开云体育

Hi Raymond,

Old fashioned electrolytic capacitors have a fairly reactive,
(to aluminum) highly conductive, water based electrolyte.

The capacitor's leakage current creates an electrolytic cell
with the aluminum plates, and removes the oxide layer from the
cathode plate, and builds an oxide layer on the anode plate...

The oxide layer on the anode plate is the dielectric (insulator)
for the capacitor.

This reaction is the "reforming" process that lives on in
electrolytic capacitor lore... Even today. It is also a process
known colloquially as anodizing aluminum.

The aluminum oxide dielectric layer has competing issues:

As the oxide grows thicker, it became a better insulator, and works
to stop the leakage current necessary to grow the oxide layer thicker.

As the oxide layer ceases growing, the electrolyte dissolves the
oxide layer, allowing the leakage currents to increase... growing the
oxide layer thicker.

In normal operation, a balance is reached between leakage current
and growing the oxide layer.... A "working voltage" rating results.

The only way to increase the thickness of the oxide layer once
it reaches equilibrium, is to increase the voltage across the
electrolytic cell, making the cell's current increase, and in turn,
the oxide grow thicker... creating a new higher voltage equilibrium,
and a new "working voltage".

As long as the heat created by the leakage current doesn't raise
the temperature of the electrolyte to a point where the it boils,
the capacitor is fairly happy.

This means that higher than working voltage surges will start to
heat the electrolyte, but as long as the surge goes away before the
electrolyte boils, the capacitor will live to see another day.

One other factor needs mentioning:

The thinner the oxide dielectric layer, the higher the capacitance,
and the lower the working voltage.

The thicker the oxide dielectric layer, the lower the capacitance,
and the higher the working voltage.

The manufacturer had to balance all of these conditions when they
wrote the specifications for their old style electrolytic capacitors.

If you used these old electrolytic capacitors at a lower than working
voltage, their capacitance would increase, and their ability to
operate safely at their specified working voltage would diminish.

Unless you reformed the capacitor... safely limiting the current
until the oxide layer thickened...

Modern electrolytic capacitors use a nonreactive (to aluminum)
electrolyte and as a result, reforming is no longer necessary.

Modern electrolytic capacitors have their oxide layer created, and
their voltage rating determined, before the capacitor is even
assembled.

-Chuck Harris

Raymond Domp Frank wrote:

On Fri, May 29, 2020 at 04:07 PM, Eric wrote:

There are 2 fundamental differences that need to be taken in to account

when

replacing “vintage” caps. And I use the term vintage loosely. One is
tolerance. In the worst case I can remember a capacitors value was +150%

-%100

tolerance. The means that for a 1 uF the measured value of anything between
2.5 uF to .1 uF would be considered “in spec” old radios still amaze me
they ever worked. Now a days +-20% is the normal. The service manual will

tell

you the tolerance of the filters. From memory I am guessing it is going to

be

+100% to -20 % so pretty wide design tolerance on the filters.

The other important spec is the voltage rating of the cap. And here vintage
caps and modern caps differ greatly. Vintage caps were very tolerant of

over

voltage especially given how tube gear warms up before the tube’s comes

in

to operation the B+ will spike some times as high as 200 to 250 volts

higher

then when the device is operating and can hang there for about 15-30

seconds.

This is not an issue in your 475 as it is solid state. However modern caps

are

completely intolerant of over voltage so if you have the physical space it

is

always good to bump up the voltage rating of the cap it wont effect

anything

to replace a 63V cap with a 400V cap excepta little cost in $ maybe one or

2

and physical space it will be slightly bigger then it’s modern lower

voltage

counterpart. However both are usually smaller then their vintage counter

part

even doubling the voltage the modern can will be smaller physically.

I don't think anyone would have accepted caps with +150/-100% tolerance, not

even electrolytics...

Any information on the inability of modern caps to withstand overvoltage

more than vintage caps used to? Maybe WV was just very conservatively spec'ed?

I'd rather have known the actual max.

For (wet) Al electrolytics, choosing a spec-voltage very much higher than

it's ever going to experience, has a deformation effect, so is not
recommended.

Raymond