Could some electronics guru please shed some light on this ? How is the voltage rating on a capacitor calculated, at design time. For any capacitor, the capacitance is the ratio of charge stored to maximum voltage across the terminals. By measuring the the discharge current and time for a fully charged capacitor to discharge, the total charge stored can be calculated. But that is possible after the capacitor is ready for use. Any hints, suggestions would be of help. Thanks in advance.
Voltage rating used to be determined by raising the voltage across the capa citor until the capacitor failed - which tended to be catastrophic, since t here's usually enough energy stored in a fully charge capacitor to establis h a low leakage path for all the energy to discharge through.
The published voltage rating was enough below the observed failure point to make failure improbable.
There's one exception that I ran into - if you put more than a couple of hu ndred volts across film/evaporated metal capacitors, you got corona dischar ge inside air-bubbles in the capacitor, which chewed out the evaporated met al, so the capacitance would be a lot lower after a couple of years of serv ice.
Manufacturers mostly got wise to this and film capacitors that were going t o see continuous high voltage AC were made as series connections of a numb er of bigger capacitors, so the voltage drop across any one connection was never big enough to create a corona discharge. There were fly-by-night oper ations that didn't bother - they figured that they'd have moved on by the t ime their customers noticed.
to >see continuous high voltage AC were made as series connections of a n umber of >bigger capacitors, so the voltage drop across any one connection was never big >enough to create a corona discharge. There were fly-by-night operations that >didn't bother - they figured that they'd have moved on by the time their >customers noticed. "
Interesting. So, what about the tuning cap across the LOPT and the yoke in a TV ? And then the yoke return cap which saw much less AC voltage, but it saw current.
Over here the circuit resonated at about 70 KHz and the cap values ran betw een like 0.0082 and 0.01 uF. This was with a half sine wave driven at 15 KH z. The amplitude was around 1 KV. Were those caps "cellular" ? I mean were they stacked caps inside, maybe in series or whatever ?
The yoke coupling cap was like only 200 volts but more like 0.47 uF but had significant current going through it. Same thing as ripple current of cour se, and they did go bad some.
But the point is a regular electrolytic would not last a second in that app lication.
I know about different capacitors, their strengths, weakness and failure mo des. I wouldn't mind knowing a little more about how the different types ar e made. But it is not a really high priority.
I find myself uncertain whether you're talking about the design of capacitors, or the choice of voltage rating in the design of something else.
Capacitor designers would know the properties of the dielectric material they intend to use, and the desired rating would feed into the thickness of dielectric required.
ng to >see continuous high voltage AC were made as series connections of a number of >bigger capacitors, so the voltage drop across any one connectio n was never big >enough to create a corona discharge. There were fly-by-nig ht operations that >didn't bother - they figured that they'd have moved on by the time their >customers noticed. "
n a TV ? And then the yoke return cap which saw much less AC voltage, but i t saw current.
tween like 0.0082 and 0.01 uF. This was with a half sine wave driven at 15 KHz. The amplitude was around 1 KV. Were those caps "cellular" ? I mean wer e they stacked caps inside, maybe in series or whatever ?
How would I know? I got my earful in 1970 about capacitors that went into f luorescent lights to compensate for the power factor of the inductor in the starter.
ad significant current going through it. Same thing as ripple current of co urse, and they did go bad some.
If they were film and foil, the foil is thick enough that corona isn't goin g to eat it quickly enough to matter.
pplication.
modes. I wouldn't mind knowing a little more about how the different types are made. But it is not a really high priority.
Charge soak in the various different sorts of plastics films used in film c apacitors is kind of interesting. Polystrene is good, but doesn't give you much capacitance per unit volume. PTFE/Teflon is good too, but hideously ex pensive.
Polypropylene is the dielectric of choice, polycarbonate is cheaper, but no t as good, and polysester is marginally worse that polycarbonate, but cheap er.
The guy at badcaps.net has a thriving business that sells replacement capacitors. He echoes what you say, but he dates a stolen formula to the turn of the century. (The OP may find it worthwhile to take a look at badcaps.net.)
There's a story about a hard drive factory that made tens of thousands of defective hard drives with horrible MTBFs. So the plant manager put on his poker face and dumped them on the other divisions of the big company that he worked for.
The "dump it on the next guy" policy seems endemic in the computer industry. Just "throw it over the wall" and let it be the next guy's problem.
