HP8569B (Displayed signal and noise floor)

[gentlerv]

Dear all,

Owning a HP 8569B spectrum analyzer and knowing many specialist here who know this machine very well, I have a question:

All functions, like calibrating, attenuator steps and so on are, as far I can see, are good working

What I do find strange is that reducing the bandwidth results in a lower noise floor, which is normal, but also the complete signal will go down.
I mean with that the amount of graticule lines stay the same but the top of the signal go's down with the noise floor.
Is this normal?

If not, what could be wrong or has to be adjusted?

[Ice-Tea]

Have you set a marker on it? That will tell you the absolute value of the peak and whether or not that does down with the noise floor..

[gentlerv]

Unknown to me how to set markers, only have the standard spectrum analyzer and as far I know it does not include markers

[gentlerv]

I use the KE5FX GPIB software, in there can set two lines as markers
If I change the bandwidth the signal go's lower then the set marker line

[gentlerv]

When I connect the calibrating signal (100 MHz -10dB) to the input and change the bandwidth, the top stay's at -10dB and the noise floor go's down
That means it should work fine.
Then why, when I connect a antenna and measure signals those maximum levels does vary with the bandwidth ?
How can I find the real value from the displayed signals?

[Ice-Tea]

.. and as far I know it does not include markers

Yeah, sorry about that. Thought it would be the same as a 8592A who has markers. Totally different beast, though..

[cncjerry]

what signal are you looking at when you said the signal will go down?  I don't think the calibration signal will change with RBW changes on that unit.  Also, changing RBW could impact the signal level depending on the span you have set, but it shouldn't.  Noise level will drop with RBW but peak signal shouldn't (within reason).  There could be an alignment issue causing that problem or it could be the way you are measuring the signal.  If the IF alignment is out of spec then the filters aren't tracking the sweep correctly and that can cause it.  There is a service manual available.  There were many of those units but a good number have cracked bandwidth and other controls so take it easy on them.  By the way, changing the span does change the accuracy of that unit I remember.

The accuracy of that unit is like .5dBm so how much is it changing?  Is it all RBWs or just a one?

I don't remember it having markers but it has GPIB and digital capture.  I thought it had trace averaging too.

I would measure an external signal through a direct connect cable but remember not to exceed like 30dBm.     

[cncjerry]

there is a preselector control on it.  You might have to peak it depending on RBW.    Also, if the span and RBW changes you might have to slow the sweep time to get an accurate level.