is it VDC or WVDC?

[mkruger]

what is the difference when a capacitor is labeled 450vdc compared to 450wvdc?

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=3834292185&ssPageName=STRK:MEWA:IT

these are advertised as just 450 volts but I just recieved them and they are labeled 450wvdc. They seem very small to be 450volts.

-mike

 

[SSLtech]

WVDC = "Working Volts DC"

Keith

 

[mkruger]

So there is no real difference i need to be worried about?

I have another 450VDC cap at 1500uF and it's physically very large... this cap is only 820uF and 450WVDC but it's significanly smaller. I can put my hand around it. The dimensions are only 35mm x 80mm. Does this seem funny (too small) for these specs?

 

[SSLtech]

Sounds a little small, but not scarily so... the 1500µF cap should be about double the volume of an 800µF cap of the same voltage...

Don't foget that size is a function of working voltage and capacitance... here's a much smaller cap with a much higher working voltage... by virtue of its minute capacitance...

Keith

 

[mkruger]

hmmm... that makes me wonder. the 1500uF cap is more than 3 times the size. and also if you noticed on the auction this 800uF cap is a "tripple capacitor" which makes me wonder why it's not 3 times bigger than the 100uF cap.

do you think it's possible that the 1500uF caps are just bigger because they are older and made with somthing different. they are both electrolytic.
the large cap is the 1500uF 450VDC
the small cap is the "tripple" 820uF 450WVDC

 

[CJ]

I think working voltage came out of the military. This is the voltage that sits on the cap most of the time. Whereas VDC might mean the max voltage that the cap should ever see, so you would want to run the cap below the VDC rating?
The most important thing to worry about is ac vs dc. A cap rated for 400 VDC does not want 400vac because that is an rms reading and the cap will see peak of that, ewhich is 1.4 times higher. They used to put ac and dc ratings on caps, but i haven't seen it in a while. mostly on your coupling caps of course, not pwr supply caps.

 

[mkruger]

i just posted some pictures, any other thoughts before i try and put these into use? we'll see what happens....

-mike

 

[PRR]

> bigger because they are older and made with somthing different{?}

Yes. The screw-terminal caps were standardized long ago, and intended for VERY high reliability with low internal stress, in a market where space and cost were not primary factors.

The new caps are made with better stuff, better quality control, and for markets where space and cost (proportional to size) are big factors. Reliability of recent caps is really better than anybody needs. They will probably out-live you, unless you keep them right at their voltage and temperature rating.

In the old days, there was surge voltage and working voltage. Rectifiers warm up faster than other tubes, and have high voltage drop when loaded, so the cap could see 525V for several seconds after start-up settling to 450V as the other tubes warmed up and started sucking current. With solid-state which does not have such large start-up surges, it is usually best to use the Working voltage as the maximum the cap will ever see. Also vacuum tube rectifiers would not pass huge current during start-up, sand-state rectifiers will. "Surge" ratings have faded away. Some label makers use the "W", some don't, but the main voltage spec is always the Working rating that it will stand in regular operation.

If start-up is a problem, the maker's specs may have a higher "surge" rating in the fine print. It would be best to just never exceed the Working rating. If you have to fudge, 10% overvoltage on an electro will just cause a small leakage and internal heating. If you are talking about the 10 seconds it takes for tubes to warm-up and pull-down the B+, the cap will not get hot enough to matter. You can probably go 15% overvoltage for a few seconds each start if you don't need lifetime reliability. In fact you may be able to go MUCH higher before caps actually pop.

 

[PRR]

http://www.cde.com/catalogs/380LQ.pdf shows a 85 deg C cap series very similar to yours, and a 450WVDC cap has a 500VDC Surge rating. Size is a major selling point.

http://www.cde.com/tech/ElectrolyticLifeMultipliers.pdf is a technical paper about electrolytic cap ratings and lifespan.

 

[mkruger]

nice! size was definatly the selling point for me. I just wanted to make sure this cap wasn't "fake". thanks for all your input. this should work out better then I expected...

-mike

 

[Dale Manquen]

Nobody has mentioned one of the primary rating factors- ripple current. The size of a capacitor determines the amount of heat the capacitor can dissipate. Units intended for high ripple currents, such as power supply applications, must be larger to keep from overheating.

Newer capacitors have significantly less equivalent series resistance, which lowers the temperature rise for a given ripple current.

 

[PRR]

> The size of a capacitor determines the amount of heat the capacitor can dissipate. Units intended for high ripple currents, such as power supply applications, must be larger to keep from overheating.

Quite true.

However it is unusual in audio to be severely pushing the Ripple Current rating. We usually want very clean DC. That forces a large capacitor to get the ripple voltage down. Ripple current is of course high, but not near enough to the typical RC rating of the cap to worry about.

