Topics

Oil Filled Capacitors


Miroslav Pokorni
 

In several instances of discussion on this forum paper caps were called 'oil
caps'. In last case that I remember, Don wrote back and pointed out that
those caps were actually paper, treated with oil but dry paper caps. There
is an auction on e bay for several oil caps with pretty good picture of
them; address is:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2534602284&category=4662

As picture shows, oil caps are enclosed in steel case, to contain oil, and
usual purpose of those caps is phase correction (or motor start), where
considerable current is drawn through the cap.

Regards

Miroslav Pokorni


 

In message <00bc01c325a1$30014da0$3b8c2e04@miroslav2>, Miroslav Pokorni <mpokorni2000@...> writes
In several instances of discussion on this forum paper caps were called 'oil
caps'. In last case that I remember, Don wrote back and pointed out that
those caps were actually paper, treated with oil but dry paper caps. There
is an auction on e bay for several oil caps with pretty good picture of
them; address is:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2534602284&category=4662

As picture shows, oil caps are enclosed in steel case, to contain oil, and
usual purpose of those caps is phase correction (or motor start), where
considerable current is drawn through the cap.

Regards

Miroslav Pokorni
Also very common in power supplies pre electrolytic. For instance the classic RCA valve receiver AR88 had a three capacitor oil filled paper cap for its HT smoothing, at least in the earlier production ones. These were very common in big communication sets of the 1930 - 1940 era.

Robin





Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

--
Robin Birch


Miroslav Pokorni
 

I have never seen them in power supplies, but I guess they would be second
choice to electrolytics because of unfavorable volume/microfarad and
price/microfarad ratio. I am surprised that oil filled were used so widely
even in 30s and 40s.

Regards

Miroslav Pokorni

----- Original Message -----
From: "Robin Birch" <@RobinBirch>
To: "Miroslav Pokorni" <mpokorni2000@...>
Cc: <TekScopes@...>
Sent: Thursday, May 29, 2003 7:53 AM
Subject: Re: [TekScopes] Oil Filled Capacitors


In message <00bc01c325a1$30014da0$3b8c2e04@miroslav2>, Miroslav Pokorni
<mpokorni2000@...> writes
In several instances of discussion on this forum paper caps were called
'oil
caps'. In last case that I remember, Don wrote back and pointed out that
those caps were actually paper, treated with oil but dry paper caps.
There
is an auction on e bay for several oil caps with pretty good picture of
them; address is:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2534602284&category=4662

As picture shows, oil caps are enclosed in steel case, to contain oil,
and
usual purpose of those caps is phase correction (or motor start), where
considerable current is drawn through the cap.

Regards

Miroslav Pokorni
Also very common in power supplies pre electrolytic. For instance the
classic RCA valve receiver AR88 had a three capacitor oil filled paper
cap for its HT smoothing, at least in the earlier production ones.
These were very common in big communication sets of the 1930 - 1940 era.

Robin





Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

--
Robin Birch


Craig Sawyers <c.sawyers@...>
 

I am surprised that oil filled were used so widely
even in 30s and 40s.

Regards

Miroslav Pokorni
Also very common in power supplies pre electrolytic. For instance the
classic RCA valve receiver AR88 had a three capacitor oil filled paper
cap for its HT smoothing, at least in the earlier production ones.
These were very common in big communication sets of the 1930 - 1940 era.
They are a famous problem with the AR88 because they leak and need to be
replaced. Problem is that the oil is really nasty - it is probably pre-PCB
(poly-chlorinated biphenyl) but you get the drift - eco-friendly disposal is
difficult.

Craig


Chuck Harris
 

Miroslav,

Sprague Black Beauties were called oil caps because they ARE
oil caps. All oil caps that I am aware of are paper caps that
have the paper soaked in oil. Not all capacitors that are molded
black plastic are black beauties. All molded black plastic caps
with color bands (like a resistor) and a lump on the banded end's
lead are Sprague Black Beauties.

After you wipe the oily goop off of a Sprague Black Beauty cap,
take it in your hand, and look at the place where the lead comes
out of the color banded end of the cap. You will see a lump where
the lead enters the black plastic.

This lump is actually a small brass tube that is the filler hole
for the oil added to the cap. The oil is added, a lead is stuck
into this hole, crimped slightly, and soldered shut.

One thing that makes these capacitors leak oil is soldering the
lead on the banded end without using a heatsink. The heat from
soldering the lead will upset the solder seal and let the oil
seep out.