They supposedly crank up the clock speed into a microprocessor until its MTBF becomes unacceptable. Then they back it off and rate the microprocessor at the reduced speed.
For solid dielectric capacitors, it's a measurement or calculation of the dielectric strength of the capacitor at a given operating temperature. Note that this temperature includes internal heating caused by ripple current and ESR (equivalent series resistance).
For liquid electrolytics, life is a bit more complexicated. The capacitors have an ESR which will cause heating. Besides overheating the capacitor, the heating can also break down the electrolyte which reduces the useful life of the capacitor. So, we have derating voltage curves and lifetime calculator for electrolytics. "Aluminum Electrolytic Capacitor Application Guide" Various derating curves: Capacitor life calculator: The lifetime calculator makes an interesting conspiracy theory. If I wanted to build the absolutely lowest cost electronic product, I would use the lowest voltage electrolytic (and other) capacitors possible for the specific circuit design. If I knew the load life, operating temperature, capacitor current, dissipation factor, etc, I could calculate how long a particular capacitor would last before it self destructs. If I wanted to offer a 1 year product warranty, I might select capacitors with a projected 5 year capacitor life. If you see a PCB with a wide variety of odd electrolytic voltage ratings, this is probably what is happening. Done properly, all the capacitors will fail at approximately the same time, which I've seen happen.
The life of a capacitor is nasty, brutish and short (Thomas Hobbs):
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
More like 2002. I read the original notice on the Nichicon web site which indicated that their magic electrolyte formula had been stolen by a former employee and sold to a competitor without the inclusion of a "stabilizer". I didn't think of saving the document and haven't been able to find it on archive.org. Personally, I think the whole thing was a coverup to explain capacitor failures due to an experiment in using water based electrolytes. Details:
Of course, it might not always be an electrolyte problem:
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
There are also dielectrics whose relative permitivity drops with bias voltage. Most ceramics do this (not sure about C0G). While manufacturers appears to have gotten pretty sloppy about how much capacitance loss they're willing to tolerate and somehow be in spec, it's a serious concern for circuit designers.
I've put >500V on some low-voltage caps without any catastrophic failure (carefully, then throw away after the test as I couldn't vouch for reliability afterwards).
On Mon, 23 May 2016 03:08:12 -0000 (UTC), Frank Miles Gave us:
Caps are rated well BELOW what they are actually capable of taking and this is so that "safety" is maintained in electrical circuit designs. Don't need no holes poked leaking 'lectrons where they should not be leaking, much less catastrophic shorting events, such as those which would occur in a breach on an HV cap. They can read normal value when tested, as a bad HV cap typically does, but in operation they look like a dead short to the circuit and can cause upstream failures as well.
The designer is ALSO supposed to choose a cap with a significantly HIGHER rating than the voltage he or she actually intends to impress upon it.
snipped-for-privacy@gmail.com wrote: >>"Manufacturers mostly got wise to this and film capacitors that were going to >see continuous high voltage AC were made as series connections of a number of >bigger capacitors, so the voltage drop across any one connection was never big >enough to create a corona discharge. There were fly-by-night operations that >didn't bother - they figured that they'd have moved on by the time their >customers noticed. " > > Interesting. So, what about the tuning cap across the LOPT and the yoke in a TV ? And then the yoke return cap which saw much less AC voltage, but it saw current. > > Over here the circuit resonated at about
70 KHz and the cap values ran between like 0.0082 and 0.01 uF. This was with a half sine wave driven at 15 KHz. The amplitude was around 1 KV. Were those caps "cellular" ? I mean were they stacked caps inside, maybe
Yes, they would have been series sections of windings for 1kV AC use in those frequencies, unless it was a ceramic cap or mica or something like that.
Years ago I designed this 28V circuit. At the time I knew nothing of the pitfalls of tantalum caps, so I stuck in
35V tant's (which is what we had in stock for 15V circuits.) and everything was fine till a few months ago. Some one had purchased some cheaper tanatalum (multi-comp... newark "junk") and one of them let go and spewed crud all over the pcb. I pulled 10 of the cheap caps from stock, stuck them on a power supply.
One let go at 16V, then another at 18, and then 20V. The rest survived to 35V, but I pitched the whole lot.
Yeah, ~1 Volt steps on the power supply. (Coarse knob)
It's interesting about the ceramic caps. Is the higher voltage true for the high-K materials, X7R, say? I was just pricing X7R at 50V and they can get speny.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.