Drop .snp files here
or click to browse — .s1p through .s8p+Loaded Files
Spec Masks
The most expensive engineering failures happen at the boundaries between different teams in a project, rarely inside a team. We find cross-domain blind spots, build models to close them, and prove the fix with data. Autonomous Systems · RF · Manufacturing · Defense · Aerospace.
See the proof →The Pattern Architect Newsletter
Weekly engineering intelligence across four tracks: boundary-aware modeling (why your sim doesn't match your line), systematic leadership (why firefighting won't stop), autonomous hardware (RF + embedded + motion under one roof), and systematic thinking (the meta-skills that compound). Subscribe and get the RAPID Starter Kit free.
Get the Starter Kit Free →Drop .snp files here
or click to browse — .s1p through .s8p+Loaded Files
Spec Masks
The most expensive engineering failures happen at the boundaries between different teams in a project, rarely inside a team. We find cross-domain blind spots, build models to close them, and prove the fix with data. Autonomous Systems · RF · Manufacturing · Defense · Aerospace.
See the proof →The Pattern Architect Newsletter
Weekly engineering intelligence across four tracks: boundary-aware modeling (why your sim doesn't match your line), systematic leadership (why firefighting won't stop), autonomous hardware (RF + embedded + motion under one roof), and systematic thinking (the meta-skills that compound). Subscribe and get the RAPID Starter Kit free.
Get the Starter Kit Free →S-Parameters: The Complete Engineering Reference
Everything you need to interpret Touchstone files, read Smith charts, and diagnose common RF integration failures. Click any section to expand.
S-parameters — scattering parameters — describe how electromagnetic energy propagates through an RF network. Unlike impedance (Z) or admittance (Y) parameters, S-parameters are defined in terms of traveling waves rather than total voltages and currents. This distinction matters because at RF and microwave frequencies, measuring total voltage at a point requires knowing the forward and reflected wave components separately — which is exactly what S-parameters provide.
For a two-port network (the most common case, and what .s2p files describe), there are four S-parameters:
| Parameter | Physical Meaning | What It Tells You |
|---|---|---|
S11 | Input reflection coefficient | How much power is reflected back from Port 1. Measures impedance match quality at the input. |
S21 | Forward transmission coefficient | How much power passes from Port 1 to Port 2. The gain (amplifiers) or insertion loss (passive networks). |
S12 | Reverse transmission coefficient | How much power leaks from Port 2 back to Port 1. Measures isolation in amplifiers, reverse leakage in filters. |
S22 | Output reflection coefficient | How much power is reflected back from Port 2. Impedance match quality at the output. |
The Complex Representation
Each S-parameter is a complex number with magnitude and phase:
Sij = |Sij| · e jφij
The magnitude and phase carry distinct physical meaning:
Magnitude represents the ratio of voltage wave amplitudes. |S11|² gives the reflected power ratio — if |S11| = 0.316 (−10 dB), then |S11|² = 0.1, meaning 10% of incident power is reflected. |S21|² gives the transmitted power ratio — this directly maps to insertion loss for passive networks or gain for amplifiers. For a lossless, reciprocal two-port network, power conservation requires |S11|² + |S21|² = 1. Any deviation indicates loss (absorbed power) or measurement error.
Phase encodes the electrical delay through the network. For a uniform transmission line of length l, the transmission phase is φ21 = −βl, where β = 2πf√εeff/c is the propagation constant. The group delay — the quantity that determines signal distortion — is the derivative of phase with respect to angular frequency:
τd = −dφ/dω
Constant group delay means all frequency components traverse the network in the same time — the signal shape is preserved. Group delay variation across the signal bandwidth causes dispersion: different spectral components arrive at different times, smearing the output waveform. A bandpass filter with 0.1 dB magnitude ripple may still introduce 5 ns of group delay variation at the band edges — for a 100 MHz wide signal, this can be the dominant source of intersymbol interference.
