63 electro cap voltages - why 63?

[anotherjim]

6.3v and 63v are common voltage ratings although 6.3v seems to be in the past now. But I've always wondered why those numbers? Has anybody seen 630v?

[Rob Strand]

Pretty sure Mullard made 630V.


10 16 25 35 (40) 50 63 75 100

I know I worked this out once but roughly voltages seems to have a bit of geometric sequence and a bit of arithmetic going on.

sqrt(10 * 25) ~ 16
sqrt(25 * 50) ~ 35

Above 50V is more linear,

(50 + 75)/ 2 ~ 63

-------
FWIW, you also get 80V, then sqrt(50*80) ~ 63 more geometric!!

[davent]

Lots of the plastic film caps, polyester, polypropylene, polystyrene,  i use in tube amps are 630v.

https://www.justradios.com/cart.html

Digikey list over 4,000, different 6.3v electrolytic caps, in stock around 1,500 different ones.

https://www.digikey.ca/en/products/filter/aluminum-electrolytic-capacitors/58?s=N4IgjCBcoEwAwHYCcVQGMoDMCGAbAzgKYA0IA9lANogBsAdAMwAEAaiALqkAOALlCAFUAdgEseAeUwBZQtnwBXAE6EQAX1WkYVEGmxd8HVUA
dave

[anotherjim]

I'm not sure I get it...
Is it because voltage rating is proportional to capacitance? That for a certain composition and can size then say a 10uf can handle 63v. a 22uF 50v while a 100uF would be 6.3v?
I mean, my point is, was some big customer asking for 63v rather than 50v? Why would the 63 ratings need to exist instead of more rational numbers?

[Rob Strand]

Is it because voltage rating is proportional to capacitance? That for a certain composition and can size then say a 10uf can handle 63v. a 22uF 50v while a 100uF would be 6.3v?

That's pretty much it.   In theory the  C*Vrating product determines the volume but there's more to it - see below.

Now and then you would find info in old data books from the 70's.

I mean, my point is, was some big customer asking for 63v rather than 50v? Why would the 63 ratings need to exist instead of more rational numbers?

The fineness of the steps comes down to what people would buy.   However, people obviously prefer smaller caps so having finer voltage steps lets you limit the size in finer steps.

If you think about it the mechanical packaging of electro has it's own finite steps in diameter and height.   That limits the number of stock aluminum casings required.    Greencaps (old school polyesters) often grow naturally with the size and voltage ratings since they are dipped.   However, encased types tend to take larger steps in size.

There's actually standards for caps (IEC and DIN).   The standards have been around since the late 60's early 70'.  The standards often pin down a lot of parameters and I guess that's where a lot of the rigidity comes from.     

I pulled out a 1974/75 Siemens Aluminum electrolytic data book (capacitors conforming to standards DIN 41240 and DIN 41332) and they show a table of voltages:  3V, 6.3V, 10V, 16V, 25V, 40V, 63V, 70V, 100V, 160V, 250V, 350V, 450V.

I don't have a copy of those DIN standards but it is quite possible that voltage sequence is pinned down in the standards.

The odder ones are:  3V, 40V , 70V and the 450V.

I've seen 70V, 75V, 80V  so I guess that area moves above, possibly because of packing a certain capacitance and voltage in a certain size.     Probably the same goes for over 350V since you see 400V and 450V, maybe even 375V.

EDIT:
Here's a datasheet which mentions one of those DIN standards but the voltages offered are different.   So that could because the voltages aren't pinned down, there standard allows many voltages, the standard has been changed since 1975.
http://static6.arrow.com/aropdfconversion/93703f8ff993e0f0e103e7bf0ee163c48f8f092b/f3304_peg1242020100416.pdf

[PRR]

....Why would the 63 ratings need to exist instead of more rational numbers?

It IS rational.
https://en.wikipedia.org/wiki/Preferred_number
https://en.wikipedia.org/wiki/Renard_series

[Rob Strand]

I suppose one question is at what in time did the voltage rating sequence become the norm?

