Seen any TV Repair Service trucks on your street lately? No? Have you ever taken a piece of electronics gear to a repair shop? Never? Really?? It's not that electronic equipment doesn't fail - it's just that the repair bill would likely exceed the cost of a replacement product. No one knows how much of the electronic trash in our landfills was put there because of technical obsolescence and how much was tossed before the end of its useful life because it failed.
I've already expressed my view about how failures due to outside surge events (ESD, induced lightning surges or incidental power line contact) could be drastically reduced with a "Surge Star" program. Today I repaired a stereo amplifier that had suffered a component failure. The amplifier was less than 8 years old and had seen only light usage. What upsets me is that, as a circuit protection designer, I was powerless to prevent this one...
The component that failed was an electrolytic capacitor. It was in a power supply circuit as a filter capacitor. The capacitor itself was 100uF and was rated at 25V. In the circuit it was in parallel with a 15V zener diode. That means that (unless the zener diode failed - which it didn't) the capacitor would never see more than 60% of its rated voltage in the forward direction and less than 1V in the reverse direction. As a circuit protection designer, I couldn't have done it better. There is no way this capacitor failed due to external circumstances!
In the past few years, there has been a quiet revolution in most AC powered equipment. This discussion will exclude incandescent light bulbs (and they're getting harder to find...), AC motors (which are losing out to DC motors - especially in appliances) and resistance heaters (is there a less green thing to do with electricity than just converting it to heat?). Just about everything else that plugs into the AC grid has a switched mode power supply (SMPS) as its first functional block.
In an SMPS, the line voltage is rectified directly - no power transformer is used to step the voltage down. You may have noticed "wall wart" power supplies used as USB chargers have gotten very tiny and lightweight. These are the most obvious examples of how losing the power transformer can make power supplies much smaller and lighter.
Once rectified, the raw line voltage is then filtered by electrolytic capacitors. Any sophomore EE student will tell you that for 120VAC (RMS) inputs, this results in a 170VDC output. This is then put through a DC/DC converter to achieve the needed output voltage. But since most equipment today is rated for up to 240VAC input, 340VDC is considered the nominal voltage of the DC supply. To allow for line input voltage variation, most designers will specify 400V rated electrolytic capacitors. Conservative (high reliability) designs will call for 450V rated capacitors.
To protect the rectifier diodes and the electrolytic capacitors, MOV (Metal Oxide Varistor) devices are almost always found on the AC input to clamp induced lightning and power surges. (When an AC motor shuts off, it can kick 6kV surges back onto the AC power line!)
But in my experience, the electrolytic capacitors are still the weakest link in the SMPS design. I personally have tossed several LCD monitors and a LCD TV due to capacitor failure. LED lighting designers are well aware that the LED's in their designs will outlast the capacitors in the driver circuits.
Electrolytic capacitors are big and expensive so they are an obvious target for cost and space cutting measures in the design cycle - squeezing design margins to the bare minimum. There are also a lot of inferior capacitors sold out there that probably don't meet their own ratings - I think I just pulled one out of this amplifier.
The bottom line here is that circuit protection can't prevent every failure. Sometimes you just have to bite the bullet and either put enough margin in your design to allow for poor performance of cheap parts or pay up front for some real quality components. Just don't blame me...
Kelly Casey is VP of Engineering for FM Technical Consulting, and holds a Bachelor of Science Degree in Electrical Engineering from the University of Nebraska, as well as a Master of Science Degree in Electrical Engineering from the Georgia Institute of Technology. Previously, Mr. Casey has held various roles at Bourns, Littelfuse, and Teccor Electronics.