Hi,

I am measuring a filter response on a PCB I made. I would like to see the response of the filter when the output is connected to a high impedance load. I am using an N5230C Network Analyzer with two 50 Ohm ports which loads the circuit and changes the filter response. Is there a simple way to measure the filter response as if it was seeing a high impedance load, using the N5230C?

I have tried adding a series resistor at the filter output to increace the load impedance seen by the filter and have the network analyzer measure the voltage divider output, but the series resistor adds significant parasitics to the measurement. I also noticed that the N5230C has an option to both change the system Zo and do a Zo conversion on the measurement results. I have tried using these features but they have given me confusing results. For example, when I increase the system impedance to a high impedance (to minimize the voltage divider ratio) the S21 measurement will increase as the system Zo arroaches the filter impedance, but will then decrease as the system Zo goes higher. I would expect S21 to continue to increase as the load impedance increases since the output voltage should increase as the voltage divider becomes more like an infinite impedance. In this case, am I really measuring the power transfered to port 2 and not the voltage seen at the filter output?

Thank you for the help,

Chris

I am measuring a filter response on a PCB I made. I would like to see the response of the filter when the output is connected to a high impedance load. I am using an N5230C Network Analyzer with two 50 Ohm ports which loads the circuit and changes the filter response. Is there a simple way to measure the filter response as if it was seeing a high impedance load, using the N5230C?

I have tried adding a series resistor at the filter output to increace the load impedance seen by the filter and have the network analyzer measure the voltage divider output, but the series resistor adds significant parasitics to the measurement. I also noticed that the N5230C has an option to both change the system Zo and do a Zo conversion on the measurement results. I have tried using these features but they have given me confusing results. For example, when I increase the system impedance to a high impedance (to minimize the voltage divider ratio) the S21 measurement will increase as the system Zo arroaches the filter impedance, but will then decrease as the system Zo goes higher. I would expect S21 to continue to increase as the load impedance increases since the output voltage should increase as the voltage divider becomes more like an infinite impedance. In this case, am I really measuring the power transfered to port 2 and not the voltage seen at the filter output?

Thank you for the help,

Chris

Since S21 is a power gain and power can be related back to voltage for a particular impedance, can we then say that S21 is equivalent to a voltage gain only when the DUT impedance matches the system impedance? Then, if I do an impedance transformation and I set the new impedance to the DUT's impedance does the new S21 values represent a voltage gain?

Am I approaching this problem correctly by trying to use a network analyzer or is evaluating a filter response better suited to an impedance analyzer?

Thanks,

Chris

The N5230C is a PNA-L; I'm not sure if that's what you meant.

In my case I'm not too concerned about the filter Q as much as just trying to verify the response across frequency and see if it has good attenuation in the stopband. Are you suggesting that I do something like a conversion from s-parameters to Z-parameters and solve for the output voltage across a high impedance load? Will this work if the DUT is not matched to the system impedance and what will the S-parameters be referenced to (the system impedance or the DUT impedance)?

Regards,

Chris