Key insight: S-parameters are measured under matched conditions — each port is terminated with the reference impedance Z0 (usually 50Ω). If your actual system impedance differs from Z0, the S-parameters alone don't predict how the component will behave in-circuit. You need to de-embed (remove fixture effects) or re-normalize (transform to a different reference impedance). The relationship between S-parameters and impedance is: Zin = Z0 · (1 + S11) / (1 − S11).
Magnitude Plots (dB)
The magnitude of an S-parameter is plotted in decibels: |S| dB = 20 × log10(|S|).
For reflection parameters (S11, S22):
- 0 dB = total reflection. Open or short circuit.
- −10 dB = 90% transmitted, 10% reflected. Acceptable for many systems.
- −20 dB = 99% transmitted. Good match. Standard RF target.
- −30 dB or better = excellent match.
For transmission parameters (S21):
- 0 dB = lossless. −3 dB = half power (filter bandwidth edge). Positive values = gain (amplifier).
Phase Plots
For S21, phase decreases linearly with frequency for a simple transmission line — the slope is the group delay. Deviations from linearity = dispersion = wideband signal distortion. Phase wrapping (+180° to −180° jumps) is an atan2 artifact, not a physical discontinuity.
Smith Charts
The Smith Chart maps complex impedance onto a unit circle. Center = perfect 50Ω match. Perimeter = total reflection.
- Center: Matched to Z0. Where you want S11 at your operating frequency.
- Right side: Impedance > 50Ω (open-circuit limit at far right).
- Left side: Impedance < 50Ω (short-circuit limit at far left).
- Upper half: Inductive. Lower half: Capacitive.
- Loops: Rapidly changing impedance — resonant structures or poorly matched lines.
Pro tip: In the viewer above, select “Smith” plot type. Tight clusters near center = broadband match. Large loops = narrowband resonant behavior.
The Option Line
Every Touchstone file contains an option line: # GHz S MA R 50 — frequencies in GHz, S-parameters, magnitude-angle format, 50Ω reference.
| Code | Format | Columns Per Parameter |
|---|---|---|
MA | Magnitude and Angle (degrees) | 2 |
DB | dB and Angle (degrees) | 2 |
RI | Real and Imaginary | 2 |
Data Lines
For a .s2p file, each line contains: freq S11_a S11_b S21_a S21_b S12_a S12_b S22_a S22_b. This viewer parses all three formats automatically.
Port Count and File Extensions
| Extension | Ports | S-Parameters | Common Use |
|---|---|---|---|
.s1p | 1 | S11 only | Antennas, single-port impedance |
.s2p | 2 | S11, S21, S12, S22 | Filters, amplifiers, cables, connectors |
.s3p | 3 | 9 parameters | Power dividers, circulators |
.s4p | 4 | 16 parameters | Differential pairs, baluns, couplers |
1. Return Loss Degradation at Connectors
Connector transitions (SMA to microstrip, U.FL to PCB pad) introduce impedance discontinuities. Appears as periodic ripple in S11. The ripple period corresponds to the electrical length of the mismatch: c / (2 × L × √εr).
Common mistake: Measuring S11 at a single frequency and declaring the match “good.” If you hit a cancellation null, the match appears 15–20 dB better than adjacent frequencies. Always look at the full sweep.
2. Insertion Loss Slope — The Cable You Forgot
S21 decreasing linearly with frequency = cable/transmission line loss. Conductor loss scales as √f (skin depth), dielectric loss scales linearly. If you don't de-embed test cables, their loss gets attributed to the DUT. Use the viewer's multi-file overlay to de-embed visually.
3. Resonant Spikes — The Parasitic You Didn't Model
Sharp dips in S21 at specific frequencies = a resonance. Common sources:
- Via stubs: Through-hole vias create stubs that resonate at
f = c / (4 × L_stub × √εr). ~30 GHz on FR-4 with 1.2mm stub — right where 5G mmWave and automotive radar operate. - Ground plane slots: Creates a slot antenna radiating at its resonant frequency.
