I have an old accelerator board used in the PDS slot of an Apple Macintosh SE/30. The board was manufactured around 1994. I want to replace the electrolytic capacitors but I'm not sure how to decode the numbers printed on the caps. Here's a photo:
I could take guess and assume that 22 means 22uF, but what about the voltage spec? Would 6A mean 6.3V? Do any of you know? There's no logo mark on the cap that I can see to indicate manufacturer.
Your first diagram doesn't seem to apply directly to my cap, but your second diagram does. According to your second diagram, my cap is what I guess it might be: 22uF and 6.3V. Thanks!
While I could just swap out the caps for identical SMD aluminum electrolytics, those too could leak some years down the line. Ceramic caps are low ESR but (other than NP0) they drop in capacitance with applied bias voltage. I am thinking about tantalum replacements, but I would need to spec the voltage high enough they wouldn't be a fire risk (trigged by over-voltage). I assume this circuit is 5V. Would a 10V tantalum SMD cap be safe, in your opinion? I could go higher in the voltage spec with tantalum caps but they would be physically too large to fit. I'd love to hear your thoughts.
10 V tantalum should be fine - as long as it is of good quality.
Apparently, there were (large) batches of defective tantalums manufactured some years ago which gave the technology a bad name. So you just need to get good ones!
When a tantalum capacitor fails it can fail short circuit. This is nothing to do with a bad batch. This is why if you want anything UL rated then you can not use a tantalum capacitor in a power supply situation as UL do consider this to be a fire risk.
DVDdoug:
You measure the voltage with a multimeter.
I measured it and confirmed it is without a shadow of a doubt 5V. I also did a continuity check on all the (+) lines attached to the five 22uF 6.3V capacitors, and they all hold up the same 5V rail, scattered throughout the board.
I cannot use a 100V or even a 50V cap in this circuit because it would be physically too large for my existing pads. Measuring pad outer edge to pad outer edge, my pads are 6.5mm apart. So any SMD component with pads that extend beyond 6.5mm would be too large to solder onto the board.
I've also considered ceramic capacitors, but the problem with those is the capacitance drops with applied bias voltage. Many ceramic cap datasheets don't even give me a curve to see the capacitance loss with applied bias voltage, making an accurate choice more challenging. So you have to get a ceramic cap with a rated voltage much higher than your application circuit voltage to ensure the rated capacitance is always the capacitance you get. Another potential problem is ESR. The existing electrolytic 22uF 6.3V caps were likely at least 3 ohms when new, so swapping them out for ceramic caps that have an 8m-ohm ESR might cause problems, I'm not sure.
What I do know is that on my other vintage Mac boards which have solid tantalum caps on the 5V rails, they are rated at 16V and are still going strong. The only issue with those is physical size at the 16V rating. They also start to get expensive at 22uF and 16V. And they fail shorted too.
I'm actually thinking about using 10V Niobium Oxide 22uF replacements. They are less expensive than tantalums, won't burn, and aren't too much more expensive (at Mouser) than electrolytics in the small quantities I need. ESR is 700m-ohm too, which is low but not too low. The only caveat is the body size is very tiny, so I need to remove an existing cap and see if the pads beneath are close enough. The datasheet says the reliability level is "Reliability level: 0.5%/1000 hours at 85°C." Not sure if that is good or bad.
ESR (equivalent series resistance) is the value of the free resistor you get in series with the capacitor you buy. As a ripple filter, you want the lowest ESR you can get in the size allotted for the least cost. The higher the ESR, the worse the capacitor will perform in this application.
There is no need to go exotic with replacements. Stick with quality electrolytics and be done with it. That said, if the board is working, you’re just wasting time and money replacing perfectly fine parts. So the parts are 24 years old. Whoop dee doo.
As mentioned in my previous post, the five 22uF 6.3V SMD aluminum electrolytic capacitors are all decoupling capacitors which merely prevent voltage dips on the 5V rail. As such, "ripple" and "low ESR" are not concerns. Even so, if the ESR is too low, there would be a larger inrush current into those capacitors when the power is first switched on, and whether that is good, bad or insignificant is something that only time and testing would determine.
Leaving very old electrolytic capacitors alone simply because your electronic device appears to work well is a recipe for disaster. That is precisely why a lot of vintage computing equipment is destroyed. The user is unaware that leaking capacitors can eventually fail, and when that happens, it is usually the case that a lot of electrolyte has leaked out over a very long time, such that some PCB traces are eaten through and the leaked fluid even gets in between tiny SMD IC pins and causes problems with them. Cleaning up that gunk is a time consuming mess, and can be expensive depending on what damage it caused. Therefore, even if a very old machine appears to be working well, it's often prudent to change the capacitors before the worst case scenario happens. And when considering replacements, all flavors of caps should be considered. This is precisely what I am doing.
As mentioned in my previous post, the five 22uF 6.3V SMD aluminum electrolytic capacitors are all decoupling capacitors which merely prevent voltage dips on the 5V rail. As such, "ripple" and "low ESR" are not concerns. Even so, if the ESR is too low, there would be a larger inrush current into those capacitors when the power is first switched on, and whether that is good, bad or insignificant is something that only time and testing would determine.
WattsThat? Even Grumpy_Mike isn't quite that grumpy.
Anyway, my thanks to "Paul__B" for the lower picture in his earlier post which matches up with my stock capacitors quite nicely. And thanks to not-so-"Grumpy_Mike" for emphasizing that tantalums often fail shorted. And again, that's why I've been pondering Niobium Oxide caps which don't fail shorted and have only a 20% voltage derated versus the 50% of tantalum.
I do appreciate the replies from everyone. Even the grumpy ones! Best wishes!
If you don't have a meter and cost is a consideration, "A cheap meter is better than no meter."