Should network analyzer always keep on for accuracy?

[chihhao]

Some enigneer they never turned off network analyzer for
the sake of measurement accuracy.

Is it true ?

Thanks a lot.

 

[jallem]

I dont think is accuracy, the motive is not to do the
calibration/setting every day. But, then again you would say
ahha...so they need accuracy..!!....
But remember some old NA does not store the setting
and every time you turn it off you will need to set the
star and stop frequency, /div, etc,etc...

 

[pisoiu]

Considering the lifetime of the CRT for older models and CCFL for newer, I do not think this is a good idea.

/pisoiu

 

[sergio mariotti]

please don't confuse accuracy and drift.
Accuracy, or better, type B uncertainty, is fixed by VNA itself and CalKit.
type A uncertainty is the statistic uncertainty and may be reduced by large averaging. Large averaging are possible only if the drift is negligeable.
Drift is negligeable if: room temperature is stable, electronic amplifilers, mixers, filters, has been turned on by 1 hour (at least), Quartz oscillators has been turned on by few days (at least), cables never moved during measurements.

About the life of VNA: what is more dangerous for VNA's life? the life time or frequent turn on - turn off ?
For some one the life time is a risky parameter tof failure, for someone is more dangerous the frequent turn on - turn off.

 

[flyhigh]

Hi,

we keep our VNA swithched on for years. It is basically to avoid shocks on power on/off. The stability of calibration is remarkable, we cheched it a few times and it is still excellent. It is HP 8510F. We have some dedicated calibrations to cover various fixtures we have, they cover all of the band and you can use callibration subset to focus on particular band of interest. The problem of cables is elegantly solved by HP (Agilent) with port extensions that are flexible enough to allow any required connection.

flyhigh