HP 8593E phase noise

[bistromath]

Hi all,

I have an old 8593E spectrum analyzer that seems OK except for a huge amount of phase noise at bandwidths closer than about 100kHz. Here's what it looks like.



I've been down a number of rabbit holes and have the Artek manuals. The PSUs have been recapped. I've attempted to confirm the power supplies are clean, I've tried driving the YIG coil from a battery supply to rule out FM noise from the power supplies, I've done all the calibrations and tried removing the counterlock board as well. The noise is on the first LO at 4.223GHz as well as every subsequent signal.

So, before I unload the results of a week's worth of diagnostics here, have any of you wily old campaigners seen something like this before? Should I resign myself to it being a bad YIG and move on?

Nick

[Perrold]

Gut feel is your YIG is fine. From the pic, it looks like the noise occupies discrete sidebands and there might be higher level noise hiding inside the filter curve at half the observed offset frequency. No idea what is causing it though and you've certainly done all the due diligence at this point. Doesn't take much noise riding on PS rails or getting in through the YIG coils to cause this though. Some unaccounted for ground loop still being present with the external supplies is a possibility. The YIG in these spectrum analyzers is also very susceptible to magnetic field pickup, at least at 60Hz. A small shaded pole motor running about a foot away caused me to do a completely unnecessary troubleshooting deep dive.

[grendle373]

looks like a noisy yto or bad res bw board set.  loop the cal out to the input and run the self cal. If it passes it's most likely the yto.  Noisy yto's are common on the 60 and 90 series. They can be found used or rebuilt.  The res bw board sets usually fail for amplitude but we have seen them get noisy a couple times.  My bet would be they yto. Third converter could be a possibility as well.