SSA3021X input blown

[wofritz]

Hello,

this is my first post in this forum, so a short introduction: I am a retired EE, worked in telecomms R&D mainly embedded software but also some digital hardware/FPGA. In the hobby area I am doing small projects with Arduino and Raspi, also some small RF projects and antennas.

My problem: My SSA3021 level display is suddenly about 15 to 20 dB low.

Hardware: Siglent SSA3021X from 2016 (Same version as in the tear down video I found here). Attached images of input stage and partly reverse engineered schematic.

Problems: Displayed level is about 15 dB down, changing over the frequency range. Input impedance not 50 Ohms.

What I've done so far:
- Read several threads about similar problems here, they helped a lot but they are not matching my hardware version.
- Measured input impedance under power with MiniVNA pro. Input impedance is far from 5o Ohms in the range 0.1 to 200 MHz as low as 6 Ohms
- Tested various diodes in the input section with HMC 8012 in diode test mode. D7 is showing around 3V forward voltage, others show expected values around 0.8V, so I think D7 is defective.
- Switched through input attenuator and preamp. Noise floor changes, but level does not change, so I assume attenuators and preamp are working correctly.

Questions:
- What type are the diodes, especially D7 (see picture). I could not find anything meaningful for 8GS, GGS, 6GS and other combinations of possible interpretations of the marking.
- What is U1, U2 etc. marked 955C? I did not find anything on the web.
- What is the purpose of U1 and U2? I see U1 can disconnect the input and/or switch 50 Ohms termination in or out, and U2 can select input signal or an unknown internal source (trace seem to disappear in an inner layer, did not follow yet), but what it this used for?

Any replies welcome, also point me to relevant threads in the forum which I have missed.

Meanwhile I'll continue my investigation.

[tautech]

Welcome to the forum.

Away from my resources for a few more days but following the links in this post might provide you with the answers you seek:
https://www.eevblog.com/forum/testgear/siglent-ssa3000x-spectrum-analyzers/msg4653826/#msg4653826

[rf-messkopf]

Problems: Displayed level is about 15 dB down, changing over the frequency range. Input impedance not 50 Ohms.

Are you able to feed in a low-frequency input signal (e.g. a couple of MHz) and do some signal tracing with a voltmeter or a scope? This would allow you to determine if the fault is actually in the SPDT RF switches you are showing in your pictures, or perhaps further down the signal path. Another possibility would be a blown attenuator or first mixer.

- What type are the diodes, especially D7 (see picture). I could not find anything meaningful for 8GS, GGS, 6GS and other combinations of possible interpretations of the marking.

No idea, but it seems to be connected to the control line of the RF switch.

- What is U1, U2 etc. marked 955C? I did not find anything on the web.

Some SPDT RF switch. It seems that nobody was yet able to identify this part in the other threads on the SSA3021X. Here is some more related stuff if you read German.

[2N3055]

Take a look here...
Maybe it helps..

http://emccompliance.co.uk/fixing-and-protecting-the-rf-input-of-a-spectrum-analyzer

[wofritz]

Away from my resources for a few more days but following the links in this post might provide you with the answers you seek:
https://www.eevblog.com/forum/testgear/siglent-ssa3000x-spectrum-analyzers/msg4653826/#msg4653826

Thank you very much.

I followed the links in your post and even the hardware versions mentioned there are different than mine, I got some useful information. The RF switch marked 955C might be a  MASWSS0181TR-3000, at least the pinout and package are the same. According to the data sheet the last characters on the packages are a date code, that explains slightly different markings of that IC in other pictures.

Are you able to feed in a low-frequency input signal (e.g. a couple of MHz) and do some signal tracing with a voltmeter or a scope? This would allow you to determine if the fault is actually in the SPDT RF switches you are showing in your pictures, or perhaps further down the signal path. Another possibility would be a blown attenuator or first mixer.

I could do this, but I find it difficult/dangerous to power it up with the board outside. I have to think about this a bit.

OTOH, I checked the attenuators and RF amp by manual configuration and the behaviour seems to be OK. The noise level moves but the signal keeps displaying a constant level.

Take a look here...
Maybe it helps..

http://emccompliance.co.uk/fixing-and-protecting-the-rf-input-of-a-spectrum-analyzer

Thanks. The hardware revision in that video is newer than mine, but it was helpful anyways.

   

[TurboTom]

I guess you nailed the microwave switch more or less -- my research ended up on this same type as well, though there may be similar parts from other manufacturers available. The main deciding point is the negative voltage drive which isn't available with all such switches (HMC595 for example, which may have matched the case markings better...). The diode likely is a BAR63-02W, provided the marking can be interpreted as "GG".

I hope you get your SA running again...

[wofritz]

I guess you nailed the microwave switch more or less -- my research ended up on this same type as well, though there may be similar parts from other manufacturers available.

I got the parts from Mouser today (2 days from Mansfield/Texas to Schwüblingsen/Germany, amazing!).
I think I can confirm that the 955C chip indeed is the MASWSS0181TR-3000. My chips are marked 952R, and from the data sheet the 95 is the type number and 2R is a date code.

Now for the tricky part - replacing the chips with two left hands with 5 thumbs each...

[wofritz]

Success! I replaced the first three RF switches and a diode and now the level display is correct again.

It was quite hairy regarding my very limited SMT rework skills and my highly professional rework equipment (an upside down flat iron with a thermo couple glued to it for pre-heating, a very cheap hot air gun and a WECP-20 soldering iron), but in the end it worked out fine without destroyed traces etc.

I wouldn't have managed it without the forum and the lots of useful info found here, so thanks to all of you!

[mawyatt]

Nice work!!

Were all 3 RF switches blown, or did you just replace in case some were slightly damaged?

Best,

[wofritz]

All three were blown. I replaced them one by one until it worked.

[tautech]

All three were blown. I replaced them one by one until it worked.

Very cool, congrats ! 

Now, others will find themselves in the same pickle as you did and would benefit from a summary of your repair including any PCB HW version markings your might have and links to sources to get the parts and any pictures you might like to provide....or even a video. 

[wofritz]

Find summary here:
https://www.eevblog.com/forum/repair/siglent-ssa3021x-hw-07-03-00-frontend-repair/