Advantest R3361A [FIXED]

[razberik]

Greetings to everyone !

One month ago, my dream became true, I finally bought my own spectrum analyzer. I have been roaming around german ebay for few months and I missed a few good listings. Once I suddenly discovered an offer located here, locally ! One quick phone call to arrange the meeting and Advantest was home before midnight. Price equally 740€ was too attractive to take a risk due to no other RF equip available to test it.

Funkwerk appeared to be functional watching numerous frequencies exactly at the place with simple antenna. However, calibration was unsuccessful due to low CAL signal. CAL signal appeared to be lower about 40-50dBm depending on ATT setting.

So investigating in next post...

Whole album:
https://picasaweb.google.com/100683312221737332313/6278363619203042561#

[razberik]

I love to do teardowns, that might be the reason I come to this forum somehow ([australian high pitch voice]ITS TEARDOWN, TUESDAAAAAY).

1. Cleaning
So I disassembled the whole machine apart before any further work. Lots of dust everywhere. Field strip wasn't comfortable, you crazy Japs.
One thing I dont understand, why I have to unmount the radio part (TG) to reveal the one of screws holding television ?
Well, I have seen skelets with more friendly disassembly. Have a look at photo.
What is very usual, CRT screen has burnt in, however it is still good readable.

One sunny Saturday terrorizing my neighbours with vacuum cleaner and then reveal my memory which screw belongs where...

[razberik]

3. Power supply

The most popular failure, flowering elyte caps. One of them was watching at me, however I didn't trust others, so DER DE-5000 has its call of duty.
Most of them were OK within specs, however I tried to replace everyone if proper replacement was available in my home stock. Some of them I kept.
I was worried about the mess on the board, but nothing extreme. I was curious about resistors with bleached marking, but there are alright. I kept them, I only reflown solder joints.

This colleague was doing something about PSU, I presume with help of translator that topic is about the effect of worn elyts on spectral purity
http://honda.way-nifty.com/pocky/2013/04/r3361a-c2e1.html
So checking all of elytes is important.

[razberik]

4. Low level problem

As I mentioned, detected signals were exactly at the place in terms of frequency, but they were too low in amplitude.
Working on this machine without documentation is little bit hassle, so I had to sent some papers to printer. Omitting computer boards and merging A3s ... still few hundreds of pages.

Attenuator - Isn't it good practice to start here ? Every step appeared correct. I presumed ATT is OK.

Follow the service manual - there is a beautiful troubleshooting flow chart.
So, is signal from 3rd CONV 3.58 MHz going into IF -25dBm or more ? No, it isn't. (IF OK)
Is signal 226.42 MHz from 2nd CONV going into 3rd CONV -27dBm or more ? No, it isn't. (3rd CONV OK)

Here comes trouble. They tell me to look at the connection between 1st CONV and 2nd CONV, that's at 4 GHz. That's a lot. We have R&S FSV in work, however 3.6 GHz only.
This means to me I have to travel a lot, I know some places where they have more GHz's, but I don't want to shake this Jap in vehicle if I don't have to.

So, is it really necessary to visit places where they have more than 3.6 GHz ? Maybe there is some little problem which can be fixed quickly...
There are a lot of stories about bad mixer diodes and bad solder joints, one successful example here: https://groups.yahoo.com/neo/groups/advantestinstrument/conversations/topics/16
Let's uncover these fragile ceramics. Using magnifier, everything looks perfect. No signs of previous repair attempts.

Turn on 34401A, Ohms. Chassis touch and then track the resistance from input and further. Everythings normal until ... BINGO !
Soldier joint between BP filter ceramic and coupler ceramic has a few more ohms, I remember about 500 ohms. That's not supposed to be, so let my Weller warm up.
Gently touch the bridge until it melt and then let it be.

Quickly attach the cover, flip the machine and turn it on ... PERFECT, CAL signal exactly as expected. Calibration successful, testing procedures OK.

I am happy on this successful repair of the radio section. Have a look on pictures.

[nctnico]

I used to own a R3161A and it had similar problems in the PSU and in my one I also replaced the electrolytics. My unit had a  problem on the same ceramic circuit as yours: one of the traces got intermittant which caused a dip around 2.2 Ghz (or something) so check your unit with a frequency sweep to check whether the response is flat.

[razberik]

Thank you for advice, that's a good idea. Bud again, I don't have such an easy access to signal generator capable of 2.5 Ghz or more. So it may take some time.

5. Tracking generator

When I make a connection between RF OUT and RF IN, I don't see a flat trace.
So, what about isolating a problem ? I connected oscilloscope to RF OUT and observed the output. DS1054Z show me almost exactly the same as seen on CRT. Well, at least to 200 MHz ...
Shall it be only the TG to be not working ?

To be continued ...

[nctnico]

Check the specs of the tracking generator first. It may be that it isn't flat for the first several MHz. Odd as it may sound but making a very wideband RF amplifier which works from tens of kHz to several GHz is hard.
Also: these spectrum analysers can be uphacked to 3GHz by altering the eeprom contents.

[razberik]

6. TG fixed

I wonder if this was an educational malfunction, anyway...
I removed bottom cover of TG and saw another ceramic LP filter. Solder joints were OK, so I left them alone.
Then I called my Rigol for his duty and tracked down where exactly the signal is leaking changing significant frequencies. 50 MHz was exactly allright, but e.g. 2 MHz was not. Level was low and there was a significant jitter. Tracking backward showed me, that the signal entering output amplifier was wrong, so I assumed that output amplifier was OK.
So I de-attached the TG housing, removed cover and then I saw another circuitry on ceramic. Solder joints were OK on ohm-meter, bud then - I don't know really why, I reflown all these solder ball.

Cover back with a few screws. Then I did extension from YTO to TG mixer (see photo) and turned on machine. TG was miraculously working again.
My neck was already hurting because of reading the screen flipped 180° (I can't really read upward text like me beloved one), so brief examination of functionality. Considering lack of proper RF equipment and the fact I wasn't sure if SA is completely alright I decided to final assembly. At least until new battery arrive or RF lab is visited.

I hope this repair log was somehow helpful (if somebody is going to repair another Advantest of this product line).
Conclusion: Always check solder joints between two pieces of ceramics. They seem to be siting firmly in aluminium cavity and there is really lack of movement to stand thermal stress over many years.

[razberik]

7. What now ?

I am aware that if I don't visit some place with proper RF equip I am only divining from crystal ball (I use this beautiful word in czech language).
What is more, few questions may come from my lack of knowledge, TBH I consider myself an RF newbie.

Anyway, if I connect any cable between RF OUT and RF IN, I obtain almost the same plot. Does it mean something ?
Whether it is N-N cable, BNC-BNC with damaged shielding, BNC-BNC hi-quality, characteristics look similar.
The wave has dynamic range about ... 1dB±0.1 ?

What should I consider now ?

Thank you, cheers !

[mrf245]

Hello razberik.
I also own R3361D, 3.6G with TG.
As your last photo show, I think you can check:
1 TG output flatness. Set span=0,let TG output some point of special freq, use your another SA (or other equipment) to check TG output flatness.
2 RF input response. Just use signal generator, output some special freq feed into RF IN port,check your receiver response.
Or maybe there are some other easier method for you to check that problem.

[DrElectro]

Hi Razberik ,

Although this thread is ancient I wanted to let you know I found exactly the same fault with the TG in my R3361A.
TYVM for your original post, it probably saved me hours (in fact I may not have even bothered otherwise).

Mike