I am no longer doing 'tear-downs' but thought some readers may be interested in a repair that I am about to undertake on a very nice portable U4941 Spectrum Analyser from the Advantest Test Equipment Stable. I used a similar model some years ago and it was very useful for portable applications as it can run on a battery pack or external 12V supply.
These analyser's normally sell for over £1000 in the UK, but this one was poorly so I risked £200 on it. I says 'risked' because there is no affordable support for these units in the UK. R&S offer a fixed (very high) price repair service. If something unusual has failed within, it becomes an expensive door stop. I have loads of spares for my R4131D SA's but none for this U series model.
The U4941 arrived and was found to be in excellent condition. It has been well cared for in its life and it came with its official Khaki protective soft case and the User Manual.
There is no service manual openly available for this model but the U3661 is similar and the service manual proves what I already expected. These units are module replacement only with no component level repair supported. No schematics are provided in the service manual. It has just basic faulty module identification routines. Not a whole lot of help really.
I love a challenge so the lack of a set of schematics is of no concern to me. I just reverse engineer what I need to better understand. All part of the fun for me.
Upon receipt of the Advantest Spectrum analyser, I carried out the usual initial checks before opening the casing. With power and an external monitor connected I was able to establish that the boot system self check showed Error 900 which is a lock problem in the Phase Locked Loop. The microprocessor board appeared to be working OK as the front panel controls all worked. This was good news for me as I dislike microprocessor related faults.
During the brief testing phase I detected the smell that I associate with degraded and leaking electrolytic capacitors. It was coming from the main units extraction fan rather than the external power supply block. The smell is not forgotten once experienced. It is a rank sweet fishy smell that can be quite overpowering in extreme cases of electrolytic capacitor leakage. The leaking of capacitors in equipment built in the 1990'sis well known and named the 'Capacitor Plague' as it is so common an can be very destructive. There is a dedicated web site detailing this nasty and all too common failure in modern electronic equipment. It is most prevalent in equipment power supplies, especially IT related equipment where such capacitors are rated close to their operating voltages. The leaking electrolyte is very conductive and corrosive and can cause all manner of havoc on a computer motherboard.
The Wiki page on Capacitor plague is here:
https://en.wikipedia.org/wiki/Capacitor_plagueThe web page dedicated to the problem is here:
http://www.badcaps.net/index.phpAs soon as I had smelt the well known failed capacitor odour I knew I would need to inspect all the capacitors on the various boards within the SA, starting with the Power supply related module. This Advantest SA is designed to be field portable so it has the ability to run from a 12V battery pack or external 12V supply. As such the unit contains a DC to DC converter board that provides all the regulated power supply rails for the unit. This SA requires multiple +12V, -12V, + 8V, -8V and +5V for analogue circuits, and +5V for the digital electronics needs. The Analogue and Digital supplies are kept separate to avoid digital noise cross contamination issues.
On the DC-DC module board there are nine separate isolating TDK DEL series DC-DC converter modules providing the requires power rails. There are three +12V, two +5V, two -12V, a +8V and lastly, a -8V. The use of individual DC-DC converters is a little unusual as normally all the required rails are provided by a multi output single DC-DC converter module. The good news is that a faulty module is easily identified and replaced or repaired. This is why the SA still appears to partially work. The digital power rail is being correctly generate and so the microprocessor is running fine.
Further investigation of the DC-DC module proved that it was the sole source of the awful leaking capacitor smell. Every electrolytic capacitor on all of the TDK DEL mini DC-Dc converter modules was leaking ! They really reeked when brought close to the nose. Some of the larger Brown coloured electrolytic's on the modules PCB also appeared to have leaked electrolyte onto the board. At that point I knew I was in for a complete replacement of all electrolytic capacitors in the DC-DC module. I will also need to clean the PCB's and repair any corrosion that has occurred. There was visible overheating damage on the -8V TDK DEL module.
A test of the complete DC-DC module was possible by supplying it with a 12V lab power supply and activating all of the TDK DC-DC modules via the remote control on/off line. Tests proved that two TDK modules were not producing an output. One +12V and the -8V rails were missing. Both modules have been removed for investigation and repair.
The faulty TDK modules were visually inspected and the -8V module is in a bad way. Its power Mosfet has seriously overheated, partially desoldering itself from the PCB, and cooking the PCB in the process. Exactly what caused this overheating is not known as the TDK module contains over-current detection and automatic shut-down to protect itself. Something has gone wrong n that little PCB. It could be just the leakage of conductive capacitor electrolyte that was soaking the PCB's surface. we shall see when it is cleaned and new capacitors and Mosfet are fitted. The +12V module has no damage evident except electrolyte leakage. It may be in over-current fold-back mode.
I ordered all new capacitors and the required power Mosfets from Radio Spares to repair the two modules. In the mean time I quickly removed the failed capacitors and faulty Mosfet. I also reverse engineered the TDK DC-DC modules so that I better understand their circuit topology. They are actually pretty simple devices but they use an obsolete switching controller so I hope neither have failed on the faulty modules. If so, I will fit an alternative DC-DC module of suitable spec. All capacitors on all TDK modules and the supporting PCB will be replaced with quality parts from Panasonic to ensure high performance and long life.
I will treat the whole DC-DC module PCB and associated TDK DC-DC modules to a good bath in my specialist Electronics compatible Ultrasonic Cleaning Bath. It is designed for delicate work and uses a swept Ultrasonic frequency and 'electronics safe' cleaning solution. It will remove all the nasty remaining leaked electrolyte that I could not get to with IPA and swabs. This will ensure that corrosion of the PCB's does not continue and ensures the long life of the module.
I need to restore the DC-DC power module to full working order before i do any further diagnosis on the SA as the correct DC powers rails are a fundamental requirement for correct operation, as you will appreciate. I can then check for over current situations on the power rails causing fold-back of the associated TDK module.
I inspected the microprocessor module and some of the electrolytic capacitors on that modules PCB have some corrosion around the pins. That is enough for me to replace them as a precaution, even if the corrosion is not leakage related. I have ordered replacements for all those capacitors as well. I shall be busy soldering new capacitors onto the PCB's, that is for sure. Prevention is better than cure however.
Fraser