All things wear out with time and use.  Our motherboards are no exception.  Most often, they die because the electrolyte used in modern electrolytic capacitors dries out and causes the capacitor to lose its ability to store charge.  Whatever the reason, when the mobo bites the dust, we generally use it as a an opportunity to upgrade to something newer and better.  Every once in a while, though, a motherboard design comes along that is such a solid performer that it makes it worthwhile to try to find out exactly what is wrong and why.  The EPoX 8RDA+ is one of those motherboards.  However, it poses some unique challenges.

The Challenge

What, then, is the challenge?  Well, put simply, the challenge is the capacitors themselves.  It would appear that the 8RDA+ is designed to use a particular size of capacitor that is rather difficult to find in the required ratings.  The main problem revolves around the capacitor diameter.  Capacitors from Panasonic, Murata, Rubycon, and virtually every other capacitor manufacturer I could find are larger in diameter than what the 8RDA+ is designed to use.

"But wait!" I hear you say.  "I know people who have replaced their capacitors with parts from those same manufacturers!"  Ah, but did the board work EXACTLY the same as it did before they replaced those parts?  In all the references I have been able to find where people used parts from these manufacturers and found parts that would fit, the board never had quite the level of performance it had previously.  This was generally chalked up to some kind of 'damage' to the board or the CPU when the caps finally blew out. "'NO!" I say!  The problem is because they didn't use the RIGHT capacitors.

So, what's the big deal with capacitors?

Capacitors are rated according to their ability to store a charge, or their capacity.  This capacity, referred to as capacitance, is measured in Farads.  Since a Farad is a very, very LARGE unit of charge, capacitors like we are dealing with are measured in micro-Farads, denoted as µF.  However, not all capacitors are created equal.  Just because a capacitor has a rating of 1,000 µF doesn't mean that it can replace any other capacitor with an equal rating.  Capacitors are also rated according to their operating voltage (typical voltages are 6.3, 10, 16, 25, 50, 63, 100, all the way up to very high voltages in the range of thousands of volts).  In addition, capacitors have temperature ratings for both heat and cold, tolerance ratings (how much the actual value of the capacitor can vary from its nominal, or ideal value), and ESR and ESL (a measure of how the capacitance will change with frequency).  ALL of these things taken together determine how well, and even if, a capacitor will perform in a given role.  The engineer responsible for the design of a given product will generally try to take all of these factors into account when selecting the capacitor to be used in a product in order to ensure that the proper component for the job is used, while also minimizing cost.

So, what are the rules for replacing capacitors?  Well, in general, you want to try to get one that is as close as possible a match as you can get for the one you are replacing.  You can usually (but not always) substitute a higher capacitance part for a lower one (e.g. use a 1200 µF part instead of a 1000 µF part).  You can usually (but not always) substitute a part with wider temperature tolerances for one with lower temperature tolerances (for example, use a 105C rated part in place of an 85C rated part).  You can substitute a part with tighter tolerances for one with looser tolerances (e.g. a part rated at +/- 20% for one rated at +/- 50%).  You can substitute a LOW ESR rated part for one with a higher ESR rating.  I have never seen an instance where doing either of these was a problem.

So, more is better, right?  Well, not always.  For example, some voltage regulator designs (generally older linear designs) will begin to oscillate if their filter capacitors get too big or too small.  In addition, their ability to respond to input transients may be impared.  Further, you should NEVER substitute a higher voltage part for a lower voltage one (at least, not when substituing electrolytic parts, which is what we are talking about here).  Electrolytic capacitors require a certain minimum operating voltage or their internal structure will begin to break down.

There.  Now I've just imparted a year's worth of learning to you in the space of one page.  LOL.

All of these factors come together to dictate the size and shape of the capacitor.  And therein lies the rub; the 8RDA+ uses capacitors that are narrow and tall, while capacitors available today tend to be more of the short and fat variety.