This information was told to me by Deane Kidd, of Tektronix.
I already knew that B.B.s were oil, and I asked him why
they were always goopy with oil, why they were always
electrically leaky, and why Tektronix used them in the first place.
(The why tektronix used them had something to do with Vollum not
liking Mallory the company, IIRC.)

After being told this by Deane, I took one apart, and SURPRISE, it
was just as he said. There was a teaspoon of oil in the cap, and
the lump on the banded terminal was a filler tube.

Oil is used in caps to improve the high voltage characteristics
and at the same time improve the dielectric constant of the paper
that is used to physically make the capacitor. By using oil you
can get more capacitance and a higher voltage rating per volume,
than if you just used dry, or waxed paper.

Some oil caps are used in motor start positions, that doesn't mean
all oil caps are motor start caps.

-Chuck


--- In TekScopes@..., "Miroslav Pokorni"
<mpokorni2000@y...> wrote:
In several instances of discussion on this forum paper caps were
called 'oil
caps'.
In last case that I remember, Don wrote back and pointed out that
those caps were actually paper, treated with oil but dry paper caps.
There
is an auction on e bay for several oil caps with pretty good
picture of
them; address is:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item
=2534602284&category=4662

As picture shows, oil caps are enclosed in steel case, to contain
oil, and
usual purpose of those caps is phase correction (or motor start),
where
considerable current is drawn through the cap.

Regards

Miroslav Pokorni


 

In message <PCEBKIFKKMBDKKBBOIJFIEHJEFAA.c.sawyers@...>, Craig Sawyers <c.sawyers@...> writes
I am surprised that oil filled were used so widely
even in 30s and 40s.

Regards

Miroslav Pokorni
Also very common in power supplies pre electrolytic. For instance the
classic RCA valve receiver AR88 had a three capacitor oil filled paper
cap for its HT smoothing, at least in the earlier production ones.
These were very common in big communication sets of the 1930 - 1940 era.
They are a famous problem with the AR88 because they leak and need to be
replaced. Problem is that the oil is really nasty - it is probably pre-PCB
(poly-chlorinated biphenyl) but you get the drift - eco-friendly disposal is
difficult.

Craig
Too true. I've had the one in mine go horribly leaky and have to be replaced. urgh

Robin




Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

--
Robin Birch


ROLYNN PRECHTL K7DFW
 

All molded black plastic caps with color bands (like a resistor) and a
lump on the banded end's lead are Sprague Black Beauties.

Not all were banded. I know the early ones were but......

In the 1969 Sprague Electronics Component catalog, C-619A, The "Difilm Black
Beauty" molded tubulars are shown with markings of voltage and capacitance
and the ad reads "Every capacitor marked twice. No twisting the case to read
ratings".

In a 1973 catalog the markings changed from actual voltage and capacitance
to their part number code which could be interpolated: Cap marked "4TM-P22"
is a 400V 0.22uf cap. Again it was claimed that "Every capacitor marked
twice. No twisting the case to read ratings".


K7DFW

..._._


eboytoronto
 

Hi,
I am a high voltage power supply and I can shed some light on oil filled capacitors.

The challenge in capacitor design is to make an A.C. cpacitor that is free from corona. This why capacitors like Wima FKP1 have an a d.c. rating and an A.C. rating. Metalized plastic and film / foil capcitors with dry construction can operate corona free up to 200 Vac / per section. A typical Wima FKP1 capacitor is designed to have two capacitor section in series giving an A.C. rating of 400 vac while the D.c. rating is typically 1600.

The corona is generated in voids between the plastic and the metal. The corona can be suppressed by filling the voids with oil.

Some capacitors, called mixed dielectric, used a combination of plastic films and oil soaked paper. The paper served to keep the oil in place.
Motor start capacitors are filled with oil to supress corona and conduct heat.

Corona leads to a wear out mechanism, where the capacitors appear to be o.k., but with time and the application of ac voltage they have a wear out mechanism. The ozone generated by the corona, degrades the plastic leading to the failures.

John Barnes

---------- Original Message ----------------------------------
From: "Rolynn Prechtl K7DFW" <k7dfw@...>
Date: Thu, 29 May 2003 13:22:27 -0700


All molded black plastic caps with color bands (like a resistor) and a
lump on the banded end's lead are Sprague Black Beauties.

Not all were banded. I know the early ones were but......

In the 1969 Sprague Electronics Component catalog, C-619A, The "Difilm Black
Beauty" molded tubulars are shown with markings of voltage and capacitance
and the ad reads "Every capacitor marked twice. No twisting the case to read
ratings".

