Electrolytic cap voltage ratings

[Tubetec]

Ive made a low powered tube HT supply with a step up transformer fed AC from my arb sig gen ,
The first cap is paper in oil ,4uf 400v then 680 ohms followed by a choke ,then an 82uf 300v , then 3 series resistors in the order of 5-10 kohms with 3 x 330uf 200v .
Owing to the fact that the AC current in the later caps is very tiny and limited by the series resistors is it safe to use it upto or maybe slightly over 200 volts ?

 

[ccaudle]

They would probably survive, but for longest life you should use the caps below the rated voltage by a significant percentage. For 200V I would use caps rated at least 250V.

 

[emrr]

The #1 question is what's unloaded voltage?

 

[soapfoot]

There’s also the question of “do you need the caps to survive for 100 hours of use or for 10,000 hours?”

My experience is that up-rating the voltage (and temperature) specs of an electrolytic cap can extend service life pretty dramatically

 

[Tubetec]

Thanks C-c and Doug ,
actual voltage comes out around 180vdc after the rectifier with the particular transformer im using and max output from the generator which is about 16 volts AC , its actually a 600 volt PIO cap after the rectifier , then a 300v electrolytic , then the larger dropper resistors bring the voltage back to around 140vdc on the 200v electrolytics , I know thats fine , but Im wondering if I can safely go to around 220v as the 200v caps only have to handle a tiny fraction of the normal ripple current there meant to see. I have a 100k resistor to ground across the output to simulate the load of a tube in circuit drawing a mA or so . Its meant as a variable supply so it wont nessesairily be cranked up full all the time .
I think I read somewhere here that if your ripple current which itself causes heating (and wear)of the capacitor is very low it may be safe to push the voltages a little say, 10% above the 200v rating .
I know that in many vintage guitar amps Ive serviced over the years , the first filter ,which also handles the highest AC ripple tends to dry out faster than the other caps further down the line , even though voltage at the screens is only fractionally lower .

 

[Newmarket]

Basically don't skimp on any Electrolytic cap rating more than you need to. They are essentially degrading from the off. Lifetime is generally defined as the point where the capacitance is reduced to 50% of nominal. Operating well below rated voltage and temp is key to realising useful lifetimes.

 

[ccaudle]

safely go to around 220v as the 200v caps only have to handle a tiny fraction of the normal ripple current

Ripple current influences heating, but maximum voltage is limited by the insulator thickness. I'm not aware that maximum voltage and ripple current interact at all. There could be secondary effects I have not heard of, but maximum voltage applies even if you have just DC with no ripple current.

 

[Tubetec]

My goal here is to have a versatile quiet bench HT supply for small signal tube stages requiring only a few mA or less , like a mic head amp or first pre stage . All of the electrolytic components are reclaimed from junked PC supplies , everything else is also reclaimed, salvaged or from surplus stocks . Even if one of the 200v caps expires and fails dead short theres enough series resistance to make sure damage to the generator is more or less impossible .
Its not going into studio gear that has to live life more or less permanently powered and withstand variations in the mains supply voltage , I can precisely control the AC supply and the DC output doesnt vary unless the current draw changes. I found a small transformer today with multiple taps 6,9,12,15 volts , I'll crank it upto 110% the dc rating of the caps and see how it goes . I'll try to never run the supply unloaded but use a two pole switch to toggle between dummy load (off) and external load (on) .
thanks also Soapfoot and Newmarket .

 

[trobbins]

Modern e-caps with >315Vdc rating comply with a common standard that allows +10% of rated voltage for 30 secs with a series resistor (ie. to cope with initial power on events). A 200V e-cap may or may not comply with such an operating condition - that would depend on the cap's datasheet.

You likely need to also appreciate that mains voltage could easily add another 5% to your e-cap voltage if your local region had that sort of variation.

Applying an over-voltage will cause partial arcing discharges to occur in the dielectric insulation - such discharges would increase in number with time and over-voltage and exhibit themselves as an increased leakage current. Your circuit would still allow such a leakage current increase, although the series resistance would sag the e-cap voltage, perhaps to maintain a steady-state increase in leakage current. But you are on your own with respect to how that changes parameters within the cap over time.

 

[JohnRoberts]

Keep in mind that cap voltage ratings are generally peak DC,,,,

Cap makers ASSume that engineers will be conservative and design in some extra headroom... Some (dumbass) engineers ASSume that cap makers over design their caps so they can survive X % over voltage. When designing big dog audio power amps the cost difference between the next step higher voltage capacitor could be significant. I knew one guy who got bit in the ass by ASSuming he could cheat the voltage spec by a few percent.

JR

 

[abbey road d enfer]

Owing to the fact that the AC current in the later caps is very tiny and limited by the series resistors is it safe to use it upto or maybe slightly over 200 volts ?

As others have said, relaxing the ripple current constraint does not take pressure off of the voltage limit.
But, for a one-off, why skimp?

 

[Tubetec]

Thanks for all the advice guys,
I have to admit I never saw an electrolytic used right upto its max voltage rating in any equipment .
Typically you find 2/3 to 3/4 of the max voltage , the exception being higher voltage tube power amps where you might find 450v caps on a 400v supply , its pretty obvious to me running close to the max voltage especially after a rectifier where you have both high inrush and high AC current does shorten the life of the component .
Its all good quality Nichicon caps Im using , looks like by adding an extra cap with a higher voltage rating earlier on I can avoid any chance of over voltages in the 200v components ,

I found a print shop that can supply me with all the disposable cameras I want,they process upto 20 units a day , the Fuji branded ones have a high quality 82uf 300v , at most these components have only been charged up a handful of times when the flash is needed , so their still in perfect condition after their usefull life as a flash is over .

