How We Measure Common-Mode Chokes With Current Probes
A current path first. A number second.
How We Measure Common-Mode Chokes With Current Probes
A practical guide to injection and receiving probes, the return loop, calibration, and extracting choking impedance. Start with the bench arrangement; continue into the mathematics when you want to know exactly what the analyzer is measuring.
RF.Guru working definition: Common-mode current is the non-cancelling phasor-sum current in a specified set of conductors, evaluated at a defined cross-section and using a declared current-direction convention. In the intended differential transmission-line mode, the outgoing and return currents are equal and opposite, so their phasor sum is zero. When they do not cancel, the remaining current must close through another reference or return path—such as the outside of a coax shield, a mast, equipment chassis, station wiring, nearby structures, earth, the operator, or distributed coupling through the environment.
This broader working definition is especially useful in practical antenna systems. On transmit, non-cancelling current on the outside of the coax can make the feedline and connected structures part of the radiating antenna system unless that path is intentional, clearly defined and properly controlled—for example by providing the required return path and placing a suitable common-mode choke at the correct boundary.
We start with the standards—and the current path
We start from established EMC principles, including ITU guidance on common-mode mitigation, then use RF current probes to investigate the actual circuit. EMC means electromagnetic compatibility: equipment working together without unacceptable interference. ITU-T K.37, particularly its discussion of high-frequency common-mode chokes, makes the essential point: a choke adds impedance to a common-mode circuit, and its effectiveness depends on the impedance already in that circuit. The number on the choke cannot describe the whole installation.
At RF.Guru we use TekBox RF probes. The useful question is not whether a probe looks professional. It is whether we know what current passes through its aperture, how the probe responds, and where that current returns.
These references have distinct scopes. K.37 explicitly excludes coaxial cables; we cite its general mitigation principle, not conformity of a coax test to that Recommendation. CISPR 16-1-2 addresses conducted-disturbance measuring apparatus, including current probes. IEC 61000-4-6 concerns equipment immunity to conducted RF disturbances. The calibrated two-probe impedance method also has its own research literature. None of these names, by itself, turns an arbitrary analyzer-and-clamps arrangement into a certified choke test. The method, calibration, geometry, frequency range, and uncertainty still have to be stated.
The diagrams and bench sequence below give a reproducible reference workflow for understanding our measurement approach. They do not specify a particular TekBox model or fixed bench dimensions, and the numerical examples are calculations rather than product test results.
Why we use both current probes and Y21
Our current-probe setup takes more work than our usual Y21 bench measurement. Y21 uses a network analyzer’s complete two-port data to recover the choke’s impedance in a controlled fixture. The probe approach adds two magnetic couplings, the probes’ frequency responses, and a deliberately closed external-current loop. Probe spacing, cable routing, the return conductor, and unwanted coupling between the probes all affect the result. We therefore calibrate and verify the probes and the complete loop as well as the analyzer. This is the EMC measurement approach discussed alongside the ITU guidance above; K.37 does not prescribe this two-probe coax test.
We use both because their different physical arrangements help expose different errors. Y21 is convenient for component characterization and, with the appropriate circuit model, separates local capacitance at the two test terminals from the wanted transfer branch. The probe setup deliberately drives and senses the external current path. With its own calibration and model, it provides another route to the choke’s impedance; it can also measure current suppression in a defined loop. More equipment and a more elaborate setup do not automatically mean greater accuracy. The methods complement each other, as explained in Y21 Choke Measurement: Strengths, Limits and Alternatives.
For an impedance cross-check, both methods must describe the same choke, current mode, and terminal conditions, with fixture effects accounted for. Over the frequency range where both methods are valid, the recovered resistance and reactance should agree within the combined measurement uncertainty—the range of error remaining in the two results. A raw probe S21 trace or an installed-current reduction is a different quantity and should not be expected to match a Y21 impedance curve directly.
If comparable results disagree beyond that uncertainty, we have something to investigate. Check the calibration, known reference impedances, unwanted current paths, probe coupling, instrument sensitivity, and the assumptions used to turn the readings into ohms. The cause may be the fixture, the calculation, or a real change in the device or its conditions; disagreement alone does not identify a defective choke. We do not choose whichever curve looks better. Agreement between different methods strengthens the evidence, especially when independent known standards also pass, but shared errors can still make two results agree.
The answer before the equations
A common-mode choke should make unwanted current on the outside of a feed line harder to drive, while allowing the wanted signal inside the coax to pass. Its choking impedance describes that opposition to the unwanted current. To test that function, deliberately excite the outside-current circuit. A normal signal sent down the center conductor into a 50-ohm coax load is testing a different path.
