HELP - HP 3562A DSA - negative spikes below nose floor

[innkeeper]

I'm looking for help on this, as i am in unfamiliar territory.

I scored a HP 3562A in unknown condition.
for the most part it seems fully functional, it needed a memory battery replacement, which i've done.

didn't know how else to describe  what i am seeing though
pictures below. 2 with no input, and one with the source output connected to the first channel.

This happens on both channels.

I don't know if this is jsut how this DSA works or if it is a problem, and if it is a problem, what to start looking at.
or if its just operator error.

It is certainly not behavior as i am used to with my HP spectrum analyzer, and not something that's described in the Service manual or operations manual.

no input (noise floor) auto range


no input noise floor manual range -100 to -180db


signal from source 500mv peak, 50khz, display  10khz span and manual Y 0.0db to -150db

[Johnny10]

I have the HP 3561A single channel.

The linear spectrum display shows this type of display.

The range you have chosen may be why it shows those lines below noise floor.


After doing some reading of the HP 3562A specs... it has quite different capabilities than the 3561.

[innkeeper]

do you see artifacts below your noise floor?

[innkeeper]

i did a bit more playing today.

I was able to get a clean full sweep  with only 10mv output from the source, auto y scale, but as i ramped up the source output, the number of artifacts increased until it ended up looking like that last picture about 500mv as i recall.

these artifacts go down to about -800db, so i know there truly artifacts of some type.  I would find it hard to believe this is normal though, i've never seen a spectrum analyzer behave like this .. but if someone familiar with this DSA says it is i can accept that. if its not i want to find the cause.

[Kleinstein]

These could be just artificial effects of the logarithmic scale. With very low levels there is slight chance that due to meter internal noise the power reading gets close to 0 or even negative. This can happen if they try to extend the readings to very low levels, e.g. by subtracting offsets. This will result in very negative dB readings.

I don't know that instrument, but at level of some - 100 dB I would not be surprised to see such artifacts in such an old instrument. So for me it would not be anything to worry about. The extend of the spikes might depend on the adjustment - less negative spikes might be more convenient, but might be also less accurate - having a higher background noise level.

[Johnny10]

I was thinking... where have I seen a working 3562A and remembered that a youtube video on updating this unit.

You might ask this gentleman if he has same experience.

D-Lab

[innkeeper]

I'll see if i can shoot him an email.