A lot of manufacturers use caps that are undersized like a 10 volt (wvdc) cap where a 16 volt one would likely last years longer.
Voltage rating is not the major factor in electrolytic failure. From Cornell Dubilier:
Aluminum electrolytic capacitors made with formation voltages at least 35% higher than rated voltage and with rated temperatures of 85 ºC or higher, don’t require much voltage derating. In applications operating at less than 45 ºC no derating is neededHeat is the major contributor to electrolytic failure. That not only means ambient temperature, but also internal heating from ripple current. Again, from Cornell Dubilier:
Some makers appear to calculate operating life saying that expected lifetime doubles with less that a 10 °C drop in temperature. The expectation that life doubles for each 10 °C reduction is well demonstrated by life tests and is consistent with the Arrhenius equation of chemical activation and the activation energy for aluminum electrolytic capacitors.(They use a more complex equation themselves that they believe is more accurate.)
The electrolytics in both of my Ocelots failed and had to be replaced after being in service about 15 years. The ones that were in there measured essentially an open circuit. I located the replacements further away from the regulator heat sink.
Both of my Tektronix 7603 oscilloscopes went down with failed electrolytics in the power supply. I'm sure that was well engineered, but heat and 40+ years of service took its toll. The replacements were tiny compared with what was in there.
Electolytics can even deteriorate on the shelf with an increase in leakage current.
While the XTB-IIR uses a 85C capacitor because of low ripple current and little temperature rise in the large case, I went to a 105C capacitor in the XTB-523 and new XTBR due to the tight case and location near the power transformer.
Jeff