One probe injects a small RF disturbance into a defined loop. A second probe senses the resulting current through its aperture—the opening around the cable. A vector network analyzer, or VNA, compares the amplitude and phase of the received signal with the signal it sends. Comparing responses with known reference impedances and with the choke can give the choke’s resistance and reactance, provided the fixture behaves as the model assumes.
The short version: close the common-mode loop deliberately, keep its geometry fixed, calibrate the instrument and the fixture, verify the result with an independent known load, and stop reporting choke impedance when leakage, resonance, or uncertainty dominates the measurement.
You do not need the algebra to understand the main lesson: a low receiving-probe signal can mean a good choke—or a bad measurement path. Calibration and checks distinguish them.
Two coax terminators and two probes: what is missing?
A reader described approximately one meter of coax on either side of the device, a terminator at each end, and two probes connected to the transmitting and receiving ports of a network analyzer. That is a useful starting point, but it leaves an important connection unspecified.
A conventional 50-ohm terminator connects the coax center conductor to its shield. It loads the differential circuit inside the cable. It does not establish a 50-ohm load between the shield exterior and the bench, and it does not provide a defined return for an externally injected common-mode current.
Without an intentional return, the loop may close through capacitance to the bench, probe housings, VNA cables, the computer’s USB connection, or nearby people. A trace may still appear. Moving your hand might even improve it. That does not make the trace a property of the choke.
For the probe method described here, VNA port 1 connects to the injection probe’s coax connector, and port 2 connects to the receiving probe’s connector. Neither port plugs directly into the center-to-shield signal path of the coax under test. A directly connected coax measurement can be useful too, but it is a different fixture and requires a different circuit interpretation.
A bench loop that you can actually draw
Start with an unpowered, disconnected coaxial 1:1 current choke. Arrange a common-mode loop comprising the shield exterior through the choke and a separate return conductor. Bond the outer shields at the two remote ends to that return. Keep the return outside both probe apertures. Place the choke on insulating supports, with no conductive mounting path that bridges its input and output.
A simple starting configuration for this shield-only impedance fixture is to remove the center-to-shield terminators and leave the center conductors unconnected at the remote ends. Use that condition for the short, reference load, and choke runs, and record it. The intentional loop then uses the shield exterior and the separate return.
Terminated differential ports can be investigated in a more complete modal fixture, but the standards and DUT must see equivalent conditions. In particular, an uninterrupted center conductor plus two end loads can create a bypass around a resistor inserted only in the shield path. Do not add those loads to the reference fixture without accounting for that circuit. A floating dummy load is also not interchangeable with one whose shell is bolted to a grounded panel.
The two probes do not measure “current before” and “current after” the choke. One excites the circuit and the other receives. In an electrically small, single series loop, essentially the same loop current passes both positions. Suppression is established by comparing the receiving current in a reference configuration with the receiving current after inserting the choke, at the same drive conditions. At higher frequencies, shunt capacitance and distributed paths can invalidate that simple series-loop picture.
If the separate return also passes through a probe, its magnetic effect may cancel the current you intended to sense. On a complete coax cable, cancellation of the normal inside-the-coax currents is desirable; cancellation of your deliberately created outside-current loop is not.
Should everything sit on a metal plate?
A metal reference plane is useful when it forms a defined part of the fixture. It gives a repeatable return geometry and helps control the electromagnetic environment. It is not a universal ingredient that makes an undefined test correct.
If the plane is the intended return, specify the two shield bonds, their width and length, the height of the cable and choke, and the probe positions. Hold the assembly above it with insulating supports. Do not let connectors, a metal enclosure, or mounting hardware create a second path around the choke. Keep the plane and all connections unchanged for the standards and the device measurement.
Bringing the plane closer also increases capacitance. Some of that capacitance can bypass a high-impedance choke. A compact return conductor with controlled spacing can be an alternative for component characterization. Use the arrangement you can model, calibrate, and reproduce; report which one you used. An immunity standard’s specified ground plane is a requirement for that particular test, not proof that your choke fixture has the same reference conditions.
One meter on each side is not automatically harmless
Those coax arms are part of the outside-current transmission structure. At 30 MHz the free-space wavelength is about 10 m, so a 2 m path already represents about 72 degrees of free-space phase length. The return path adds geometry, and nearby dielectric and metal change propagation. The velocity factor printed for the coax’s internal mode is not automatically the velocity factor of current on its exterior.
