TABLE OF CONTENTS In this chapter we cover the generic measurement considerations around PDNs. Some of the specific measurement setups for PDN structures will be covered in later chapters where we describe the modeling and characterization details.
With increasing power levels and constantly dropping supply voltages, the PDN impedance of high-power systems must be low, sometimes in the milliohm or submilliohm range. Even if we do not necessarily need very low PDN impedance (i.e., in the case of lower-power systems and low-current power rails) dc-dc converters with high dc loop gain will create submilliohm output impedances in the Hertz and low kilohertz frequency range. Moreover, low-loss test fixtures and low-ESR bypass capacitors may have effective impedances in the milliohm range also. These low impedance values themselves create unique challenges in the selection of instruments, setups, and connections. As we will see, a big part of the challenge is practical: the constraints of high-density product designs are in conflict with the size and cost of connections needed for repeatable and consistent measurement of low PDN impedance.
There are several basic methods for the purposes of impedance measurements. Reference [1] lists six commonly used options:
Bridge method;
Resonant method;
I-V method;
RF I-V method;
Network analyzer method;
Autobalancing method.
Figure 5.1 shows the core concept behind each method. According to [1], the bridge method has high accuracy, low cost, but narrow frequency coverage; so, the balancing may need to be done manually. The resonant method inherently assumes...