In a 1973 catalog the markings changed from actual voltage and capacitance
to their part number code which could be interpolated: Cap marked "4TM-P22"
is a 400V 0.22uf cap. Again it was claimed that "Every capacitor marked
twice. No twisting the case to read ratings".


K7DFW

..._._





Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/



ROLYNN PRECHTL K7DFW
 

Hi, I am a high voltage power supply and I can shed some light on oil
filled capacitors.


I'll be dipped!

I've worked on a few of these critters but never had one talk to me, much
less send an email.

What will they think of next?

K7DFW

..._._


Richard Solomon, W1KSZ <w1ksz@...>
 

Amongst the "Boatanchor" afficionado's, those things are called "BBOD's".
"Black Beauties of Death" , due to their nasty habit of shorting out and
taking everything associated with it.

73, Dick, W1KSZ

-----Original Message-----
From: Rolynn Prechtl K7DFW [mailto:k7dfw@...]
Sent: Thursday, May 29, 2003 4:22 PM
To: TekScopes@...
Subject: Re: [TekScopes] Re: Oil Filled Capacitors



All molded black plastic caps with color bands (like a resistor) and a
lump on the banded end's lead are Sprague Black Beauties.

Not all were banded. I know the early ones were but......

In the 1969 Sprague Electronics Component catalog, C-619A, The "Difilm Black
Beauty" molded tubulars are shown with markings of voltage and capacitance
and the ad reads "Every capacitor marked twice. No twisting the case to read
ratings".

In a 1973 catalog the markings changed from actual voltage and capacitance
to their part number code which could be interpolated: Cap marked "4TM-P22"
is a 400V 0.22uf cap. Again it was claimed that "Every capacitor marked
twice. No twisting the case to read ratings".


K7DFW

..._._





Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/


Miroslav Pokorni
 

I have never seen a Sprague's Black Beauty and from all that I saw on this
forum, do not have any desire to, either. My first glimpse at wacky Sprague
caps were 'Vitamin Q', those orange, plastic foil jobs, but did not see any
failing of them. If Black Beauties are constructed as you describe them,
that is quite unique, an oil cap that is not contained in steel envelope is
a gutsy enterprise that no one else tried. I can see it is cheaper, because
you avoid metal-glass-metal seal, but result is what you paid for. When you
have a container full of oil, thermal cycling would exert pressure, positive
and negative and hardly anything short of steel can stand that. Once you
have a leak, i.e. something gets out, moisture will come in, no doubt and
caps fortified with moisture are poor performers.

However, if failure mechanism is as you said Deane Kidd described, i.e.
soldering without heat sink (did anyone see such a thing as a heat sink used
on production line?), then those caps were open to failure few months after
manufacturing, as opposed to few years for thermal cycling induced failure.
I wander how long was run of those caps, for how many years they were
manufactured?

Make no mistake, all paper caps are impregnated with oil or wax, paper alone
would not cut it, but that is very different from having a cap in a
container filled with oil.

Initially, paper caps were wound from at least two layers of paper and
aluminum foils; later on, aluminum was vacuum deposited, so we got
self-healing caps. Paper used for dielectric was very thin craft paper and
two layers were used because paper always had defects, but probability of
defects in two sheets of paper coinciding was very small. The paper normally
contained 15% moisture (I knew there had to be some moisture, but had to
look up the value). The wound cap element was vacuum dried and then
impregnated with oil or wax and then encapsulated, initially in phenolyc
(Bakelite), later on, more modern (and cheaper) resins were used. Some
manufacturers did not see that phenolyc is good enough to keep moisture at
bay, so they used metal tube for body of envelope and thick phenolyc for end
caps; I have seen even epoxy end caps.

On the other hand, if cap was to be 'oil cap', then oil impregnated element
was placed in steel can (I guess, only the Sprague was 'smart'), cap leads
were attached to lid mounted penetrators and lid attached to can, soldered
or welded, depending on the service rating of end product. After that, units
were placed in vacuum chamber, oil was injected and filler tube was crimped
and soldered or welded; vacuum was used to avoid residual air bubbles in
oil. The purpose of having a cap floating in oil was to increase power
dissipation; all those power circuit caps see quite a bit of current, hence,
dissipation is a factor.

