AnsweredAssumed Answered

newcomer questions

Question asked by K on May 5, 2007
Latest reply on May 11, 2007 by johanna_larmour
Hi Everyone,

I've started working as a graduate student at a lab and I have some basic questions about the calibration procedure for a network analyzer. 

We have bought a used HP 8752C network analyzer and no calibration kit for it.  I've been told that another student has performed a calibration for it by using standard connector shorts and 50ohm load.  For the open circuit, they left the connector not connected to anything.  Afterwards, the student was saying that he checked the response by attaching a waveform generator and sending a sinusoidal to it.  He said that when he set the frequency to ~100kHz (low frequency), the frequency peak was at the expected value.  But when he sent a higher frequency signal at ~20MHz, the peak was very broad (span was about 100kHz). 

My questions relate to whether or not the steps he took were correct?  Will using standard connector shorts, opens and loads yield acceptable calibration results?  Also, it doesn't seem like feeding a signal directly to the network analyzer is a good way to check the calibration, am I correct?  

Afterwards, I've tried checking the response using a passive device (a low pass filter) and the response was as expected though.

I also tried using the "System Verification Tests" that is available to the network analyzer using the same short, open, and load connectors that were used for the initial calibration attempt and it failed all the tests.

My previous experience with network analyzers is limited.  The last time I calibrated a network analyzer was with a proper calibration kit and was a few years ago.  Any advice would be greatly appreciated!  Thanks kindly.