Keep a component fixture compact. More importantly, check it across the complete frequency range. A resonance that moves when you change arm length or height may belong to the fixture. Do not simply subtract a guessed number of cable ohms from a dB trace.
If you keep long arms to reproduce a particular installation, that is a legitimate system experiment. Report the assembled system response. To claim the choke’s terminal impedance separately, you need appropriate calibration at those terminals and a model that handles the distributed fixture.
Choose and place the probes deliberately
TekBox offers both current-monitoring probes and bulk-current-injection probes. They are not interchangeable merely because both have a hole in the middle. Choose the models for the frequency range, aperture, sensitivity, permissible drive, and expected conductor current. Check the individual probe’s calibration data and the manufacturer’s fixture instructions.
The receiving probe normally works into a specified 50-ohm input. A 1-megohm oscilloscope input gives a different response unless the manufacturer explicitly supports that arrangement; see the TekBox current-probe FAQ. An injection probe needs enough coupling at the lowest frequency without exceeding its drive or temperature limits. Small-signal reciprocity can sometimes permit a monitoring probe to inject, but it does not confer a bulk-injection power rating.
Mark probe orientation, spacing, cable position, and clamp closure. A small gap in a split core changes its response. Keep the conductor at the calibrated aperture position; keep probe cables away from the test loop and from each other. Mechanical strain or moving the VNA leads can change phase and leakage.
Calibration is three jobs, not one button
| Job | What it establishes | What it does not establish |
|---|---|---|
| VNA calibration | The instrument’s complex response at the ends of its test cables. | The probes’ magnetic coupling or the external-current fixture. |
| Probe characterization | Receiving transfer impedance, orientation/phase where characterized, and usable probe range. | The complete loop impedance or a choke’s attenuation. |
| Complete-fixture calibration | How known impedances at the device position map into the measured probe response. | Validity after the fixture geometry changes, or at power levels not tested. |
First let the VNA stabilize and perform the appropriate two-port calibration at the ends of the cables that will attach to the probes. Use the actual adapters and reference impedance. Save the calibration and avoid flexing the leads afterward. A trace normalization alone does not provide the same error correction as a full vector calibration.
Next check each probe using its suitable calibration fixture and the manufacturer’s procedure. TekBox supplies monitoring-probe fixture instructions and separate BCI probe fixture instructions. A damaged core or poorly closed clamp can be exposed by this check. A factory magnitude curve is useful, but it is not a complete complex calibration for every fixture you build.
Finally calibrate the complete two-probe loop with standards at the intended choke reference planes. A 50-ohm resistor in this step goes in the common-mode test gap. It is not either of the coax’s center-to-shield terminators. If a resistor connects the shield segments, prevent the center conductor or another conductor from providing a parallel path through the surrounding terminations. Document any temporary center-conductor bonds or disconnections and preserve the defined modal boundary.
A repeatable small-signal bench sequence
- Define the result. Are you measuring the choke’s complex common-mode impedance, suppression in a specified loop, or differential-mode insertion loss? Record the frequency range and reference planes before connecting anything.
- Make the passive fixture. Disconnect transmitters and external power. Fix the return, support height, arm lengths, probe spacing, differential terminations, and instrument-cable routes. Photograph the arrangement.
- Calibrate the VNA and check the probes. Use suitable fixtures, the specified receiver termination, and the same operating band. Keep the receiver safely within its input rating.
- Measure the standards. At the device position measure a characterized short and a characterized load, retaining complex data. An open and additional loads are useful for detecting leakage and checking whether the simple model is adequate.
- Measure the choke without rearranging the bench. Preserve the reference planes and surrounding geometry. Replacing a large choke with a tiny resistor while moving both probes changes more than the unknown impedance.
- Verify with something you did not use to calibrate. A second known resistance and, where practical, a known reactive standard test different parts of the extraction. Use RF-characterized standards, not just their DC markings.
- Check repeatability and linearity. Repeat clamp opening/closing, reconnect the device, and repeat at several small drive levels within all ratings. Reduce VNA IF bandwidth or average if needed, then check that the result has converged rather than simply become smoother.
- Export the evidence. Save complex S-parameters, calibration-standard values, the calculation, fixture photographs, verification residuals, and the frequency interval that passed. Keep the untouched measurement files.
This procedure is for a passive, small-signal bench. Connecting a transmitting amplifier to this arrangement is a different experiment, with separate coupling, protection, thermal, and instrument-safety requirements.