Paper caps have been used, I understand very successfully, from the
beginning of radio, everyone seem to agree that Leiden Bottles were too
bulky. The impregnated paper as dielectric has some drawbacks, but leakage
is not one of them, if manufacturing process is right. It is only the advent
of plastic foil type that made production yields better and replaced paper
type.

As an aside, encapsulation in plastic, in general, as a mean of protection
from moisture penetration, was a holly grail of IC industry, eventually
abandoned. It was RCA's 15 years of efforts that ended those false
expectations. Semiconductor dies are normally passivated using low
temperature glass, which is phosphor based. Over period of time, moisture
accumulates in die cavity and acts with phosphor in glass to form phosphoric
acid. When power supply voltages of 15 V or higher are applied, an
electromigration of phosphoric acid takes place and when enough molecules
are in the path between two electrodes, a full blown electrolyzes develops
and IC gets destroyed. Those facts were well known, false expectations came
from wishful thinking that plastic can be formulated so that it would serve
as an efficient moisture barrier. This problem effected only analog, CMOS, p
ower circuits and discrete power devices (transistors, SCRs etc.), possibly
early MOS, because standard logic was at safe 5 V. The RCA's conclusion was
not to go for better plastic but to change passivation glass to something
that does not have phosphor. They did that and rest of industry followed.

This might be a clue as to what goes wrong with paper caps that have been in
use for many years. Simply, oil might get contaminated with moisture that
penetrates through phenolyc encapsulant and develops electrical leak. At one
point, there is enough leakage current to heat up the cap and enough reserve
moisture to feed thermally accelerated oil contamination process. Maybe,
those people who enclosed caps in metal tubes, knew something and kept it to
themselves.

Regards

Miroslav Pokorni

----- Original Message -----
From: "chuck_585a" <cfharris@...>
To: <TekScopes@...>
Sent: Thursday, May 29, 2003 12:46 PM
Subject: [TekScopes] Re: Oil Filled Capacitors


Miroslav,

Sprague Black Beauties were called oil caps because they ARE
oil caps. All oil caps that I am aware of are paper caps that
have the paper soaked in oil. Not all capacitors that are molded
black plastic are black beauties. All molded black plastic caps
with color bands (like a resistor) and a lump on the banded end's
lead are Sprague Black Beauties.

After you wipe the oily goop off of a Sprague Black Beauty cap,
take it in your hand, and look at the place where the lead comes
out of the color banded end of the cap. You will see a lump where
the lead enters the black plastic.

This lump is actually a small brass tube that is the filler hole
for the oil added to the cap. The oil is added, a lead is stuck
into this hole, crimped slightly, and soldered shut.

One thing that makes these capacitors leak oil is soldering the
lead on the banded end without using a heatsink. The heat from
soldering the lead will upset the solder seal and let the oil
seep out.

This information was told to me by Deane Kidd, of Tektronix.
I already knew that B.B.s were oil, and I asked him why
they were always goopy with oil, why they were always
electrically leaky, and why Tektronix used them in the first place.
(The why tektronix used them had something to do with Vollum not
liking Mallory the company, IIRC.)

After being told this by Deane, I took one apart, and SURPRISE, it
was just as he said. There was a teaspoon of oil in the cap, and
the lump on the banded terminal was a filler tube.

Oil is used in caps to improve the high voltage characteristics
and at the same time improve the dielectric constant of the paper
that is used to physically make the capacitor. By using oil you
can get more capacitance and a higher voltage rating per volume,
than if you just used dry, or waxed paper.

Some oil caps are used in motor start positions, that doesn't mean
all oil caps are motor start caps.

-Chuck


--- In TekScopes@..., "Miroslav Pokorni"
<mpokorni2000@y...> wrote:
In several instances of discussion on this forum paper caps were
called 'oil
caps'.
In last case that I remember, Don wrote back and pointed out that
those caps were actually paper, treated with oil but dry paper caps.
There
is an auction on e bay for several oil caps with pretty good
picture of
them; address is:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item
=2534602284&category=4662

As picture shows, oil caps are enclosed in steel case, to contain
oil, and
usual purpose of those caps is phase correction (or motor start),
where
considerable current is drawn through the cap.

Regards

Miroslav Pokorni


Chuck Harris
 

Hi Miroslav,

Miroslav Pokorni wrote:
I have never seen a Sprague's Black Beauty and from all that I saw on this
forum, do not have any desire to, either. My first glimpse at wacky Sprague
caps were 'Vitamin Q', those orange, plastic foil jobs, but did not see any
I think that Sprague called just about everything they made "Vitamin Q"
at one point or another. I have seen waxed paper caps with that name,
and molded plastic caps, and I was just checking my junk box, and I have
some "Vitamin Q" caps that are metal tubes with glass end seals.