- Connector cavity modes: Higher-order modes above cutoff frequency absorb energy.
4. Group Delay Variation
Flat magnitude response doesn't guarantee signal integrity. Group delay (τg = −dφ/dω) variation smears wideband signals. Look at S21 phase — curvature = group delay variation. Worst near filter band edges.
5. Poor Isolation — The Amplifier That Oscillates
S12 measures reverse isolation. If too high (insufficiently negative), the amplifier can oscillate. Rollett stability: if K < 1 at ANY frequency (including out-of-band), potentially unstable. Check S12 across the full measured range.
Using the viewer: Create two panels — S21 + S12 on magnitude, S11 on Smith. If S12 approaches S21 at any frequency, investigate stability.
Use this viewer when you need to:
- Quickly inspect a
.snpfile from a vendor or measurement - Compare S-parameters from two devices, or before-and-after measurements
- Overlay spec masks to check pass/fail against requirements
- Export publication-quality PNG plots for reports
- Share RF data with non-RF engineers who don't have licensed tools
- Triage production measurements without reserving an ADS seat
Use licensed tools when you need to:
- De-embed fixtures mathematically (TRL, SOLT calibration post-processing)
- Compute stability circles, noise figures, or power gain contours
- Cascade S-parameter networks (series, parallel, or mixed-mode)
- Convert between parameter types (S → Z → Y → ABCD)
- Perform time-domain reflectometry (TDR) via inverse FFT
What file formats are supported?
Touchstone Version 1: .s1p through .s16p. All three data formats (MA, DB, RI) parsed automatically. Frequency units (Hz, kHz, MHz, GHz) detected from the option line.
Is my data uploaded to a server?
No. All processing happens locally in your browser. Your .snp files never leave your machine. No backend, no API calls, no data collection.
Can I use this for differential S-parameters (.s4p)?
Yes. The viewer parses .s4p files and displays all 16 single-ended S-parameters. Mixed-mode conversion (SDD11, SCC21) is not implemented — use a licensed tool for modal decomposition.
Why do my Smith chart traces look different from my VNA?
This viewer normalizes to the reference impedance in the Touchstone file's option line. If your file says R 75, the Smith chart is 75Ω-referenced — different positions than a 50Ω VNA display.
What does the spec mask feature do?
Define pass/fail boundaries: frequency range, parameter, min/max dB limits. The viewer shades the pass region for instant visual screening. Useful for production triage.
Build systems that cross domain boundaries?
The Pattern Architect Newsletter delivers weekly engineering intelligence across RF, embedded systems, autonomous hardware, and systematic thinking. One framework per week.
Subscribe Free →S-Parameters: The Complete Engineering Reference
Everything you need to interpret Touchstone files, read Smith charts, and diagnose common RF integration failures. Click any section to expand.
S-parameters — scattering parameters — describe how electromagnetic energy propagates through an RF network. Unlike impedance (Z) or admittance (Y) parameters, S-parameters are defined in terms of traveling waves rather than total voltages and currents. This distinction matters because at RF and microwave frequencies, measuring total voltage at a point requires knowing the forward and reflected wave components separately — which is exactly what S-parameters provide.
For a two-port network (the most common case, and what .s2p files describe), there are four S-parameters:
| Parameter | Physical Meaning | What It Tells You |
|---|---|---|
S11 | Input reflection coefficient | How much power is reflected back from Port 1. Measures impedance match quality at the input. |
S21 | Forward transmission coefficient | How much power passes from Port 1 to Port 2. The gain (amplifiers) or insertion loss (passive networks). |
S12 | Reverse transmission coefficient | How much power leaks from Port 2 back to Port 1. Measures isolation in amplifiers, reverse leakage in filters. |
S22 | Output reflection coefficient | How much power is reflected back from Port 2. Impedance match quality at the output. |
The Complex Representation
Each S-parameter is a complex number with magnitude and phase:
Sij = |Sij| · e jφij
The magnitude and phase carry distinct physical meaning:
Magnitude represents the ratio of voltage wave amplitudes. |S11|² gives the reflected power ratio — if |S11| = 0.316 (−10 dB), then |S11|² = 0.1, meaning 10% of incident power is reflected. |S21|² gives the transmitted power ratio — this directly maps to insertion loss for passive networks or gain for amplifiers. For a lossless, reciprocal two-port network, power conservation requires |S11|² + |S21|² = 1. Any deviation indicates loss (absorbed power) or measurement error.