What the receiving probe actually measures
The probe produces a voltage related to the current through its aperture. Its transfer impedance, usually written ZT, expresses that relationship:
Z_T(f) = V_receiver(f) / I_loop(f) I_loop(f) = V_receiver(f) / Z_T(f)
ZT has units of ohms, but it is not the impedance of the choke. It is the probe’s current-to-voltage conversion. To recover phase as well as current magnitude, use complex probe calibration and a consistent polarity reference.
With compatible RMS voltage/current references, the logarithmic version is:
I [dBµA] = V_receiver [dBµV] − Z_T [dBΩ]
In the particular matched 50-ohm calibration-line arrangement where the probe senses the known line current, the derivation gives ZT = 50 Ω × S21, after the applicable cable/reference-plane corrections. In magnitude units:
Z_T [dBΩ] = S21 [dB] + 33.98
The assumptions matter: the calibration-line current must be the known current implied by the incident wave, and the receiver must have the specified termination. With mismatch or a different fixture, use the actual current rather than that shortcut. A weaker probe response means a smaller transfer impedance: with S21 = −20 dB, this example gives 5 Ω, not 500 Ω. The “add 34” rule calibrates a probe in that fixture. It does not turn a two-probe choke trace into choke impedance.
Watch the probe-calibration step
In this supporting demonstration, Mach One Design EMC shows how to characterize an RF current-monitor probe with a matching TekBox calibration fixture and a VNA. It illustrates the probe-transfer-impedance step discussed here, including the matched 50-ohm fixture’s S21 + 34 dB conversion. Calibrating that probe is separate from calibrating the complete two-probe loop to recover a choke’s complex impedance. The video’s through normalization serves its probe-response demonstration; it does not replace the complete vector calibration described above.
Watch “How to Calibrate RF Current Probes | Step-by-Step Tutorial” on YouTube, by Mach One Design EMC. The transfer-impedance derivation starts around 2:04; the comparison with manufacturer data starts around 12:08.
From the probe response to choking impedance
Now we can make the deeper step. A vector network analyzer records amplitude and phase. Call the chosen complex transmission response H(f). In the simple calibrated loop model below, H is the VNA’s S21 at the stated probe-port reference planes. Export its complex values; a screenshot of magnitude in dB is insufficient.
When the fixture can be represented by a single series loop, with stable probe coupling and negligible direct leakage, its response can be written:
H = K / (Z_b + Z_x)
Zx is the unknown common-mode impedance at the device position. Zb is the effective background impedance, including the fixture and probe loading. K is a complex coupling factor. Both K and Zb depend on frequency. Neither is automatically 50 or 100 ohms because the VNA has 50-ohm connectors.
Take two characterized standards Za and Zr, with responses Ha and Hr. At each frequency:
K = (Z_r − Z_a) / (1/H_r − 1/H_a) Z_b = K/H_a − Z_a Z_x = K/H_x − Z_b
For an ideal zero-ohm short, Za = 0 and its response is H0. The same calculation becomes:
Z_x = Z_r × (1/H_x − 1/H_0) / (1/H_r − 1/H_0)
Every quantity in that expression is complex. Convert a VNA magnitude M in dB and phase φ to H = 10M/20 × (cos φ + j sin φ), using consistent angular units. Do not subtract dB numbers and call the result resistance. Use the real short’s impedance in the general formula when its inductance and resistance matter.
This is a stated circuit model, not a universal formula for every two-clamp arrangement. Published two-probe methods also use probe network models, S11 information, and matrix de-embedding. Follow the calibration variables and equations of the method you implement; do not silently substitute raw S21 into a formula defined for a different normalized response. The original Liu and colleagues dual-current-probe work is a useful starting reference.
A calculated example you can reproduce
These are synthetic values for checking the calculation, not measurements of an RF.Guru product. At one frequency, let K = 1 Ω and Zb = 20 + j10 Ω. Use an ideal 100-ohm reference, then an unknown choke of 1000 + j500 Ω.
| At the device position | Complex response H | Magnitude, approximately |
|---|---|---|
| Short, 0 Ω | 0.040000 − j0.020000 | −26.99 dB |
| Reference, 100 Ω | 0.008275862 − j0.000689655 | −41.61 dB |
| Unknown choke | 0.000784314 − j0.000392157 | −61.14 dB |
Substitution into the complex short/load equation recovers R = 1000 Ω and X = +500 Ω. Its impedance magnitude is about 1118 Ω and its phase is +26.6 degrees. The short-to-choke change is approximately 34.15 dB in this particular loop. That is not a promise of 34.15 dB suppression on an arbitrary antenna.