The orange drop caps are quite reliable, and as such well regarded...
and were copied by just about everyone.

failing of them. If Black Beauties are constructed as you describe them,
that is quite unique, an oil cap that is not contained in steel envelope is
a gutsy enterprise that no one else tried. I can see it is cheaper, because
you avoid metal-glass-metal seal, but result is what you paid for. When you
have a container full of oil, thermal cycling would exert pressure, positive
and negative and hardly anything short of steel can stand that.
The caps have a small bubble of air in them that cushions them from the
expansion properties of the oil. In other words, if there was no air,
there would be the hydraulic pressure of the incompressable oil pressing
at the seals. But there is air, so the bubble just gets bigger or
or smaller as the pressure changes. This kind of cap doesn't generate
a significant amount of internal heating, so the pressure that results
from oil's expansion isn't much at all.

Once you
have a leak, i.e. something gets out, moisture will come in, no doubt and
caps fortified with moisture are poor performers.
However, if failure mechanism is as you said Deane Kidd described, i.e.
soldering without heat sink (did anyone see such a thing as a heat sink used
on production line?), then those caps were open to failure few months after
manufacturing, as opposed to few years for thermal cycling induced failure.
I wander how long was run of those caps, for how many years they were
manufactured?
The first time I saw the BBs was in a tek 513D. According to Stan, the
513 was made in 1953. The last time I saw them in Tek scopes was the
545 (no suffix letter). According to Stan, the 545 was made in 1958.
I think it is safe to assume that Tek used BBs from the very beginning
until 1960, or so.

I suspect that more failures were noted in the big caps that could only
fit in the terminal blocks with short leads, than in the little caps
that would have 1" of leads. I seem to remember Deane saying that they
started to heatsink them in production, and the problems went away, sort
of. At least, the problems were postponed. I suspect that all of them
fail sooner or later due to the metal to plastic seal.

...

Paper caps have been used, I understand very successfully, from the
beginning of radio, everyone seem to agree that Leiden Bottles were too
bulky. The impregnated paper as dielectric has some drawbacks, but leakage
is not one of them, if manufacturing process is right. It is only the advent
of plastic foil type that made production yields better and replaced paper
type.
Virtually every wax impregnated paper dielectric cap that I have
found in radios that I have serviced has been leaky to the point
of drawing high hundreds of microamps at rated voltage. That does
wonderful things when the cap is a coupling cap to a high impedance
grid circuit.

...

This might be a clue as to what goes wrong with paper caps that have been in
use for many years. Simply, oil might get contaminated with moisture that
penetrates through phenolyc encapsulant and develops electrical leak. At one
point, there is enough leakage current to heat up the cap and enough reserve
moisture to feed thermally accelerated oil contamination process. Maybe,
those people who enclosed caps in metal tubes, knew something and kept it to
themselves.
The leakage problem comes from two directions. First is due to the oil
leaving the cap. A very large part of the dielectric's insulating value
is due to the oil. Second is from the introduction of moisture into the
oil.

Hermetically sealed oil caps, or even paper caps, last virtually
forever. The only problem is affording the cost.

-Chuck, WA3UQV


Miroslav Pokorni
 

Corona discharge would be pretty hard to induce at 200 Vac. You would have
to hit frequency around cap's resonance, but then Q factor would be quite
miserable, so there would not be much of voltage amplification.
I can see that if you are getting corona, that can ruin a day for a high
voltage power supply, like yourself. Corona would cause impedance of a cap
to drop to next to zero and burn everythin in the path, it is just that you
need pretty high voltage for that.

Regards

Miroslav Pokorni

----- Original Message -----
From: "jbarnes" <jbarnes@...>
To: <TekScopes@...>; "Rolynn Prechtl K7DFW"
<k7dfw@...>
Sent: Thursday, May 29, 2003 2:53 PM
Subject: Re: [TekScopes] Re: Oil Filled Capacitors


Hi,
I am a high voltage power supply and I can shed some light on oil
filled capacitors.

The challenge in capacitor design is to make an A.C. cpacitor that is
free from corona. This why capacitors like Wima FKP1 have an a d.c. rating
and an A.C. rating. Metalized plastic and film / foil capcitors with dry
construction can operate corona free up to 200 Vac / per section. A typical
Wima FKP1 capacitor is designed to have two capacitor section in series
giving an A.C. rating of 400 vac while the D.c. rating is typically 1600.

