I read everywhere that loosing the CAL data when the battery dies is like planning a ride through hell but nobody actually demonstrated what is involved when such thing happens.?

while troubleshooting my S.A, I did a DEFAULT CAL recall and it recalibrated itself fine after.

I even read the manual and it explains in details what to do and srite down measurements compensation. I did not read however how those manual readings are then set into the instrument’s memory. ??

Does someone knows why is everyone afraid so much about the battery and the involved procedure?

I Think I found the right section in the manual :

Reloading the Correction Constants
This procedure assumes that you have valid correction constant data
from a previous backup. Without backup data, new correction constants
must be generated by performing the adjustments in Chapter 2.

Hi Nicolas,

I can report from my calibration experience with an HP8593E. I assume there is no big difference in procedures to your HP8594E.
I purchased a used HP8593E to find that is was only calibrated to 18GHz. For the frequency range 18...27GHz there were no correction coefficients, so level readings were way off.

Calibration required a stable generator for the frequency range in question, a high quality splitter and an accurate power meter. You have to take a level measurement with the SA at every single correction grid frequency (there are many!) and write it down. You find those lists in the manual.
Then you have to enter all those numbers point by point into the instrument. I used the front panel keys to do that.
This may be automated via GPIB, but you still need to get the numbers into a computer.
It was very tedious and I refrained from re-calibrating the range below 18GHz. I do not ever want to do this again.
That's why in the same pass I replaced the CMOS memory battery. Now, I should be safe for another 10 years.

Best regards,
Tom

开云体育

Much depends on how the equipment was designed.? For a scope, you could need a controlled square or sine wave source, various amplitudes and various frequencies.? Often it needs to have a particular command set and a controller (computer) to automate the tests.? Often, the tests are not specified, simply programmed.? The code source is not always documented.? All of this is true if and when the code is even available.

It can get nasty.

It also, as you note, depends on which set of constants you need to regenerate.?

Not all calibration procedures are documented to the extent you can perform them, even if you had the equipment.

Harvey