I suspect it arose when low voltage electrolytic caps hit the scene.   Before say 1960 the demand for low voltage caps was low to non-existent, most  caps were 100V or more.    However as transistor technology become more popular the demands for smaller electrolytic caps increased and that's when the industry kicked off and started to standardize their products.   Even in the mid 60's I remember the voltage ratings for low voltage electros were still quite  patchy and not standardized like we know it today.   When Japanese transistor radios were popular the standard voltages were in place.

[PRR]

I suppose one question is at what in time did the voltage rating sequence become the norm?...

1943-1952. Read footnotes of my first link.

OK, *capacitor voltage* came to the game late. I sure remember 400V-600V-1000V-1600V sequence.

[Rob Strand]

OK, *capacitor voltage* came to the game late. I sure remember 400V-600V-1000V-1600V sequence.

I guess that's the subtlety.  The preferred number sequence was around along time before the cap voltages.

And yep, 400V, 600V, 1000V, 1600V were around.  There was also 125V, 200V, 250V.  (FYI, there's a few earlier cap catalogues on the web maybe Ducon, Cornell Dubilier, Sprague.)

For low voltage caps  I remember seeing 10V, 16V  but also 12V and 20V.   The earlier radios had these, usually axial packages.  Then at some point, perhaps around the late 60's, things moved more towards what we see today.    In particular I remember Elna brand caps and also the radial packages.   Unfortunately not much in terms of catalogues in this era.

[Rob Strand]

Just looking through projects in electronics mags the modern values, modern voltages and radial caps start appearing around 1973 and by 1976 they are the norm.   Mostly Elna brand.  Elna made axials and radials.

In 1972 there's a lot of Mullard axial electrolytics with the older style voltages and values  (eg 6.4V, 40V).

[anotherjim]

When modern production allows small electro caps than of your, wouldn't it have been an opportunity to rationalize the range of avaialble voltage ratings? Do we need to care if the rating is well above our working voltage? I don't think I'd cry if the next higher choice above 50v was 100v.

[garcho]

^ i think a lot of sticking to standards in manufacturing, besides the obvious, also has to do with outfitting your factory with machinery and materials built to some standard or another. If a new standard makes more sense but you have to drop 10 million on new machines and parts, while still dealing with your "legacy" products and customers, just to assuage that little OCD voice in your head, then it doesn't actually end up making sense. Like the imperial system in the US.

[Rob Strand]

When modern production allows small electro caps than of your, wouldn't it have been an opportunity to rationalize the range of avaialble voltage ratings? Do we need to care if the rating is well above our working voltage? I don't think I'd cry if the next higher choice above 50v was 100v.

The size of the 50V vs 100V is the only issue.

i think a lot of sticking to standards in manufacturing, besides the obvious, also has to do with outfitting your factory

Very true.   To some degree it's also why EU and US have taken so long to converge on standards.    In the early days there were companies who developed technology and lead the way, after that were the 'copiers' who copied what they did.    Eventually everyone was doing the same thing.  However, the same thing happened in different countries and it happened slightly differently, so each side doesn't want to change.

[davent]

Was searching for info on toroid transformers today and came across a circuit in the neutral power line feeding the transformer calling for two 33,000uf/6.3v caps.


dave

[Rob Strand]

Was searching for info on toroid transformers today and came across a circuit in the neutral power line feeding the transformer calling for two 33,000uf/6.3v caps.

I stared at both those circuits for about 2 mins and I don't get it.   At first I thought the Crown one was some sort of nasty hack to pass EMC but the caps in the Bryston throws that out the window.

The ground one I can understand. That's pretty common.

[PRR]

......I don't think I'd cry if the next higher choice above 50v was 100v.

If it cost a penny or a millimeter, and you were making *millions* of units, you'd cry for a barely-good-enough part.

[PRR]

> I don't get it.

Blocking DC on the utility power.

[Rob Strand]

Blocking DC on the utility power.

That's probably it.   Big amps with low resistance transformers won't like DC.

https://sound-au.com/articles/xfmr-dc.htm

[davent]

Blocking DC on the utility power.

That's probably it.   Big amps with low resistance transformers won't like DC.

https://sound-au.com/articles/xfmr-dc.htm

> I don't get it.

Blocking DC on the utility power.

Yes and yes.

https://www.diyaudio.com/forums/power-supplies/161428-variations-dc-main-filter-buzzing-toroid-transformers.html#post2088079
dave