Recently, I have seen the same problem on several 32-095 and 32-096 X-ray source controls which are used on older Physical Electronics PHI X-ray photo electron spectroscopy systems.
The issue is that C9, a 680 uF electrolytic capacitor blows out and the electrolytic material leaks out on the board. Left unattended, the electrolytic etches and oxidizes the traces on the board.
If you have an older PHI XPS system that uses a 32-095 or 32-095 X-ray source control you should pull if out of the rack, remove the cover and inspect the board immediately.
If corrosion is present, then remove the board and remove C9. Note the polarity of C9 as the + indicator on the board may be etched away. Then, carefully clean the corrosion from the board as best as you can. If in the shop I use some Alconox and let it sit on the board for a while, then rinse with DI water and let the board dry overnight. In the field I have used isopropanol or methanol and cotton swabs. Note that if the traces are corroded badly then they may come off the board as you clean it. If so, you will need to use some fine copper wire to rebuild the traces.
Once the board is clean and dry, replace C9 with a new one. I will dig into this issue some more and try to determine why this problem occurs so often and come up with a permanent solution. In the meantime, I would recommend that the C9 capacitor be replaced every 5 years.
The pictures below show where C9 is located on the control board and what the corrosion looks like.
Thanks Jonathan! I looked at the schematic and this capacitor is not used as part of a rectifier circuit. It is used as part of a TTL / transistor circuit that turns on a Crydom solid state relay. So I think that it is related more to temperature and age.
Randy, without the schematic in front of me I cannot tell you with certainty why these capacitors fail so often, but I can intelligently speculate. A good quality modern electrolytic should last beyond five years – closer to fifty than to five. If it isn’t then the part is either being used outside it’s temperature range or it is being electrically overstressed.
Some aluminum electrolytics get unhappy above 45C though most decent ones thrive up to 85C or even 105C if you go for premium capacitors. If the capacitor tends to be in an environment almost as warm as it’s maximum rating it is a good ideas to use a higher voltage part (in other words derating for temperature).
Electrical overstress can be related to voltage which is the one we always think of first. This can mean excessive voltage, or it can mean pulses of reverse polarity on a capacitor that is not intended for such use. But electrical overstress can also mean excessive current, for example using a part intended as an AC power supply first filter with a full wave rectifier in a circuit with a half wave rectifier killed many capacitors in cheap consumer electronics back in the vacuum tube era. Another classic example is using a part intended for use with a 60 hz circuit in a 400 hz circuit. But excessive ripple can also be caused by the load side of the circuit as it often is in cheap LED light bulbs. The solution is to substitute a capacitor that is rated for high ripple current or as having “low ESR”. These used to be referred to as “audio output capacitors” or as “computer grade” capacitors. If this capacitor is in fact feeding a high ripple load such as a switching power supply then you can often make its life easier by placing a lower capacitance non electrolytic in parallel such as a ceramic or plastic film capacitor with the required voltage rating but 5-10% of the capacitance rating.