Advantest R3265A Frequency offset

[neuschs]

Good day,

I have a R3265A which worked till last week. Now the self calibration doesn't work anymore - the calibration signal ( 25 MHz at -10 dBm) shows up as 28.74MHz (attached image).
I've checked the calibration signal and it is perfectly fine on another spectrum analyzer and the oscilloscope. I've generated a 230MHz signal and it is alsooff by 3.86 MHz.

I potentially connected the 421 MHz IF Out to a trigger generator (not totally sure but it  is possible that this happened). I am no professional regarding the innards of this device and the service manual are nearly too much information.
I dont know where to start searching.

Is it possible that some intermediate frequency generator shifted that amount ?

I would be thankful for any advice.

Greetings
Sebastian

[bidrohini]

It is possible that connecting the 421 MHz IF Out to a trigger generator could have caused a shift in the intermediate frequency generator. The intermediate frequency (IF) generator is responsible for converting the input signal to a lower frequency that can be more easily processed and measured by the spectrum analyzer. If the IF generator is not functioning correctly, it can result in an incorrect frequency measurement.

In addition to checking the IF generator, it would be worth checking the other internal components of the spectrum analyzer, such as the mixers, amplifiers, and filters, to see if they are functioning correctly.

It is also worth checking the internal reference oscillator of the spectrum analyzer, as it provides the reference frequency for the IF generator and other internal components.

Since you are not a professional with this device and the service manual is too much information for you, it would be best to contact the manufacturer or a qualified repair technician for further assistance with troubleshooting and repairing your spectrum analyzer.

[neuschs]

Thanks for your answer, I am shure under commercial circumstances this would make sense, but I am very certain that this would be way more expensive than the initial cost.

I am surely no professional regarding spectrum analyzers, but I sure know a little bit about RF mixers, intermediate frequencies and synthesizers.

I went through the performance validation routine but I have to stop for example because I dont have the right gear (e.g. a frequency synthesizer which goes up to 6 GHz and beyons).  Therefore I asked if someone with more experience could nail down the issue to a smaller block in the block diagram.

Additional information since last time: I tried to calibrate the spectrum analyzer with a signal he thinks is 25 MHZ so 21.5 MHz but I get an Attenuation Error otherwise with the correct 25 MHz I get Calibration Signal undetected.

The reference frequecy out is fine (measured 9.9993 MHz) as well as the generated 25 MHz signal.

[neuschs]

For anyone having a similar issue: https://loetlabor-jena.de/doku.php?id=projekte:r3271-repair:start
It is in german, but a short break down: The 2nd LO generates 3810 MHz. The VCO is controlled via a coupled feedback which gets mixed with 3800 MHz the difference (10 MHz) is then phase compared to the a reference.
The generation of the frequency 3800 MHz (and mixing) happens on assembly THD296. This ceramic pcb was manufactured with silver ink or something similar, but it oxidizes and the trace resistance grows ridiculous, I can verify that for my part.
This results in generation of none reference frequency for the phase detector which therefore lets the VCO drift upwards and hitting the upper boundary.

Sadly this isn't my only problem. The THD296 works with a step recovery diode for generating 3800 MHz from 200 MHz. But it seems this diode is broken and for specially this assembly there is no parts list or schematic in the service manual.
I will verify my judgement by changing the whole assembly for a working one, but I intend to get the original one running.
One step I am planning is to "reverse-engineer" the pcb and try to fix it.

Another Idea that came to my mind: Is there maybe a building block which multiplies 200 MHz to 3800 MHz to easily replace this part. Step recovery diodes are practically not available at all.