 

[abbey road d enfer]

[VEER]I don't care much about photographs. For me they're just a tool to present things.
I was conscious of the existence of disposable cameras in the past, but I had not noticed the renewal of interest for the format. I am shocked that in a context of shortage of raw materials, pollution by plastics and existence of more environment-friendly alternatives (digital cameras), this aberration is not denounced. What do they think?* [\VEER]

*IMO they don't think. They live their lives of contented idiots.

 

[Newmarket]

[VEER]I don't care much about photographs. For me they're just a tool to present things.
I was conscious of the existence of disposable cameras in the past, but I had not noticed the renewal of interest for the format. I am shocked that in a context of shortage of raw materials, pollution by plastics and existence of more environment-friendly alternatives (digital cameras), this aberration is not denounced. What do they think?* [\VEER]

*IMO they don't think. They live their lives of contented idiots.

Agree that the disposables are a waste esp environmentally. Seems to me that the use is targeted at social situations eg wedding reception where one is put on each tables for the guests' use. Obvs low cost risk when the wine is flowing etc. Host can then gather them for processing into actual photographs and/or scanned to digital.
The thing that seems odd to myself is that 35mm film has risen rapidly in price (I have quite a few SLR and compact cameras of various description. Would like to use them more but film and processing costs and availability means that I don't). So I imagine that the disposables might become too expensive even while understanding that they might not have the best film stock in them.

 

[trobbins]

I have to admit I never saw an electrolytic used right upto its max voltage rating in any equipment .

That is likely because it is mains powered, and the mains input has to be designed for circa +10%, and then there may be load variation so that may vary the cap voltage, plus a designer will typically include some safety margin.

Reforming or testing e-cap performance usually includes checking leakage current at the max rated DCV (ie. operate for many minutes at max DCV, and some allow hours depending on the series resistance used and the age of the cap).

 

[JohnRoberts]

Things may have gotten better this century but I recall designing for more that a 10% mains voltage variation.

IIRC OZ was one of the outliers for "high" over voltages ( mains voltage spurs in the outback pushed higher voltage into the long lines, to make up for losses). Peavey sold product to 100(?) countries so SKUs had to work reliably everywhere.

JR

 

[emrr]

The only things i see with very close ratings are early SS when cap size was a constraint, lots of 25V filters on 24V circuits, expecting a regulated PSU. Still….

 

[soapfoot]

The datasheet of a good capacitor will often have a "lifetime multiplier" chart or table that's a function of ripple current and operating temperature.

Operating a capacitor very conservatively will often apply as much as a 200x multiplier to the rated lifetime of a cap.

Operating a 10,000 hour electrolytic with very little ripple current at something close to room temperature will (on paper) afford something close to 2 million hours of rated service life.

I'm not sure the impact of DC voltage is as-often quantified in this manner, but it does seem to have an effect as well--and it goes to show that conservative operation can have rather dramatic positive effects on longevity.

(PS--the reduced ripple current is probably the reason downstream filters sometimes appear to last longer than input filters... more so than slightly-reduced DC operating voltage)

 

[Tubetec]

As I mentioned Im running a step up transformer off the output of my arb generator so mains variations in this case arent an issue, the DC voltage is rock solid ,it only changes when the load changes , initially there is some fractional drift due to the capacitors re-forming but again over days that seems to settle down .
With the large caps and high series resistances the HT supply ramps up over the course of about 5 minutes , theres no massive current spike at switch on like you get when you have a mains power and a bridge rectifier followed by a large capacitor .

I was thinking of trying a couple of gas discharge regulator tubes in series so that when you hit around 250dc at the rectifier output the neons arc up ,give a visual cue and dont allow it rise much further , I'm not sure how much HT current the generator will give out and at least 5-10 ma would be required for gas discharge tubes, is it even worth the penalty in terms of noise to dump maybe 10-20 times the HT current of the actual circuit ? , in any case all the caps in this area of the PSU are 300v or more rated and I can rely on the downstream series resistors in conjunction with an on/off (100k dummy load on ) switch to keep the voltages in check on the larger 330 ,470, 470uf which are 200 volt parts ,

Seems like making provision for leakage current measurement with the meter might be a good plan , is it as simple as connecting the meter set to the uA scale in circuit ? say at the grounded end of the cap ? be interesting to take a closer look at the forming process caps go through ,maybe experiment with frequencies other than 50/60hz and waveforms besides sine and see how leakage current is affected . And maybe making the unit capable of 110% of the max cap voltage isnt such a bad idea, after all it is for bench purposes ,

 

[abbey road d enfer]

I was thinking of trying a couple of gas discharge regulator tubes in series so that when you hit around 250dc at the rectifier output the neons arc up and doesnt allow it rise much further , I'm not sure how much HT current the generator will give out and at least 5-10 ma would be required for gas discharge tubes, is it even worth the penalty in terms of noise to dump maybe 10-20 times the HT current of the actual circuit ?

At the risk of being considered a heathen , may I suggest HV zeners (actually avalanche diodes)? They don't need to draw current permanently .