HP 8568a input level low.

[Simon Spiers]

Hi everyone, wonder if the experts can help with a little problem.
Work have finally scraped the HP8568a SA after the power supply blew up a few years ago, so it came home with me.
After replacing a shorted capacitor on the PSU the thing seemed to work fine.

However when running calibration I can't get the -10dB level high enough for the calibration to continue and it stops with message adjust attenuator input.
I have run the various self calibration tests and none complete.
It looks to me as if the input has lost gain, or maybe the limiter has failed?
The attenuator looks like it's OK despite my first thoughts otherwise.
Any help would be gratefully appreciated.
Here's the result

[cncjerry]

Put a known signal into the SA and post the results.  Or measure your calibration signal output on a scope into 50ohms.

Also, check the selection relay for the two inputs buy putting the signal into both to make sure the relay is working properly.  I have one that has a problem with the relay and I have to punch it several times to get the right level.

A lot of the 8568a/b analyzers have blow inputs.

[MadTux]

If different step attenuator settings don't change the level displayed, chances are high that either the AC/DC input selector relay is broken (in front of attenuator) or that something further in is toasted.

You might do a sweep from a few Mhz to 1.5GHz and see whether that changes anything (level should get progressively better if input selector relay is dead, since impedance of broken relay (capacitive coupling) decreases with higher freq. 8568 has a coax RF limiter diode in front of input. It probably gets fried first, if input is overloaded. Remove it and see whether it changes anything, but install new one if it's broken before use, so the front end mixer lives longer. I have bought and rebuild a few 8568s and none had a broken input mixer.

BTW, if you dial in 0dB into attenuator setting, you can basically remove any attenuator elements from the equation. Also that noise free baseline IMO doesn't look promising. Didn't see you are in 1db/div, so that is normal

[cncjerry]

yes, just watched the video and agree with MadTux.  That noise baseline looks like way too low.

[Simon Spiers]

Thank you all for your input and time.
I have resolved the problem.
The problem was me really.
The link lead I was using was a scope lead designed to be used on a 1 meg input.
The lead is resistive and as a result makes a great attenuator when fed into 50 ohms!
Sorry everyone!