Advantest R3465 Repair Help

[Sesta]

Hi guys,
I bought a nice R3465 Adantest 8GHz spectrum analyzer cheaply. The instrument seems works but the trace shows a noise even when the signal is apply.
After a rapid check I saw that the IF section does not work well, for any signal injected (421,4MHz) nothing was shown on display.
I opened the IF cover (see picture below) and now, with IF cable disconnected, if I touch the PCB trace I see the noise moving on display.
I need your help to investigate better or to clarify the circuit on the photo.
The problem seems to be on the first IF filter but I don't know if we can open it without creating additional problems.

Thanks a lot for any help-information about it.

Alberto.

[chick0n]

This is the second converter. I think, it takes in the 4231.4MHz from the 1st mixer, through a bandpass filter (The 2 Golden Screws). Then it gets mixed with 3810MHz, from the LO (Bottom Golden Screw). The Mixer Diode is marked "2R".

Then it puts out the 421.4MHz 2nd IF.

The coax cable on the bottom left takes a "sample" from the LO, for the PLL.

Looks very familiar to my R3273.

Better dont open the Filter, or turn the Golden Screws!!


If im understanding correct, the analyzer doesnt show a applied signal?

And you injected 421MHz into the 2nd converter? Thats wrong...the 2nd converter needs 4231.4MHz input.

Did you test the input step attenuator? Maybe its broken...

[Sesta]

Thanks chickOn for the explication, after a better analysis I also came to your conclusion.
You have understood correctly, applying the signal to the second IF (4231.4MHz) nothing is displayed.
I looked for LO and noticed that the second amplifier was not working properly.

Not having information about it, I modified the final stage (as shown in the photo).
I used a HBFP0420 (found in the drawer) with its voltage regulator and I finally saw the display moving.

I applied the input signal and it WORKS !!!

The video signal is slightly lower than expected (about 30db less) so I will have to modify the circuit again to increase the gain.

[chick0n]

Nice.. I think the original transistor Q705, is a "2SC3604". Hard to obtain....

[Sesta]

After some other check I think the mixer diode is broken.
I replaced HBFP0420 with BRF181 because the signal at 3,8GHz was very low.
To test it I connect the signal 4231.4MHz (-10dbm) to the input of the second converter and my spectrum analizer to the IF output .
On the IF output we see the above signal (3,81GHz -70dbm and 4,2314GHz -60dbm) both with the same attenuation (see picture) but at 421,4MHz I only see noise.

Do you agree with my analysis?
Do you have any suggestions about which diode I can use?
Any advice is appreciated.

[chick0n]

The Diode marked "2R" should be a hsms-8202. Still available. Maybe you can test with diode tester.

I dont know if your 1st IF really is 4,2314GHz. I just concluded, because in R3273 it is.

Maybe try to inject 4,2314GHz, after the bandpass?

Maybe LO power is still too low? BFR181 datasheet says, it starts working at 6GHz. Maybe 4GHz is too low?

[Sesta]

Yes, also on R3465 has the first IF is at 4.2314GHz, I have anyway try and also bandpass is fine tuned at this frequency.
I'am not able to test real power of LO but I checked the voltage ratio and is similar to first LO stage.
The BFR181 the FT is at 8GHz, on R3465 the PLL run at 200MHz and harmonic mixer works at 3,8GHz.
I assumed that LO power is OK because on 2nd IF out the power level is similar at the 4,231GHz signal.