Where you get in deep trouble (implied by the CDE paper cited above) is when you "just want DC", don't care about ripple until it is so bad that your average voltage falls. In welders and motor-drivers, 20%-30% ripple is no big deal. And these apps are high power so they need need huge caps to work at all. The economic tendency is to use a cap "just big enough" to keep the arc hot or motor turning steady. Huge slugs of energy flow in and out of the cap every half-cycle. The paper uses the term "wear-out".

However, the rating for these small ones is similar to the old monster screw-terminal "Computer Grade" caps. Something like 3.4A vs 3.7A for similar capacity and voltage, even though there is a huge difference in surface area.

As you say, ripple current is less an issue if losses are small. So a low-loss cap can take more ripple current with smaller outside area than a high-loss cap. They have really reduced series resistance since I was a boy. And that is when the specs for those screw-terminal "Computer Grade" caps were set: decades ago. As far as I can remember, the standard values and physical sizes have not changed in decades. If you need to freshen-up your 1978 DEC-10, you can buy an exact replacement. But capacitor technology sure has improved. These little snap-cans are lower DCR, lower leakage, similar I(ripple), and appear to have real-world life similar to the old Computer Grade. At least in "easy" work like audio where we love to super-size caps. Perhaps for heavy-duty work like arc-welders, the big old cans fail a little slower. (OTOH, if made with big old technology to ensure exact equivalence, maybe they work a little worse than you could do with snap-cans.)

 

[Dale Manquen]

Two other points worth making:

The ripple current rating is the rms current, not the average current. When dealing with power supplies with silicon diode bridges, the charging current is of short duration and high amplitude, leading to an RMS value that can be several times the average value of the current. The heating effect is directly related to the RMS current. The filter capacitor must be chosen with this in mind.

Also, substituting a silicon diode or two for a tube rectifier can increase the RMS ripple current significantly since the diode has much lower dynamic impedance. The charging current spike narrows up and gets taller, raising the RMS current value and capacitor heating. Inserting a resistor in series with the diode(s) can reduce the peak current and keep the RMS ripple current to about the original level.

 

[PRR]

> an RMS value that can be several times the average value of the current. The heating effect is directly related to the RMS current. The filter capacitor must be chosen with this in mind.

I've never seen it be a problem in "good audio". I'm trying to find a bad example.

Take a cap like Mike's, 600uFd 450V. Specs for a similar cap say ESR is 0.35 ohms, ripple current should be under 2.8A.

What would I do with 600uFd 450V? Sounds like vacuum tube stuff. Since really great amps have been built with 40uFd CLC or CRC filters, either a very large amp (parallel 6550) or trying to avoid the LC or RC and get clean B+ in one step. Or both?

After poking at Duncan PSD a bit, I tried ~450VDC, 1ADC. I assumed the power transformer's AC current rating was twice the DC current I needed, a good rule of thumb for initial design. Since this is not a small transformer, I assumed 10% regulation. PSD said it would have 16 ohms winding resistance for those specs.

I get 10V p-p ripple, 2.4%. Pretty clean, and in line with the motivation to use a 600uFd cap.

Cap current is -1A, +3.4A peaks, 2.5A mean, 1.6A RMS. This is 2/3rds the cap's rating.

Changing cap ESR makes little difference to the I(RMS).

Changing transformer resistance changes I(RMS). But transformer resistance is related to size and price. If using an "oversized" transformer, this may be a problem. But it looks like to "oversize" the iron feeding a 600uFd 450V cap is unlikely: to reach 2/3rds of rating I used a 330V 2A or 660VA transformer, which is a heck of a lump.

Using that same iron but with a 80uFd cap (ESR=2.4), sticking to the 1A load, I get I(RMS)= 1.5A. This 80uFd cap is rated 0.83A ripple so we are in trouble. BUT the ripple voltage is 71Vp-p or over 15%. I'm sucking a 460+V no-load down to 397V in the bottom of the ripple. If I only wanted heat, that might be fine (except the poor cap!). If I wanted best use of available transformer voltage, this sucks: I'm getting 80% of the test-tone power I could have got with a bigger cap.

I have always suspected (but never got round to proving or finding loopholes) that in "good audio" our ripple voltage needs usually ensure a cap that will not be in real danger of excess current. Apparently an oversized transformer can be a problem. It won't be a problem when designing a 1VA supply off a 20VA transformer because such small caps have large surface/volume ratio, and because small iron is wound for 20% regulation and has high series resistance.

A point I noted in the charts: if the ripple current rating is low, you need a much bigger cap. To get twice the I(RMS) rating, you buy a cap with about 4 times the uFd, Considering the economics of 10-for pricing, if you do kiss a ripple current problem, put the same bucks into a 10-pack of smaller caps. The increased surface area gives higher ripple current handling.