Time Domain Reflectometry For Impedance Measurement

A controlled impedance requirement is a promise that the board will present a defined impedance to a signal, and the only way to keep that promise is to measure it. The measurement is made on a test coupon, which is a structure built into the panel that reproduces the trace geometry of the design, and the instrument is a time domain reflectometer.

This article explains how the measurement works, what limits its accuracy, and how the coupon and the acceptance window should be defined.

Why A Coupon Is Measured

The finished trace is buried under mask and surrounded by components, so it cannot be probed directly. The coupon is placed in the panel border or in a dedicated area, and it carries a trace of the same width and layer arrangement as the design. Measuring the coupon characterises the process that produced the board, so it is a process control measurement rather than a check on an individual net.

The coupon also carries the structures needed to characterise the material, because the impedance of a trace depends on the dielectric constant and on the thickness of the dielectric beneath it. A coupon that measures within the window demonstrates that the lamination and the etch produced the geometry that the design assumed.

What Time Domain Reflectometry Does

The instrument launches a fast step or pulse into the trace and records the voltage that returns as a function of time. A change in impedance along the trace produces a reflection, and the amplitude and sign of the reflection give the magnitude and the direction of the change. The time axis can be converted to a distance using the propagation velocity of the trace.

For a uniform trace the waveform shows a flat region whose level corresponds to the impedance of the line, referenced to the impedance of the instrument and the cable. The value is read in the middle of the trace rather than at the ends, because the ends are affected by the launch and by the termination.

Rise Time And Resolution

The rise time of the instrument sets the spatial resolution. A fast edge resolves a short feature, while a slow edge averages over a length of the trace and cannot see a short discontinuity. The length that the instrument resolves is approximately the product of the rise time and the propagation velocity divided by two, so an edge of a hundred picoseconds resolves features of the order of a centimetre on a typical laminate.

The same relationship sets a limit on how accurately a short coupon can be measured, because a trace that is comparable in length to the resolution is dominated by the transitions at its ends. The coupon should be long enough that the instrument can resolve a flat region in the middle, and the acceptance value should be read from that region rather than from the whole waveform.

The Launch And Its Effect On The Result

The connection between the probe and the coupon is a discontinuity, and it appears in the waveform as a disturbance at the beginning of the trace. A well designed launch minimises it, and the coupon is arranged so that the measurement point is far enough from the launch for the disturbance to have died away. A probe that presses on the pad with an uncontrolled force gives a different result on each measurement.

The pad at the end of the coupon is itself a capacitance, and a large pad produces a dip in the waveform that can be mistaken for a low impedance trace. The coupon structures are designed with a defined pad and a defined launch, and the measurement is repeatable only when the same probe and the same technique are used. The test procedure should therefore specify the probe and the launch as part of the method.

Reading The Trace And Setting Limits

The impedance is read as an average over the middle portion of the trace, and the average is compared with the target. The acceptance window is set from the requirement of the design rather than from the capability of the process alone, and it should account for the fact that the measurement itself has an uncertainty.

A window of plus or minus ten percent is common for a general purpose design, and tighter windows are used for a radio frequency board. The tighter the window, the more the process has to be controlled, because the impedance depends on the trace width, the dielectric thickness and the dielectric constant, and each of those has its own tolerance. The way those tolerances combine is the same reasoning that applies to microstrip and stripline geometry, and to the length matching described for matched traces.

Coupon Design For A Meaningful Result

The coupon should reproduce the design geometry exactly, including the trace width, the layer, the dielectric and the copper thickness. It should be long enough to give a measurable flat region and short enough to fit the available panel area, and it should be placed where the process conditions match the product area.

A coupon in the panel border may experience a different plating thickness from a coupon in the middle of the panel, because the current distribution varies. Where the impedance requirement is tight, coupons in more than one location give a better picture of the panel than a single coupon, and the results are recorded with the panel identification so that a drift can be traced. The fabrication route that produces those coupons is described under PCB design and fabrication.

Additional Considerations for This Build

Practical attention to impedance measurement pays for itself here, because it decides whether the finished board behaves as the drawing intended. Where the requirement is not stated on the fabrication drawing or in the assembly notes, the shop has to assume a default, and that default is rarely the value the design was simulated with. Stating impedance measurement explicitly, together with the tolerance that applies, removes the assumption and keeps the result predictable from batch to batch.

Process Control and Verification

On a design of this kind, test coupon is the item that decides how the rest of the board is arranged. Where a value sits close to a process limit, the drawing should say so, since the shop can then open the process window rather than working to a nominal figure that carries no tolerance. A stack-up that is drawn rather than described removes most of the ambiguity from a quotation, and it lets the fabricator price the board against the dielectric and copper weights that will actually be used.

Where the requirement is not written down, the shop supplies its own default, and the default is chosen for the process rather than for the design.

FAQ

Why is the impedance read in the middle of the trace? Because the ends are affected by the launch and by the termination. The middle region is the part of the waveform that represents the uniform line.

Does the coupon measure the actual signal traces? No. It measures the process that produced them, on a structure built from the same geometry. It is a process control measurement rather than a check on an individual net.

How does rise time limit the measurement? A slower edge averages over a longer length, so it cannot resolve a short discontinuity and it smooths the waveform. The coupon has to be long enough for the instrument to resolve a flat region.