Advantest R3361B S.A. sampler: built to fail - advises needed to fix it

[egirland]

Built on a ceramic substrate, this sampler it is very prone to crack failures.
My own not being an exception (see pic).

I've 40 years experience in soldering, but this is a very special case (up to 7GHz here!), so I kindly request the group for help in applying the best possible fixing procedure.
I thought I'd clean the part with IPA and then wipe it over very lightly with a multi-fiber rag (the one I regularly use to clean my glasses);
then proceed with soldering.
Is it advisable to use some flux here??
Is it advisable to put a very thin copper strand between the parts?
How to avoid the solder from spreading too much along the traces?
.. and so on..


Thank in advance.
Emanuele.

[nctnico]

I'd solder a thin strand on it. Having done such a repair myself before I strongly suggest to:
1) put kapton / polyimide tape on the areas that you don't want to have solder on. The gold plated traces transport tin very easely
2) use hot air from the underside to pre-heat (evenly !) and solder very carefully because the traces lift up very quickly
3) take your time; it is a repair that is easy to mess up

Use cotton swabs only to clean flux residu. Stay away with anything that can build up static electricity (like fiber cloth).

[SilverSolder]

+1 for thin strand.  I would unravel some desoldering wick and use one or two strands of that.

No lead free solder within 100 miles of this project!

Using a vinyl cutting machine (or an X-acto knife and a cutting board) you can make a little mask that leaves only the part of the gold traces exposed that you want to solder, with a slight clearance (think: crude solder mask).  This is to prevent solder going "too far" but also to help contain the flux to the traces only, instead of spreading to unwanted areas.

Conservative tip heat setting, enough to melt the solder and not much more.  Pre-heat with conservatively hot air.

Paste flux, minimal quantity (speck on end of a toothpick) to avoid it spreading/seeping to somewhere it's not wanted, as cleaning this module may be a delicate affair.

Tin the traces, then add the (lightly fluxed) thin strand(s) without adding more solder.

[egirland]

SUCCESS!
See pics for the procedure steps. I didn't put any strand (too difficult).
Final result is not as clean as I would have liked to be but the S.A. is working again.
I should now check it for flatness on the entire frequency range but I do not have any suitable sig-gen... (I only have an HP8640B-no doubler..).
May be I can perform some test using the internal tracking generator (any advise?).
For the frequency ranges I usually work (< 0,5GHz) I believe this can be considered a success.

Thank you all for help and advises.

Until next problem... ciao!
Emanuele.

[egirland]

no no no!!
Correction: it was not a crack!
It was the joint between the input LPF and the sampler: they are two distinct components (THP118 and THP202) joint together by a solder joint.
But what I have found inside when I first took the RF module apart was what you can see in the picture in my first post: no solder between the traces!!
When I made the fix I've found a solder drop inside the unit... I though it was a residue of my fix. It was not: it was the original joint that simply detached from the traces.
Oh my God!
Can anybody confirm my assumption? Has anybody access to a unit and tell me what the original setup really is?

Thank you.
Emanuele

[TurboTom]

Congratulations on your success! I'm not sure if I would have been brave enough to do some soldering inside this microwave module...

I'ld probably have tried -- after some really decent cleaning of the gold traces -- to just cold weld a tiny foil of indium to the surfaces. This metal sticks to (almost) everything and especially to precious metals. For example indium seals are used for (almost-permanent) sealing of vacuum appliances since it bonds to other metal surfaces so well.

But anyway, since it apparently had been soldered before and the joint just came lose, your solution is perfectly adequate! Enjoy your repaired spectrum analyzer!

Cheers,
Thomas

[serg-el]

Here is a crack in the ceramic.  These tracks need to be restored!

[egirland]

Ups! You're right Sir!
The unit is currently working, but now I expect it to fail again soon..

You say:"These tracks need to be restored!"
HOW?!
Indium, as TurboTom suggested?