When a two-standard calibration is not enough
The probes can talk directly to each other through magnetic or electric fields. Signal can also couple between their coax leads. The receiver then sees a mixture of wanted loop response and leakage. At a high choke impedance the wanted signal becomes small while that direct path may remain.
A simple extension is H = Hleak + K/(Zb + Zx). If an ideal open really produces Hleak, without changing the surrounding coupling, define G = H − Hopen and apply the preceding equations to G instead of H. The subtraction must be vector subtraction, including phase.
An actual open has capacitance, and opening the fixture can change its fields. Therefore an open trace is immediately useful as a leakage warning, but not automatically a perfect correction. Where necessary, use three characterized standards and fit the more general bilinear relationship H = (aZ + b)/(cZ + d), or use a full two-port/network de-embedding method. Verify against further standards. The research on improved two-probe calibration addresses why probe coupling and calibration-standard selection deserve this attention.
A calibration made with only a short and a small resistor can be poorly conditioned when extrapolated to several kilohms. Include verification impedances near the intended range, and check their parasitic capacitance. Calibration is not permission to recover an arbitrarily large impedance from an arbitrarily small received signal.
The 100-ohm shortcut belongs to a different circuit
For a pure series impedance Z inserted directly between two matched ports of real impedance Z0, the familiar expression is:
S21 = 2Z_0 / (2Z_0 + Z) Z = 2Z_0 × (1/S21 − 1)
With 50-ohm ports, 2Z0 is 100 ohms. That explains the familiar constant. It does not establish that an inductively coupled probe loop has a 100-ohm source-plus-load impedance.
Nor is S21 inherently a differential-mode measurement, and Y21 inherently a common-mode one. The physical connections define the mode; S and Y describe the resulting network. A correctly constructed direct fixture can characterize common-mode impedance. A wrongly constructed probe fixture can measure mostly leakage. Our related Y21 and S21 articles explain those complementary methods and their boundaries.
Choking impedance and current suppression are different results
Report the common-mode impedance as R + jX, together with magnitude and phase. Resistance describes the dissipative part of the impedance; reactance describes its energy-storage behavior. A negative reactance above a resonance is possible. A large magnitude alone does not describe heating or the risk of resonance with the rest of the system.
For a fixed Thevenin driving voltage in a simple series common-mode circuit, with the original circuit impedance Zenv, the current suppression caused by adding Zchoke is:
A_I = 20 log10 |I_reference / I_with_choke|
= 20 log10 |(Z_env + Z_choke) / Z_env|
For a purely resistive 100-ohm environment, a 1000-ohm resistive choke gives about 20.83 dB reduction. A purely reactive j1000-ohm choke has the same impedance magnitude but gives about 20.04 dB in that same environment. With a reactive environment, partial cancellation can even increase current at a particular frequency. These are calculated circuit examples, not antenna predictions.
A sweep of 20 log10|Zchoke/50 Ω| is an impedance referenced to 50 ohms. It is not automatically insertion loss, current reduction, or a system common-mode rejection ratio. State exactly which quantity your dB plot represents.
The traps that produce convincing-looking nonsense
| Observation | Possible cause | Useful check |
|---|---|---|
| A deep, exceptionally narrow null | Vector cancellation between leakage and loop response. | Inspect phase; compare the open/leakage trace and independent standards. |
| The result changes when a hand approaches | An uncontrolled capacitive return or fixture coupling. | Define the return and supports; repeat without touching the assembly. |
| A peak follows coax-arm length | A distributed fixture resonance. | Repeat with a separately calibrated geometry; identify what follows the fixture. |
| Impedance changes with probe spacing | Direct probe coupling, altered loading, or changing cable geometry. | Recalibrate each geometry and compare a known standard. |
| Negative resistance on a passive choke | Wrong phase/polarity, reference-plane error, leakage, or an inadequate model. | Verify calibration and passivity before interpreting the DUT. |
| Very high impedance with noisy phase | Receiver noise floor or ill-conditioned subtraction. | Improve sensitivity safely; report a bound or invalid interval when extraction is unstable. |
| The result changes with drive level | Probe/core nonlinearity, receiver compression, heating, or insufficient signal-to-noise ratio. | Repeat the standards at the same levels before blaming the choke. |
| Unexpected loss of choking after mounting | Metalwork, a ground terminal, or cable shield bypasses the intended path. | Draw every conductive connection and repeat the continuity check. |
Even a small capacitance can matter. At 30 MHz, 10 pF has a reactance magnitude of about 531 Ω. A real 10 pF path bridging a multi-kilohm choke would therefore be substantial. That is an illustrative calculation—not a claim that every fixture has 10 pF.
