Does anyone know of anything or any way available to calculate the mismatch uncertainty improvement that you can get from using an attenuator pad? It is generally accepted that an attenuator pad with a good VSWR in front of a load with a poor VSWR will decrease the mismatch error, but I can't find anything that quantifies the improvement. Best I can find are broad statements like "a well matched attenuator pad with very low VSWR will virtually eliminate uncertainty due to mismatch". Hard to put that into a mathematical error analysis. Especially when "perfect" attenuators are so hard to find.

For example if you have a load with a 2:1 VSWR being driven from a source with a 1.5:1 VSWR the mismatch uncertainty is +0.56dB/ -0.6dB. What if I insert a "real world" 2dB pad with a 1.25:1 VSWR what the resultant mismatch uncertainty now be and how can I calculate it for various real world values?

For example if you have a load with a 2:1 VSWR being driven from a source with a 1.5:1 VSWR the mismatch uncertainty is +0.56dB/ -0.6dB. What if I insert a "real world" 2dB pad with a 1.25:1 VSWR what the resultant mismatch uncertainty now be and how can I calculate it for various real world values?

http://cp.literature.agilent.com/litweb ... 2-1087.pdf

The equations that you want are:

Plus=Pad_S21/(1-SM*Pad_S11-LM*Pad_S22-

SM*LM*Pad_S21*Pad_S12+SM*LM*Pad_S11*Pad_S22);

Minus=Pad_S21/1+SM*Pad_S11+LM*Pad_S22+

SM*LM*Pad_S21*Pad_S12-SM*LM*Pad_S11*Pad_S22);

SM=source match, LM = Load Match, Pad... You get it.