Issue link: https://iconnect007.uberflip.com/i/1040234
62 DESIGN007 MAGAZINE I OCTOBER 2018 opposed to a foil lamination at the PCB fabri- cator. Controlling the signal conductor width is more critical for circuits that use thin sub- strates because thinner substrates usually have narrower conductors. A normal conduc- tor width tolerance is ± .5 mil, but this can depend on the copper thickness and circuit density. For a circuit using a thick substrate and a wide signal conductor, the common con- ductor width tolerance has less influence on impedance variation than a thin circuit with a narrow conductor. The impedance of a circuit using a narrower conductor is impacted more by the conductor width tolerance than a circuit using a wider conductor. Copper plating is applied in the process to generate plated through-hole vias and has some thickness variation. Even though the copper thickness variation can impact the impedance performance of the circuit to a lesser degree than substrate thickness and conductor width, a larger concern is usually how thicker copper can increase the conduc - tor width tolerance. Thicker copper cannot be etched as accurately for good conductor width control as thinner copper; however, there are exceptions, depending on the image and etch- ing processes. The most influential variables should be considered for impedance control. If the PCB design has a tight specification for imped- ance, a substrate with strong thickness control should be considered along with a well-con- trolled process for generating the signal con- ductor width. Copper thickness control and Dk control are important as well, but these are typically less critical for impedance accuracy. However, Dk consistency is extremely influ- ential for the phase response of the radio-fre- quency (RF) circuit. Additionally, the measurement of imped- ance has many issues to consider. Typically, a time-domain reflectometer (TDR) is used to measure the impedance of a PCB. There are many types of TDRs with different capabili - ties. In general, the rise time of the TDR is essential and a faster rise time will allow more accurate impedance measurements. Further, there can be an issue with masking, which is a term used to explain that a large impedance anomaly can mask the true impedance anom - aly behind it. The basic operation of a TDR is to send a pulse down a circuit and look at the reflected energy to determine the impedance. It is not uncommon to have a large impedance anom- aly where the test probe meets the circuit being tested for impedance. This impedance anomaly will reflect a lot of energy back to the TDR, and less energy will propagate down the circuit. This makes other impedance measure- ments after the impedance anomaly somewhat altered. It is relatively easy to evaluate this issue by testing a circuit that has a good impedance anomaly where the probe meets the circuit, then use a tool to alter the circuit pattern in the area where the probe touches the circuit and causes a worse impedance anomaly. After this, retest the circuit and a change in imped- ance will often be detected. Depending on the TDR's capabilities and the circuit construction, the impedance of the trace will change from the original measurement. In some cases, it may be very little, and in others, more signifi- cant. DESIGN007 John Coonrod is technical marketing manager at Rogers Corporation. To read past columns or contact Coon- rod, click here. The impedance of a circuit using a narrower conductor is impacted more by the conductor width tolerance than a circuit using a wider conductor.