I got a 8560E which was working (sort of) except the display was wobbly and signal trace was distorted (not the graticule and characters) and it was giving errors all related to LOG AMP. However, it was showing the signal and I could do everything with it except the distorted shape of signal and noise floor. Certainly the LO and converters were good.
So I opened it up and turned out all power supply lines are OK except the -15V which was oscillating badly (405khz from ~+4V to ~-17V or so). Sure enough the output 220uF cap on that line was dead and the linear regulator had started to oscillate. I took the PSU out (left the HV module attached inside the unit) replaced the 220uF caps on all 4 rails, tested the PSU outside the box by attaching a DC load to the 5V rail (the switcher gets it feedback from this rail) with only 1A load and the PSU started and all voltages were good (didn't go beyond 1A).
I put the PSU back in and .... nothing :-( the SA is dead. I checked the voltages, +110V is good, -12.6, 28V and -15V are all perfect. But 5V is down to 3V and +15 is down to 9.6V. They are clearly overloaded because I checked the overcurrent protection circuit and sure enough it has overtaken the error amp in the linear regulator which means the supply is in CC mode. The +15V is giving about 2.37A (judging by the voltage drop across the 0.07 ohm current sense resistor) Since the +5 rail is down, the PWR_UP signal remains down and the system does not even boot. The resistance of the 5V rail to ground is 1.5 ohm :-(
After several hours of trying to locate the possible short by disconnecting/connecting various combinations of boards and connectors I think at least for the 5V rail I have narrowed it down to the Freq Control board and in particular the LODA. DC resistance of this rail to ground is only 1.5 ohm as soon as I connect the freq control board. now, on that particular board, the DC resistance of 5V rail to ground is 4.4 ohm if LODA is disconnected and 1.5 ohm when connected. That 4.4 ohm increases to 13+ ohm if I disconnect the YTO as well.
Now my question is how much current these components really draw (LODA and YTO)? Is this kinda rail resistance normal? How could the LODA be bad when it was working before I repaired the power supply. I never touched ANYTHING on that side of the spectrum analyzer. Also where could be the fault on the +15V rail? that is not connected to LODA.
How much current this power supply should deliver before going into CC mode? maybe it is just not able to supply the required current for some reason...but why? it was OK before I replaced the output caps.
Any help is appreciated