Phase encodes the electrical delay through the network. For a uniform transmission line of length l, the transmission phase is φ21 = −βl, where β = 2πf√εeff/c is the propagation constant. The group delay — the quantity that determines signal distortion — is the derivative of phase with respect to angular frequency:
τd = −dφ/dω
Constant group delay means all frequency components traverse the network in the same time — the signal shape is preserved. Group delay variation across the signal bandwidth causes dispersion: different spectral components arrive at different times, smearing the output waveform. A bandpass filter with 0.1 dB magnitude ripple may still introduce 5 ns of group delay variation at the band edges — for a 100 MHz wide signal, this can be the dominant source of intersymbol interference.
Key insight: S-parameters are measured under matched conditions — each port is terminated with the reference impedance Z0 (usually 50Ω). If your actual system impedance differs from Z0, the S-parameters alone don't predict how the component will behave in-circuit. You need to de-embed (remove fixture effects) or re-normalize (transform to a different reference impedance). The relationship between S-parameters and impedance is: Zin = Z0 · (1 + S11) / (1 − S11).
Magnitude Plots (dB)
The magnitude of an S-parameter is plotted in decibels: |S| dB = 20 × log10(|S|).
For reflection parameters (S11, S22):
- 0 dB = total reflection. Open or short circuit.
- −10 dB = 90% transmitted, 10% reflected. Acceptable for many systems.
- −20 dB = 99% transmitted. Good match. Standard RF target.
- −30 dB or better = excellent match.
For transmission parameters (S21):
- 0 dB = lossless. −3 dB = half power (filter bandwidth edge). Positive values = gain (amplifier).
Phase Plots
For S21, phase decreases linearly with frequency for a simple transmission line — the slope is the group delay. Deviations from linearity = dispersion = wideband signal distortion. Phase wrapping (+180° to −180° jumps) is an atan2 artifact, not a physical discontinuity.
Smith Charts
The Smith Chart maps complex impedance onto a unit circle. Center = perfect 50Ω match. Perimeter = total reflection.
- Center: Matched to Z0. Where you want S11 at your operating frequency.
- Right side: Impedance > 50Ω (open-circuit limit at far right).
- Left side: Impedance < 50Ω (short-circuit limit at far left).
- Upper half: Inductive. Lower half: Capacitive.
- Loops: Rapidly changing impedance — resonant structures or poorly matched lines.
Pro tip: In the viewer above, select “Smith” plot type. Tight clusters near center = broadband match. Large loops = narrowband resonant behavior.
The Option Line
Every Touchstone file contains an option line: # GHz S MA R 50 — frequencies in GHz, S-parameters, magnitude-angle format, 50Ω reference.
| Code | Format | Columns Per Parameter |
|---|---|---|
MA | Magnitude and Angle (degrees) | 2 |
DB | dB and Angle (degrees) | 2 |
RI | Real and Imaginary | 2 |
Data Lines
For a .s2p file, each line contains: freq S11_a S11_b S21_a S21_b S12_a S12_b S22_a S22_b. This viewer parses all three formats automatically.
