Capacitors

[KhanTyranitar]

I know this is a bit of a subjective question, but for my project I'm going to need a bunch of capacitors. I own a C128 and it was repaired by a previous owner due to damaged traces from leaking capacitors, and I've seen videos of others being repaired for the same reasons. I've also owned a truck which had a known history of the same problem when the ECM would fail due to capacitors leaking in the board.

So in a nutshell, to you electronics gurus, is their a brand recommendation for capacitors that won't fail over time and ruin my board?

[Arlet]

All electrolytic capacitors fail, but some fail faster than others. Most manufacturers will specify a minimum lifetime @ certain temperature.

If you go to a site like digikey.com, and search for aluminium electrolytic, it will show you a "lifetime @ temp" column that you can use for selection or sorting.

[KhanTyranitar]

Ok, I did not know I could filter by lifespan. Thank you.

[BigDumbDinosaur]

KhanTyranitar wrote:

So in a nutshell, to you electronics gurus, is their a brand recommendation for capacitors that won't fail over time and ruin my board?

For what it's worth, I use Panasonic's low ESR electrolytics, such as their part number EEU-FR1A101B, which is a 100µF 10 volt part.

As Arlet hints, do pay attention to life vs. operating conditions, as well as the operating voltage vs. rated voltage. Factors such as ripple riding on the DC source can shorten capacitor life in some cases.

Also, when I design electrolytics into something I try to use capacitors whose voltage rating is no more that 50 percent higher than the maximum continuously expected operating voltage. For example, in five volt circuits I typically specify 6.3 volt capacitors if available in the required capacity. In some cases, continuously operating an electrolytic at a voltage that is substantially lower than its maximum rating may cause it to develop less capacity.

[GARTHWILSON]

See this lecture from an engineer at Kemet, the capacitor manufacturer: https://www.youtube.com/watch?v=ZAbOHFYRFGg

I was a repair tech at TEAC's (home audio and semi-professional tape recorders) main western hemisphere location in Montebello, CA in 1982 and '83 and fixed over a thousand units there. The only capacitor failures we had were the electrolytic capacitors whose WVDC rating was close to the voltage that was on them in operation, for example a 25V or 35V capacitor for a 24V application (for the motors). These were always power supply capacitors of at least 1000uF (usually more) where a higher voltage would have taken more space and cost more. When there was a wide margin, like a 10uF 50V capacitor in a 12V application (which was common in the audio circuits), they never failed. Ever.

In our aircraft communications electronics (for which I have been the sole circuit designer since 1992), the only electrolytic capacitor failures we've had were from a particular batch of 220uF 16V 6x11mm capacitors we got in the mid- to late-1990's which had 6V on them in the circuits. Those started going down after 10 years. They definitely were not "dropping like flies," but I would prefer that the failure rate be zero. They would short out inside, meaning they became like a hard-wire connection from one lead to the other. On the rare occasion I get a unit for repair that has those, I just replace them now automatically, even if they have not failed yet after 20 years.

The only failures we've had of monolithic ceramic capacitors were infant mortalities. If they made it through the first year or two, they never went down. IIRC, those failed in the open-circuit mode. (I think it has been quite a few years since we've had any of these failures.)

Quote:

I've also owned a truck which had a known history of the same problem when the ECM would fail due to capacitors leaking in the board.

I'm sure that was from the heat.

[KhanTyranitar]

GARTHWILSON wrote:

See this lecture from an engineer at Kemet, the capacitor manufacturer: https://www.youtube.com/watch?v=ZAbOHFYRFGg

I was a repair tech at TEAC's (home audio and semi-professional tape recorders) main western hemisphere location in Montebello, CA in 1982 and '83 and fixed over a thousand units there. The only capacitor failures we had were the electrolytic capacitors whose WVDC rating was close to the voltage that was on them in operation, for example a 25V or 35V capacitor for a 24V application (for the motors). These were always power supply capacitors of at least 1000uF (usually more) where a higher voltage would have taken more space and cost more. When there was a wide margin, like a 10uF 50V capacitor in a 12V application (which was common in the audio circuits), they never failed. Ever.

In our aircraft communications electronics (for which I have been the sole circuit designer since 1992), the only electrolytic capacitor failures we've had were from a particular batch of 220uF 16V 6x11mm capacitors we got in the mid- to late-1990's which had 6V on them in the circuits. Those started going down after 10 years. They definitely were not "dropping like flies," but I would prefer that the failure rate be zero. They would short out inside, meaning they became like a hard-wire connection from one lead to the other. On the rare occasion I get a unit for repair that has those, I just replace them now automatically, even if they have not failed yet after 20 years.

The only failures we've had of monolithic ceramic capacitors were infant mortalities. If they made it through the first year or two, they never went down. IIRC, those failed in the open-circuit mode. (I think it has been quite a few years since we've had any of these failures.)

Quote:

I've also owned a truck which had a known history of the same problem when the ECM would fail due to capacitors leaking in the board.

I'm sure that was from the heat.

Ok, good to know. I don't mind spending a little more for cushion versus getting something that is barely adequate.

And yes on the truck heat was the biggest factor, however many other ECMs of the same era were not plagued by this problem. In this vehicle the computers were made by Mitsubishi, and only models with capacitors made by Rubicon were prone to this failure. ECMs that used other capacitors rarely failed.

In my application the hisghest voltage they will be used on is 12V DC so I'm guessing if it make sure the ratings are significantly higher then I should be ok. Appreciate it.