Substrate permittivity measurement for PCB Printed Technology Circuits
Substrate permittivity measurement for PCB Printed Technology Circuits – 2016/2017 José Carlos Martínez Durillo
Index
Background €€ Manufacturer, substrate… Depending on Not too accurate sometimes.
Velocity Factor Ratio of the speed at which a certain signal passes through a medium, such as a PCB.
Also, if affects to… üImpedance üCapacitance üStrip width …
Also, if affects to… üImpedance üCapacitance üStrip width …
Given a choice, lower permittivity is nearly always BETTER Usual assumption WRONG
Design phase Signal type Design requirements Low loss Steady behaviour … … Design prototype Simulation & analysis ac Mo cu re ra te Refine requirements Methods
Methods Hard to get • Capacitance method -> Very size-dependant -> acceptable results • Resonant cavity-like treatment -> PCB≈2 D open-sided cavity • Using an oscilloscope -> Voltage across a resistor between supply and capacitor is measured -> Complex procedure • Time Domain Reflection -> By using reflectometry, velocity factor is estimated and from it, permittivity.
Methods • Resonant cavity-like treatment Based on treating the substrate like a 2 D open-sided cavity Each resonance corresponds to an integral “wave number” or “mode number” in each direction (length and width)
Methods • Depending on the dimensions of the substrate, resonances will vary Countless combinations Does not affect the results!
Methods Improved accuracy Neglectable if Q>100 Correction factor to take losses into account Howell & Wang
Methods Improved accuracy Neglectable if Q>100 Correction factor to take losses into account Howell & Wang
Setting up the substrate • We will measure FR 4 substrate of 1. 5748 mm width. • Substrate is cut of 284 x 35. 60 mm using LPKF S 62 PCB Prototyping Milling Machine
Connection process • Again, different possibilities: Needs two SMA
Connection process • Again, different possibilities:
Connection process
Connection process
Connection process
Connection process For simplicity, we try the simplest and cheapest one: reflection SMA Connection in a corner
Theoretical results
Experimental Results. Comparison. We must calibrate the network analyzer before: Open, short and load
Experimental Results. Comparison. Measured Adjusted
Experimental Results. Comparison. Mean: 3. 963
Experimental Results. Comparison.
Experimental Results. Comparison.
Experimental Results. Comparison. 0. 227 % deviation
Conclusions ØKeeping in mind the importance of the permittivity is a vital matter when designing PCB. ØDepending on the quality of the substrate, the tolerance given by the manufacturer may not be accurate enough. ØMeasurement techiques must be carried then, to ensure a correct behaviour of the substrate. This paper has gone through some of them.
References http: //www. ce-granollers. com/placa_virgen 1. htm FR 4 Datasheet: http: //www. bungard. de/images/downloads/ipc 4101 d. pdf http: //www. wetterlin. org/sam/SA/Operation/PCB_Diel_Const. pdf Electromagnetic Waveguides, S. F. Mahmoud, 1991. http: //www. sglabs. it/public/Sg. Labs_HP_85036 B_5. JPG Wang, “Determining Dielectric Constant and Loss-Tangent in FR-4”, UMR, 2000.
Any questions?
- Slides: 32