The corona is generated in voids between the plastic and the metal.
The corona can be suppressed by filling the voids with oil.

Some capacitors, called mixed dielectric, used a combination of
plastic films and oil soaked paper. The paper served to keep the oil in
place.
Motor start capacitors are filled with oil to supress corona and
conduct heat.

Corona leads to a wear out mechanism, where the capacitors appear
to be o.k., but with time and the application of ac voltage they have a wear
out mechanism. The ozone generated by the corona, degrades the plastic
leading to the failures.

John Barnes


---------- Original Message ----------------------------------
From: "Rolynn Prechtl K7DFW" <k7dfw@...>
Date: Thu, 29 May 2003 13:22:27 -0700


All molded black plastic caps with color bands (like a resistor) and a
lump on the banded end's lead are Sprague Black Beauties.

Not all were banded. I know the early ones were but......

In the 1969 Sprague Electronics Component catalog, C-619A, The "Difilm
Black
Beauty" molded tubulars are shown with markings of voltage and
capacitance
and the ad reads "Every capacitor marked twice. No twisting the case to
read
ratings".

In a 1973 catalog the markings changed from actual voltage and
capacitance
to their part number code which could be interpolated: Cap marked
"4TM-P22"
is a 400V 0.22uf cap. Again it was claimed that "Every capacitor marked
twice. No twisting the case to read ratings".


K7DFW

..._._





Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/



Morris Odell <morriso@...>
 

Miroslav wrote:

Corona discharge would be pretty hard to induce at 200 Vac. You would have
to hit frequency around cap's resonance, but then Q factor would be quite
miserable, so there would not be much of voltage amplification.
I can see that if you are getting corona, that can ruin a day for a high
voltage power supply, like yourself. Corona would cause impedance of a cap
to drop to next to zero and burn everythin in the path, it is just that you
need pretty high voltage for that.

I am a high voltage power supply and I can shed some light on oil
filled capacitors.

The challenge in capacitor design is to make an A.C. cpacitor that is
free from corona.

<snip>

I'm pretty sure the power supply was talking about corona initially on a
pretty small scale. It's well known that the voltage stress across
insulating voids is higher than that across a corresponding volume of
insulator due to the distribution of the electric field between areas of
different dielectric constant. The problem would be worse for AC because of
the changing field - perhaps due to dielectric loss effects too. Once a
micro corona develops it would lead to local destruction and eventually
failure of the entire unit. One of the reasons for impregnating the
dielectric with oil is to fill the voids and prevent this from happening. I
suppose the oil would really be the dielectric and the cellulose fibres of
the paper merely a carrier/spacer.

Plastic film dielectrics don't have the porous structure of paper with its
inherent voids and so doesn't need to be impregnated. I suppose you can
think of plastic films as "solid oil" anyway.

Morris


Craig Sawyers <c.sawyers@...>
 

Corona discharge would be pretty hard to induce at 200 Vac. You would have
to hit frequency around cap's resonance
That reminds me of another failure mechanism - acoustic. Try the following
experiment. Make a simple power amp for square waves with a couple of
transistors or FETS, driving +/- 30V (say). Connect capacitor from output
to low impedance load to get a reasonable current flowing. Now sweep an
audio generator.

Many capacitors howl like a banshee at particular frequencies where the
internal mechanics resonate. These are usually ones where the windings
inside are relatively "loose". Very tightly wound capacitors are very
quiet - but there are some real shocks out there when you try the experiment
(like some respected 10kV polypropylene which were more like a loudspeaker
than a capacitor).

Needless to say, such mechanical vibration does nothing to help the
reliability of the capacitor.

Craig


Miroslav Pokorni
 

I wonder whether they talk when you squeeze them?

Regards

Miroslav Pokorni

----- Original Message -----
From: "Rolynn Prechtl K7DFW" <k7dfw@...>
To: <TekScopes@...>
Sent: Thursday, May 29, 2003 2:59 PM
Subject: Re: [TekScopes] Re: Oil Filled Capacitors




Hi, I am a high voltage power supply and I can shed some light on oil
filled capacitors.


I'll be dipped!

I've worked on a few of these critters but never had one talk to me, much
less send an email.

What will they think of next?

K7DFW

..._._


Miroslav Pokorni
 

Yes, Morris, you are right, I did not consider that an area where there is a
transition between air and high dielectric constant material makes for some
very ugly field distribution. However, that is at the fringes of the cap,
not between 'plates' of it.