Port Count and File Extensions
| Extension | Ports | S-Parameters | Common Use |
|---|---|---|---|
.s1p | 1 | S11 only | Antennas, single-port impedance |
.s2p | 2 | S11, S21, S12, S22 | Filters, amplifiers, cables, connectors |
.s3p | 3 | 9 parameters | Power dividers, circulators |
.s4p | 4 | 16 parameters | Differential pairs, baluns, couplers |
1. Return Loss Degradation at Connectors
Connector transitions (SMA to microstrip, U.FL to PCB pad) introduce impedance discontinuities. Appears as periodic ripple in S11. The ripple period corresponds to the electrical length of the mismatch: c / (2 × L × √εr).
Common mistake: Measuring S11 at a single frequency and declaring the match “good.” If you hit a cancellation null, the match appears 15–20 dB better than adjacent frequencies. Always look at the full sweep.
2. Insertion Loss Slope — The Cable You Forgot
S21 decreasing linearly with frequency = cable/transmission line loss. Conductor loss scales as √f (skin depth), dielectric loss scales linearly. If you don't de-embed test cables, their loss gets attributed to the DUT. Use the viewer's multi-file overlay to de-embed visually.
3. Resonant Spikes — The Parasitic You Didn't Model
Sharp dips in S21 at specific frequencies = a resonance. Common sources:
- Via stubs: Through-hole vias create stubs that resonate at
f = c / (4 × L_stub × √εr). ~30 GHz on FR-4 with 1.2mm stub — right where 5G mmWave and automotive radar operate. - Ground plane slots: Creates a slot antenna radiating at its resonant frequency.
- Connector cavity modes: Higher-order modes above cutoff frequency absorb energy.
4. Group Delay Variation
Flat magnitude response doesn't guarantee signal integrity. Group delay (τg = −dφ/dω) variation smears wideband signals. Look at S21 phase — curvature = group delay variation. Worst near filter band edges.
5. Poor Isolation — The Amplifier That Oscillates
S12 measures reverse isolation. If too high (insufficiently negative), the amplifier can oscillate. Rollett stability: if K < 1 at ANY frequency (including out-of-band), potentially unstable. Check S12 across the full measured range.
Using the viewer: Create two panels — S21 + S12 on magnitude, S11 on Smith. If S12 approaches S21 at any frequency, investigate stability.
Use this viewer when you need to:
- Quickly inspect a
.snpfile from a vendor or measurement - Compare S-parameters from two devices, or before-and-after measurements
- Overlay spec masks to check pass/fail against requirements
- Export publication-quality PNG plots for reports
- Share RF data with non-RF engineers who don't have licensed tools
- Triage production measurements without reserving an ADS seat
Use licensed tools when you need to:
- De-embed fixtures mathematically (TRL, SOLT calibration post-processing)
- Compute stability circles, noise figures, or power gain contours
- Cascade S-parameter networks (series, parallel, or mixed-mode)
- Convert between parameter types (S → Z → Y → ABCD)
- Perform time-domain reflectometry (TDR) via inverse FFT
What file formats are supported?
Touchstone Version 1: .s1p through .s16p. All three data formats (MA, DB, RI) parsed automatically. Frequency units (Hz, kHz, MHz, GHz) detected from the option line.
Is my data uploaded to a server?
No. All processing happens locally in your browser. Your .snp files never leave your machine. No backend, no API calls, no data collection.
Can I use this for differential S-parameters (.s4p)?
Yes. The viewer parses .s4p files and displays all 16 single-ended S-parameters. Mixed-mode conversion (SDD11, SCC21) is not implemented — use a licensed tool for modal decomposition.
Why do my Smith chart traces look different from my VNA?
This viewer normalizes to the reference impedance in the Touchstone file's option line. If your file says R 75, the Smith chart is 75Ω-referenced — different positions than a 50Ω VNA display.
What does the spec mask feature do?
Define pass/fail boundaries: frequency range, parameter, min/max dB limits. The viewer shades the pass region for instant visual screening. Useful for production triage.
Build systems that cross domain boundaries?
The Pattern Architect Newsletter delivers weekly engineering intelligence across RF, embedded systems, autonomous hardware, and systematic thinking. One framework per week.
Subscribe Free →