Surface Roughness Compensation, Gradient, and Other Methods Application Note AP8206

btty038

Surface Roughness Compensation, Gradient, and Other Methods Application Note AP8206 [Copy link]

Choosing the Right Surface Roughness Measurement
There are many surface roughness compensation models that can help designers and manufacturers simulate the effects of copper roughness on insertion loss. The Polar Si9000e offers several methods, depending on your or your OEM requirements. Huray (Canonball) accepts Rz or SEM data if available, and Gradient accepts Rq (RMS) roughness – as do the traditional methods Groisse and Hammerstad, which are only valid below a few GHz
Just as important as choosing the right model is choosing the right surface roughness measurement for the model input.
Common measurements are Ra, Rq, and Rz:
Ra is the absolute average of the profile values
Rq is the root mean square (RMS) of the profile values.
Rz is the peak to valley height.
Different surface roughness compensation models require different inputs. Some require Rz, while others require Rq. It is important to note that these numbers come from different methods of summarizing the surface profile. Any conversion of values from one method to another should be handled with care and with the understanding that the conversion is only an approximation and will result in garbage going into the model.
To complicate matters further, there are different ways to calculate Rz. The most common methods for calculating Rz come from the German Standards Institute (DIN), the Japanese Industrial Standards (JIS), and the International Organization for Standardization (ISO). Obviously, everyone has their own methodology. Figure 1 shows an example of a cross-sectional profile used for roughness measurement. The existence of various summary methods becomes easy to understand.
Rz (DIN) utilizes the absolute average of the five highest peaks and five lowest valleys over the sample length.
Rz (JIS) utilizes the absolute average of the five highest peaks and five lowest valleys over the five sample lengths.
Rz (ISO) is the maximum peak-to-valley distance over the sample length.

Figure 1 shows a graphical representation of the cross-sectional profile.
It is easy to understand why there are multiple Rz methods to summarize peak-to-valley roughness.
Generally speaking, Rz DIN and Rz JIS are comparable, and Rz JIS is always smaller than Rz DIN because Rz JIS contains more data points.
Rz ISO is not recommended because of the small number of data points.