Your reply made me go and look up dielectric constants, what I should have
done in the first place. Paper shows values between 3.3 and 3.7, while oils
that are used for caps, chlorinated phenyls and naphthalenes (trade names
HB-40 Oil, Pyranol 1476, Arochlor 1260, Halowax Oil 1000), are between 2.6
and 5.0. There is not much disparity there, ratio is less than 1:2, so there
would not be much of electric field change going from solid paper particles
to liquid oil, at least not between plates in an oil field cap. On the
fringes of a cap, oil is held between dielectric layers by capillary action
so there is a continuity of dielectric there. Besides, applicable to both,
inter-plate and on fringes, oil being liquid and remaining so, is kind of
self-healing, through diffusion.

Looking at the dielectric constant of most frequently used plastic foil
materials, (polyethylene, polypropylene and polystyrene), there is nothing
dramatic, they are bunched at 2.4. The only other common capacitor
dielectric material, polycarbonate, stands at 3.2. These are not quite the
values that would make field discontinuity which might lead to local corona.

So, I guess, that corona stuff at 200 Vac is an artifact of Wima's marketing
department. I was aware that Wima was trying to tell world that there is no
such a thing as Wima when it comes to line voltage caps, but Taiwanese seem
to have held pretty good.

Regards

Miroslav Pokorni

----- Original Message -----

From: "Morris Odell" <morriso@...>
To: <TekScopes@...>
Sent: Thursday, May 29, 2003 11:15 PM
Subject: RE: [TekScopes] Re: Oil Filled Capacitors


Miroslav wrote:

Corona discharge would be pretty hard to induce at 200 Vac. You would
have
to hit frequency around cap's resonance, but then Q factor would be quite
miserable, so there would not be much of voltage amplification.
I can see that if you are getting corona, that can ruin a day for a high
voltage power supply, like yourself. Corona would cause impedance of a cap
to drop to next to zero and burn everythin in the path, it is just that
you
need pretty high voltage for that.

I am a high voltage power supply and I can shed some light on oil
filled capacitors.

The challenge in capacitor design is to make an A.C. cpacitor that
is
free from corona.

<snip>

I'm pretty sure the power supply was talking about corona initially on a
pretty small scale. It's well known that the voltage stress across
insulating voids is higher than that across a corresponding volume of
insulator due to the distribution of the electric field between areas of
different dielectric constant. The problem would be worse for AC because
of
the changing field - perhaps due to dielectric loss effects too. Once a
micro corona develops it would lead to local destruction and eventually
failure of the entire unit. One of the reasons for impregnating the
dielectric with oil is to fill the voids and prevent this from happening.
I
suppose the oil would really be the dielectric and the cellulose fibres of
the paper merely a carrier/spacer.

Plastic film dielectrics don't have the porous structure of paper with its
inherent voids and so doesn't need to be impregnated. I suppose you can
think of plastic films as "solid oil" anyway.

Morris


Miroslav Pokorni
 

I guess, it does not help, but then everything has a resonant frequency. The
only way to cope with it, if you can not avoid operating at that frequency,
is to make cap mechanically lossy so that Q factor is down in mud.

Possibly, this acoustical phenomena, together with high current through cap,
made resonant power supplies pretty much disappear. The last one that I have
seen was a Hewlett Packard unit that used a cap branded with their name,
what I guess was a custom cap.

There seems to be no such a thing as a simple component, there is only the
simple demand.

Regards

Miroslav Pokorni

----- Original Message -----
From: "Craig Sawyers" <c.sawyers@...>
To: "TekScopes" <TekScopes@...>
Sent: Friday, May 30, 2003 12:27 AM
Subject: RE: [TekScopes] Re: Oil Filled Capacitors


Corona discharge would be pretty hard to induce at 200 Vac. You would
have
to hit frequency around cap's resonance
That reminds me of another failure mechanism - acoustic. Try the
following
experiment. Make a simple power amp for square waves with a couple of
transistors or FETS, driving +/- 30V (say). Connect capacitor from output
to low impedance load to get a reasonable current flowing. Now sweep an
audio generator.

Many capacitors howl like a banshee at particular frequencies where the
internal mechanics resonate. These are usually ones where the windings
inside are relatively "loose". Very tightly wound capacitors are very
quiet - but there are some real shocks out there when you try the
experiment
(like some respected 10kV polypropylene which were more like a loudspeaker
than a capacitor).

