resistors too large. In the megohm range, thermal noise and input bias current become significant. For low noise, keep the values down while holding the ratio.
  • Exceeding the output swing. Even rail-to-rail outputs do not quite reach the rails. Check that the gained-up signal stays inside the supply at maximum input.

Frequently asked questions

Why is the gain negative?
The signal enters the inverting (−) input, so a rising input produces a falling output — a 180° phase reversal. For the same phase, use a non-inverting stage, which has high input impedance but cannot have a gain below one.
Is there a point to unity gain (Rf = Rs)?
Yes, when you want phase inversion or are generating a differential signal. If you only want a buffer, a non-inverting voltage follower uses fewer parts and has higher input impedance.
How do I read the gain in dB?
It is a voltage ratio, so 20·log₁₀(Rf/Rs). Two times is about 6dB, ten times is 20dB, a hundred times is 40dB. dB makes it easier to compare against the open-loop gain and GBW curves in the datasheet.
Why does gain fall at high frequency?
The op-amp's open-loop gain falls with frequency, and you need ample open-loop gain to hold a given closed-loop gain. Above GBW divided by your gain, the amplifier can no longer sustain it. That quotient is the practical upper frequency limit.

Standards and references

  • IEC 60747-5 — Characterisation of semiconductor devices including operational amplifiers.

Last updated: 2026-08-29