Impedance Measuring of a 24 GHz Rectifier
时间：04-05
整理：3721RD
点击：
Hey all,
I have designed a 24 GHz full wave Greinacher-Rectifier (with diode connected transistors). Now I want to measure it?s impedance. But: The measured impedance deviates from the simulated impedance! The left picture shows my simulation results during a frequency sweep (100 MHz to 26.5 GHz). The right picture shows the measurement. The deviation at low frequencies seems acceptable. But I totally don?t understand the deviation at high frequencies (especially why the imaginary part is positive)! The simulation results are more or less coincide with values from literature.
I used for measuring an Agilent/Keysight N5242A PNA
http://www.keysight.com/en/pd-867173...twork-analyzer
with ?Integrated True-Mode Stimulus Application?
http://www.keysight.com/en/pd-141668...msa-option-460
http://na.support.keysight.com/pna/h...MSA.htm#Create
For Topology I selected BAL-BAL and for Stimulus ?Single Ended?. The right picture shows Sdd11 which I achieved with this settings.
Has anyone an idea where this deviation come from?
Do I have to do a ?non linear measurement?? I do not have any experience with NVNA-Measurements. The problem is that additional devices seems to be needed (Measure Reference, Cal Reference, Splitter) which we had to buy first.
For clarity: The two differential HF-inputs of the rectifier are connected to Port 1 and Port 3 respectively. No output is connected with the VNA.
Thank you very much, and sorry about the English!
I have designed a 24 GHz full wave Greinacher-Rectifier (with diode connected transistors). Now I want to measure it?s impedance. But: The measured impedance deviates from the simulated impedance! The left picture shows my simulation results during a frequency sweep (100 MHz to 26.5 GHz). The right picture shows the measurement. The deviation at low frequencies seems acceptable. But I totally don?t understand the deviation at high frequencies (especially why the imaginary part is positive)! The simulation results are more or less coincide with values from literature.
I used for measuring an Agilent/Keysight N5242A PNA
http://www.keysight.com/en/pd-867173...twork-analyzer
with ?Integrated True-Mode Stimulus Application?
http://www.keysight.com/en/pd-141668...msa-option-460
http://na.support.keysight.com/pna/h...MSA.htm#Create
For Topology I selected BAL-BAL and for Stimulus ?Single Ended?. The right picture shows Sdd11 which I achieved with this settings.
Has anyone an idea where this deviation come from?
Do I have to do a ?non linear measurement?? I do not have any experience with NVNA-Measurements. The problem is that additional devices seems to be needed (Measure Reference, Cal Reference, Splitter) which we had to buy first.
For clarity: The two differential HF-inputs of the rectifier are connected to Port 1 and Port 3 respectively. No output is connected with the VNA.
Thank you very much, and sorry about the English!
No idea? Nobody?
Did you simulate and measured the circuit considering the same input power? Because if you are using Schottky diodes then their input impedance will not only vary with frequency but also with the input power available.
Either real circuit inductance or wrong reference plane (wrong VNA calibation).