Needless to say, such mechanical vibration does nothing to help the
reliability of the capacitor.

Craig


Stan & Patricia Griffiths <w7ni@...>
 

At least one of the caps in that auction is a Tek timing capacitor and I am
not sure Tek ever used oil in those or not. I never took one apart.

I have sometimes referred to "Black Beauty" caps as oil caps since the
actually have oil in them. Technically they may be "dry paper, treated with
oil" but there is so much oil in them that it often leaks out by the wire
seals and gets all over the outside of the cap and sometimes drips onto the
instrument chassis. It is hard for me to think of that as a "dry" cap . . .

Stan
w7ni@...

----- Original Message -----
From: "Miroslav Pokorni" <mpokorni2000@...>
To: <TekScopes@...>
Sent: Wednesday, May 28, 2003 10:14 PM
Subject: [TekScopes] Oil Filled Capacitors


In several instances of discussion on this forum paper caps were called
'oil
caps'. In last case that I remember, Don wrote back and pointed out that
those caps were actually paper, treated with oil but dry paper caps. There
is an auction on e bay for several oil caps with pretty good picture of
them; address is:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2534602284&category=4662

As picture shows, oil caps are enclosed in steel case, to contain oil, and
usual purpose of those caps is phase correction (or motor start), where
considerable current is drawn through the cap.

Regards

Miroslav Pokorni






Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/



eboytoronto
 

Hi,
I am not really a high voltage power supply !!

I am a high voltage power supply engineer. I work for CPI designing power supplies for Medical X-ray generators. These are up to 100kW, 150kV and 1000mA.
Some of my work can be seen at www.cpii.com/cmp
These incidently use resonant technology.

I learned about the corona effects from a company called ASC. This company was TRW. They make and have made many custom capacitors for HP, Tek, Varian, Fluke etc. They are generally white wrap and fill construction.

They explained the corona effects in capacitors to me. They have an application note at:

http://www.ascapacitor.com/PDF/corona%20phenomena.pdf

Wima built their capacitors with two sections in series to help with the self-healing properties. When one of the sections arcs the other still presents impedance.

ASC has built some snubber capacitors for me that have four sections in series to obtain an a.c. rating of 800 v rms. The dc. rating of these was 1600 v.

As someone pointed out the corona arises because the electric field is not uniform in the area of a void or mismatched dielectric constants.

These comments on capacitors also apply to high voltage transformers. This why vacumn impregnating is used.

I have done some work with solid dielectric materials, potting, matching the dielectrics and preventing voids is quite a challenge.


Regards,


John Barnes

P.S. just got home from the Rochester Hamfest with a pile of 7K plugins to work on.

---------- Original Message ----------------------------------
From: "Miroslav Pokorni" <mpokorni2000@...>
Reply-To: "Miroslav Pokorni" <mpokorni2000@...>
Date: Fri, 30 May 2003 02:42:39 -0700

I guess, it does not help, but then everything has a resonant frequency. The
only way to cope with it, if you can not avoid operating at that frequency,
is to make cap mechanically lossy so that Q factor is down in mud.

Possibly, this acoustical phenomena, together with high current through cap,
made resonant power supplies pretty much disappear. The last one that I have
seen was a Hewlett Packard unit that used a cap branded with their name,
what I guess was a custom cap.

There seems to be no such a thing as a simple component, there is only the
simple demand.

Regards

Miroslav Pokorni

----- Original Message -----
From: "Craig Sawyers" <c.sawyers@...>
To: "TekScopes" <TekScopes@...>
Sent: Friday, May 30, 2003 12:27 AM
Subject: RE: [TekScopes] Re: Oil Filled Capacitors


Corona discharge would be pretty hard to induce at 200 Vac. You would
have
to hit frequency around cap's resonance
That reminds me of another failure mechanism - acoustic. Try the
following
experiment. Make a simple power amp for square waves with a couple of
transistors or FETS, driving +/- 30V (say). Connect capacitor from output
to low impedance load to get a reasonable current flowing. Now sweep an
audio generator.

Many capacitors howl like a banshee at particular frequencies where the
internal mechanics resonate. These are usually ones where the windings
inside are relatively "loose". Very tightly wound capacitors are very
quiet - but there are some real shocks out there when you try the
experiment
(like some respected 10kV polypropylene which were more like a loudspeaker
than a capacitor).

Needless to say, such mechanical vibration does nothing to help the
reliability of the capacitor.

Craig





Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/