IP3 and SFDR Calculator
Estimate IIP3 and OIP3 with IM3 clearance and SFDR for an equal-tone RF test based on gain, noise bandwidth and detection margin.| RF quantity | Value | Interpretation | Copy |
|---|---|---|---|
| {{ row.label }} | {{ row.value }} | {{ row.note }} |
What the selected two-tone level means
{{ marginHeading }}
- {{ metric.label }}
- {{ metric.value }}{{ metric.note }}
Calculation path
SFDR = ⅔ × (IIP3 − MDS)
- {{ step }}
This equal-tone, small-signal estimate does not model compression, mismatch, filtering, analyzer overload, ADC clipping, unequal tones, or other distortion orders. Confirm real drive limits and measured P1dB before applying power.
Fundamental, IM3, and MDS crossing
See the 1:1 fundamental slope and 3:1 IM3 slope against the selected detection boundary.
Two strong radio-frequency tones can mix in a nonlinear receiver or device and create new signals close to the originals. The most troublesome third-order products often fall at 2f1 − f2 and 2f2 − f1, where filtering may not be able to separate them from a wanted channel.
Third-order intercept point (IP3) summarizes the small-signal relationship between a fundamental tone and its third-order intermodulation product (IM3). On an extrapolated log-power plot, the fundamental rises 1 dB for each 1 dB increase in input while IM3 rises 3 dB. Their projected crossing is IP3. It is a comparison metric, not a level that should be applied to hardware.
- IIP3
- Input-referred IP3, measured or translated to the input reference plane.
- OIP3
- Output-referred IP3. Small-signal gain separates it from IIP3.
- MDS
- The minimum detectable signal used here: integrated input noise plus the chosen detection signal-to-noise margin.
- SFDR
- Spurious-free dynamic range between MDS and the equal-tone level where predicted IM3 reaches MDS.
Noise bandwidth matters because a wider measurement collects more noise. Raising bandwidth or the required detection margin raises MDS and reduces the modeled SFDR window. A better IIP3 widens the window, while each 1 dB increase in per-tone power reduces fundamental-to-IM3 separation by 2 dB.
The equal-tone model is useful for comparing a receiver chain, mixer, amplifier, or test setup under consistent conditions. Real operation can depart from it because of compression, unequal tones, mismatch, filtering, analyzer distortion, ADC clipping, temperature, bias, and frequency-dependent behavior.
How to Use This Tool:
Keep the IP3, gain, noise, and tone measurements tied to the same device condition and reference planes.
- Choose Input IP3 (IIP3) or Output IP3 (OIP3), then enter the IP3 value and small-signal gain. Use negative gain for a passive loss.
- Select a Noise basis. Enter measured input noise density when available, or use noise figure with the −174 dBm/Hz room-temperature planning reference.
- Enter the physical measurement bandwidth, equal per-tone input power, first-tone frequency, and tone spacing. Changing the bandwidth unit preserves the same physical bandwidth.
- Set a nonzero Detection SNR margin only when detection requires signal above integrated noise. Compare the current IM3 clearance with the maximum clean input tone, then confirm hardware compression and absolute ratings separately.
Interpreting Results:
Input-referred SFDR is the modeled separation between MDS and the per-tone crossing where IM3 equals MDS. A nonnegative window means IIP3 is at or above MDS; it does not prove that the clean-tone crossing is safe to apply.
Current IM3 clearance is MDS minus predicted IM3. Zero is the boundary. A value of 0 dB or more places IM3 at or below MDS, while a negative value places it above the selected detection boundary. Reducing each tone by 1 dB lowers modeled IM3 by 3 dB.
Use the frequency outputs to see where the lower and upper IM3 products land, and use the intercept map to check the 1:1 fundamental slope against the 3:1 IM3 slope. Then compare the proposed drive with measured P1dB, maximum input, analyzer headroom, and filtering at the actual frequencies.
Technical Details:
All power calculations use logarithmic dB or dBm arithmetic at an input reference plane. Gain translates an input-referred quantity to the output plane without changing the dB separation between corresponding quantities.
Formula Core:
First resolve the selected intercept to IIP3 and integrate input noise density across bandwidth B. The detection margin S raises MDS above that integrated noise.
For the noise-figure path, N0 = −174 dBm/Hz + noise figure. For the measured-density path, N0 is the entered input noise density. Bandwidth is converted to hertz before the logarithm.
The SFDR and clean-tone crossing follow from the IIP3-to-MDS separation. These equations assume equal input tones and third-order small-signal behavior.
| Symbol | Meaning | Unit |
|---|---|---|
| G | Small-signal gain from input to output reference plane | dB |
| B | Measurement bandwidth after conversion | Hz |
| S | Detection SNR margin | dB |
| Ptone | Input power of each equal tone | dBm |
Frequency Mechanism:
If the second tone is one spacing Δf above the first, the third-order products fall one spacing below the first tone and one spacing above the second.
Validation Bounds:
| Quantity | Accepted range |
|---|---|
| IP3 and small-signal gain | −200 to +200 dBm for IP3; −200 to +200 dB for gain |
| Noise input | −300 to +100 dBm/Hz for density, or 0 to 100 dB for noise figure |
| Measurement bandwidth | Greater than 0 Hz and no more than 1 PHz after unit conversion |
| Per-tone power | −300 to +200 dBm |
| Tone frequency and spacing | Both must be greater than zero; each is capped at the equivalent of 1 PHz |
| Detection SNR margin | 0 to 100 dB |
| Display precision | 2 to 5 decimal places |
The calculation retains full precision. The selected decimal places affect visible and document-output rounding only. Accepted bandwidth units are Hz, kHz, MHz, and GHz; all frequency spacing is converted consistently before the IM3 locations are calculated.
Accuracy and Safety Notes:
IP3 is extrapolated from small-signal slopes and can lie far beyond the device's compression or damage level. Never use IIP3, OIP3, or the modeled clean-tone crossing as an absolute maximum rating.
- The model assumes two equal continuous-wave tones and one third-order response.
- It omits compression, mismatch, filters, frequency response, unequal tones, harmonics, second-order products, blocker desensitization, ADC limits, and test-instrument distortion.
- Frequency products are calculated algebraically; verify that both IM3 frequencies are positive and fall in the device and measurement passband.
- The −174 dBm/Hz path is a room-temperature planning reference, not a temperature-aware noise calculation.
- Compare results only when reference plane, gain, bandwidth, detection margin, frequency, bias, temperature, and measurement method are consistent.
Worked Example:
Receiver with measured input noise
With IIP3 at 10 dBm, input noise density at −150 dBm/Hz, and 100 kHz bandwidth, integrated noise is −100 dBm. A 0 dB detection margin keeps MDS at −100 dBm, so SFDR is about 73.33 dB and the equal-tone IM3 crossing is about −26.67 dBm per input tone. At −35 dBm per tone, predicted IM3 is −125 dBm, leaving 25 dB of clearance below MDS.
References:
- Use Selectivity to Improve Receiver Intercept Point, Analog Devices.
- Spurious-Free Dynamic Range (SFDR), Analog Devices.
- Noise Figure, Keysight Technologies.