Hi, I have used 40V 68 000uf caps on 43.7V rails for a 14 years. It was my daily background amp so it had around 20-35hours a week or total 25 000+ working hours. it was class A amp with lot of heat inside. I thought it was time to recap and maybe replace them with 63V rating. Measured them 8psc in series and capasistance were within spec (for a sngle cap) Epcos sikorel caps.
I had also accident when i wrongly soldered in 25V 220uf Nichicon KZ instead of 100V 47uf (same cap size) in 66V suply . I did noticed this AFTER 2 days (around 2-3 working hours) when after adjustments i decided to assemble amplifier completely.
having gone trought this all I can do is to recomend take manufacturers specified ratings with grain of salt. this definately looks like a scam to me (forcing customer to buy higher voltage cap=more expensive, by spreading wrong information)
I had also accident when i wrongly soldered in 25V 220uf Nichicon KZ instead of 100V 47uf (same cap size) in 66V suply . I did noticed this AFTER 2 days (around 2-3 working hours) when after adjustments i decided to assemble amplifier completely.
having gone trought this all I can do is to recomend take manufacturers specified ratings with grain of salt. this definately looks like a scam to me (forcing customer to buy higher voltage cap=more expensive, by spreading wrong information)
I think you are being paranoid.
43.7 V is not that far above 40 V, so taking into account production spread, it is not that strange that a 40 V-rated capacitor survives 43.7 V for a long time if you are lucky.
25 V and 66 V is a big difference, but you don't know how long the capacitors would have lasted at 66 V.
43.7 V is not that far above 40 V, so taking into account production spread, it is not that strange that a 40 V-rated capacitor survives 43.7 V for a long time if you are lucky.
25 V and 66 V is a big difference, but you don't know how long the capacitors would have lasted at 66 V.
Caps like that are rated in "working voltage", meaning a 40v cap will be happy running the full 40v forever. They also have a surge rating higher than the working voltage. So your 40v cap might have a 55v surge rating. SO the cap isn't going to blow right up a pinch over 40v. VOltage ratings on caps are not like a cliff to fall off. It is not like x volts is OK and X+1 volts will blow it up. What does happen at excess voltage is what MArcel just suggested. Life expectancy. Instead of lasting 100,000 hours, maybe your 100v over rating cap might last 10 hours. Your 43v over 40v is less than a 10% variance. Not really a concern. I would not leave 25v caps in a 66v supply, but I would expect them to work for a while.
Getting away with something does not make it a good idea.
Conspiracy theories really have no place here. Engineers are not stupid. Circuits are designed and tested thoroughly. A company just cannot make 100,000 VCRs and have them crap out four months after purchase. And when products do fail prematurely their analysis determines what parts fail. And if it was a cap, customers who bought caps by the million will have very serious discussions with the cap maker over his specs.
Getting away with something does not make it a good idea.
Conspiracy theories really have no place here. Engineers are not stupid. Circuits are designed and tested thoroughly. A company just cannot make 100,000 VCRs and have them crap out four months after purchase. And when products do fail prematurely their analysis determines what parts fail. And if it was a cap, customers who bought caps by the million will have very serious discussions with the cap maker over his specs.
I have decided to share my findings on this nice early morning with a tea and slight troll. What is really funny -how SELECTIVE we are. any engineer would say any 50v 100uf cap over 0.5usd/psc is scam and consipracy theory to take more money from stupid audiophiles. but when oposite scam theory is pointed out(of course i was joking).. 🙂
more on the seriuos note-14 years on heated class a a map with overvoltage WITHOUT loosing capacity is impressive . some PS caps in ML amps fails within 5years and it didnt used even up to 90%
I have also witnessed Toshiba transistors over engineering- they worked with 2x max specified voltage..
more on the seriuos note-14 years on heated class a a map with overvoltage WITHOUT loosing capacity is impressive . some PS caps in ML amps fails within 5years and it didnt used even up to 90%
I have also witnessed Toshiba transistors over engineering- they worked with 2x max specified voltage..
I don't know if this applies to electrolytic capacitors, but when you read up on time-dependent dielectric breakdown in MOS gate oxides or on high-voltage cable insulation, you see that in both cases, the reliability people mainly worry about the long-term effects of imperfections in the insulation: crystal defects in gate oxides or voids in high-voltage insulation. The voltage that can be handled without imperfections is often substantially higher than with imperfections, but the rated voltages have to take into account that there could be imperfections present.
Absolutely not the case - the bottom line literally is that that derating components unnecessarily costs the manufacturer money and reduces profit margin, so this is no incentive to do this.
However if even a tiny fraction of a percent of the products you sell burst/explode in circuit within their rated voltage you, as a reputable manufacturer, are in deep trouble, because buyers will simply blacklist you.
Add to this aging effects and manufacturing spread (which can be quite large) and you'll see than the vast majority of components sold will be somewhat better than the worst-case ratings (should be fairly obvious!). Its the 0.001% that might blow up in service that risk putting the manufacturer out of business, so quality control is used to prevent this and ensure many sigmas of reliability. https://news.mit.edu/2012/explained-sigma-0209
In my experience EPCOS caps are very well made, and the "go to" brand here in Argentina when you want high quality-to-cost ratio, since they are made in neighbouring Brazil (in a formerly Siemens plant) and both Countries being Mercosur free trade members, they have a huge price and availability advantage relative to others ... what´s not to like?
I measure them, as with all parts I commercially use, and they consistently stand about 10% more of rated voltage, for as long as you wish.
Not suggesting they are used that way on purpose, but meaning that gives me extra peace o mind.
So what you measured matches 100% what I have found here, my EPCOS powered amps last for decades and I do not get problems from them.
I also use lesser brands such as TREC or SAMWHA but there a 50V rated one gets no more than 40V DC.
Meaning: you can use high quality parts for real peace of mind, or average "commercial quality" but then derate them.
I measure them, as with all parts I commercially use, and they consistently stand about 10% more of rated voltage, for as long as you wish.
Not suggesting they are used that way on purpose, but meaning that gives me extra peace o mind.
So what you measured matches 100% what I have found here, my EPCOS powered amps last for decades and I do not get problems from them.
I also use lesser brands such as TREC or SAMWHA but there a 50V rated one gets no more than 40V DC.
Meaning: you can use high quality parts for real peace of mind, or average "commercial quality" but then derate them.
over time the electrolitic caps form and adjust their oxide thickness, when a larger voltage is supplied. The result is a slightly smaller capacitance, but usually these caps will stand their projected life.
these were some actual tests we did almost 50 years ago, just fresh at university
just googled some stuff:
https://www.kendeil.com/Portals/0/Tech-Academy/BUILDING_AN_ELECTROLYTIC_CAPACITOR.pdf
just googled some stuff:
https://www.kendeil.com/Portals/0/Tech-Academy/BUILDING_AN_ELECTROLYTIC_CAPACITOR.pdf