Background subtraction is not a cure for an unstable background. If the leakage is almost as large as the received signal, small changes in phase or geometry produce large changes in the inferred impedance. Keep meaningful margin above the noise and leakage limits and quantify repeatability; no single “looks clean enough” rule applies to every analyzer and probe pair.
What this test does not tell you about a balun
A balun may also transform impedance, support a balanced load, or fail under an asymmetric load. A common-mode impedance test addresses only part of that job. Measure differential insertion loss and matching separately, and investigate current balance or mode conversion with an appropriate balanced or mixed-mode fixture.
A coaxial 1:1 current choke is the straightforward case for this guide. For a two-wire device, both conductors must be excited in the intended common-mode sense, with a return outside the probe aperture. Tying terminals together can be appropriate in a defined common-mode fixture, but it is not a universal connection recipe for a 4:1 transformer or voltage balun. Draw its winding and terminal circuit first.
Likewise, small-signal impedance is not a kilowatt power rating. Differential RF power can heat coax, conductors, and connectors even when their magnetic effects largely cancel in the ferrite. Common-mode current causes additional loss, and ferrite properties can change with temperature and excitation. Voltage stress, mismatch, duty cycle, enclosure temperature, and insulation require separate assessment. Never infer a continuous-power rating from a low-level VNA sweep.
What should accompany a published result?
A useful report identifies the DUT, probe models and individual calibration data, VNA reference planes, receiver termination, drive level, sweep settings, and temperature. It shows the loop and return dimensions, probe placement, plane bonds, center-conductor terminations, mounting state, and any external grounding or computer connections.
Include the standard models, calibration date, unused verification loads, complex raw data, extraction equation, repeatability results, and usable frequency range. Plot R, X, and |Z| rather than only the most attractive dB curve. State whether the result is terminal impedance, complete-loop impedance, or suppression in a particular test circuit.
If the intended claim is an on-air current reduction, test that installation separately and measure at more than one feed-line position. A low reading at one location can be a current minimum. A bench measurement becomes useful to an operator when its reference conditions are clear enough to explain what changes after installation.
The useful conclusion
The answer to “Do I have everything connected?” begins with the return conductor. The answer to “How do I calibrate?” includes the analyzer, the probes, and the complete fixture. The answer to “How many ohms is my choke?” requires complex data and a model that survives verification.
Current probes are valuable because they let us excite and observe the current system of interest. Their presence alone does not make a measurement correct. Draw the loop, establish the reference, check a known impedance, and keep the limitations attached to the result. Then the trace becomes engineering evidence.
Sources and further technical reading
- ITU-T K.37—EMC mitigation principles; common-mode choke effectiveness depends on the surrounding circuit.
- CISPR 16-1-2 and IEC 61000-4-6—respectively measuring-apparatus requirements and conducted-RF immunity testing; different scopes from standalone choke characterization.
- TekBox RF probes and current-probe calibration fixtures—probe families, suitable fixtures, and transfer-impedance characterization.
- Liu et al., dual-current-probe common-mode loop impedance method—calibration and removal of probe frequency response.
- Zhao et al., improved calibration for a two-probe setup—calibration accuracy and the effects of the measurement arrangement.
Mini-FAQ
- Do the two 50-ohm coax terminators close the common-mode loop? No. They terminate center-to-shield differential mode. The external-current circuit needs its own defined return and reference conditions.
- Do the probes measure current before and after the choke? In this method one probe injects and the other receives. In a simple series loop the current is essentially the same around the loop; suppression is a reference-versus-DUT comparison.
- Can I calculate choking impedance from S21 in dB alone? Generally not. Complex extraction needs amplitude and phase, characterized standards, and a valid circuit model.
- Does putting everything on a metal plate fix the setup? Only when the plate and its bonds are a defined part of the fixture. It can also introduce bypass capacitance or conductive paths that change the result.
- Is a low-power probe sweep enough to establish the power rating? No. Thermal behavior, common-mode excitation, insulation, connectors, mismatch, and duty cycle require separate checks.
- Is this an ITU-certified two-probe measurement? No certification is claimed. ITU EMC guidance, probe apparatus standards, immunity methods, and calibrated component impedance measurements have distinct scopes.
Found this useful? Support my writing and experiments: Buy me a coffee.