As the sharp eyed will have noted, this repair is not yet complete
I returned the unit to correct operation with the work on the DC-DC module, but that is not the whole story. As I stated at the beginning of this thread, the capacitors on the microprocessor and another digital PCB also looked suspect due to corrosion on the leads. Experience tells me that this is likely caused by electrolyte leakage rather than atmospheric effects (no other surface corrosion was visible on other metal parts or solder joints).
Today I had a few minutes to tack the two outstanding boards. I removed all of the SMT electrolytic capacitors using the twist method and all went well with no adverse effects on the PCB. As expected, several of the capacitors had started to lead around their lead seals. The fluid was starting to pool under them on the PCB but no obvious corrosion had occurred. A lucky escape fot the PCB and testament to the laquer coating quality.
The cleaning and preparation approach for these two PCB's is a little different to that of the DC-DC module as the contamination is far less severe and a full clean process is not justified. The full clean process comes with risk and so should be carried out employing common sense rather than automatically, as I will explain.
The areas of contamination on the two PCB's are limited to directly under the capacitors and no fluid spread could be found further afield. The cleaning process used is as follows:
1. Electrolytic capacitors removed
2. Lead remnants removed with fluid flux and a soldering iron.
3. A good layer of fresh solder applied to ll capacitor pads
4. New solder removed from capacitor pads
5. PCB area around capacitor pads hand cleaned with a Q-Tip and IPA.
6. PCB placed into orbital mixing bath filled with IPA to wash away any contamination on the PCB. Duration 10 minutes for microprocessor PCB and 3 minutes for PCB fitted with Lithium cell.
7. PCB allowed to drip off excess IPA
8. PCB placed into orbital mixing bath filled with De-ionised water. Duration 10 minutes for microprocessor PCB and 3 minutes for PCB fitted with Lithium cell.
9. PCB dries in warm air drying cupboard.
Some readers may wonder why I did not use the Ultrasonic Cleaning Bath this time. Ultrasonic cleaning is justified where heavy contamination exists and where components are relatively robust and without special needs. Using such cleaning technology casually and without due consideration invites problems. The microprocessor PCB contains many sensitive processing components that may, or may not respond well to Ultrasonic cleaning. If they do not respond well, the SA could be effectively trashed by unnecessary ultrasonic cleaning. The other digital PCB contains a AA size Lithium cell. Immersing such in cleaning fluid heated to 40C is not particularly unsafe, but it could lead to unpredictable conduction and electrolytic action between the battery and areas around it. Loss of the supported memory data is also a possibility. The conductivity and purity of the cleaning solution is not known so it is best to play safe with batteries that cannot be removed from a PCB without vital data loss. IPA and De-ionised water present little risk to the battery or memory chip data contents. IPA and clean water are appropriate cleaning mediums for this limited electrolyte contamination scenario.
Please also note that I did not use solder wick on the PCB containing the Lithium battery in order to avoid the risk of shorting out the battery or the power rail that it supports. Common sense stuff really.
Off of the cleaning discussion..... take a look at the number of bodge wires and components on the PCB that contains the Lithium battery. I am both disappointed and amazed to see such a PCB installed in a production issue Advantest Spectrum Analyser originally costing many thousands of Pounds. IIRC they were around £10K
To my eye, the PCB contains an awful amount of botchery that should have been addressed by a new revision of PCB. There are even IC's glued on top of others and bodge wires used to connect to them. Floating resistors on the end of bodge wires.... oh dear
Whilst the soldering and wires are well executed, there are too many of them and such would not normally pass testing on a vibration table. The mastic application on the bodges is also poor. Pretty nasty all round really........but it does function so I will not stress about such. It was just shocking to see this from Advantest, a normally high quality equipment manufacturer.
Note: most of these pictures were taken before the PCB's were cleaned in the orbital mixing bath.
Fraser