Question about Capacitors

[Lostgallifreyan]

Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote in
news:48FFCDEC.D52D064@hotmail.com:

Andre Majorel wrote:

Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote:
DJ Delorie wrote:

Don't use ceramic with high ripple currents

Ceramics are NEVER used with high ripple currents you brain dead
defective.

Go away and die.

Is that the real Graham Stevenson or a forgery ? The Path: looks
genuine to me.

It's genuine.

Anyone who thinks you use ceramic caps as reservoir caps needs their
brain examined.

Graham

Are there no other situations where ripple current matters? Perhaps unusual
power HF systems?

 

[Jamie]

Lostgallifreyan wrote:

Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote in
news:48FFCDEC.D52D064@hotmail.com:



Andre Majorel wrote:


Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote:

DJ Delorie wrote:


Don't use ceramic with high ripple currents

Ceramics are NEVER used with high ripple currents you brain dead
defective.

Go away and die.

Is that the real Graham Stevenson or a forgery ? The Path: looks
genuine to me.

It's genuine.

Anyone who thinks you use ceramic caps as reservoir caps needs their
brain examined.

Graham




Are there no other situations where ripple current matters? Perhaps unusual
power HF systems?

Don't mind the Ham, I mean Graham, he's living in the audio world only.

I think anything above 50 Hz is challenging to him.

Btw, we used 1 Mhz signals in faraday cages at various levels of power
to test ripple currents..

1 Mhz was also the freq used in most of the other test.


http://webpages.charter.net/jamie_5"

 

[Lostgallifreyan]

Jamie <jamie_ka1lpa_not_valid_after_ka1lpa_@charter.net> wrote in
news:5IQLk.7051$Zc.1676@newsfe09.iad:

Btw, we used 1 Mhz signals in faraday cages at various levels of power
to test ripple currents..

Wouldn't that have tested the dielectrics in various entertaining ways quite
beyond ripple currents? I imagine some cap parts would have parted
moorings given the kind of stresses this might involve. Not that I know a
lot, just got visions of taking this to Tesla levels of experimentation..

 

[Phil Allison]

"Lostgallifreyan"

Eeysore to Fake

Anyone who thinks you use ceramic caps as reservoir caps needs their
brain examined.


Are there no other situations where ripple current matters? Perhaps
unusual
power HF systems?

** The " ripple current" rating matters whenever ANY capacitor is passing
significant amounts of current - such that its ESR causes it to self heat.
Cap makers do not publish ripple current figures for all types, but usually
do publish ESR values and curves from which the amount of self heating can
be calculated - at least approximately.

Ignore this at your peril - eg:

The common X cap for mains suppression is well designed for its job, with
250 volts AC at 50 or 60Hz - but what say if you apply that same voltage
at 10kHz ??

Answer - it smokes and explodes.



...... Phil

 

[Lostgallifreyan]

"Phil Allison" <philallison@tpg.com.au> wrote in
news:6ma65lFdr6s2U1@mid.individual.net:

** The " ripple current" rating matters whenever ANY capacitor is
passing significant amounts of current - such that its ESR causes it to
self heat.

Agreed. Small mica caps seemed to me to be at risk but the smallness of their
values just seemed to imply that HF would be needed to make enough current
pass to put them at risk.

 

[Eeyore]

Lostgallifreyan wrote:

Eeyore wrote
Andre Majorel wrote:
Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote:
DJ Delorie wrote:

Don't use ceramic with high ripple currents

Ceramics are NEVER used with high ripple currents you brain dead
defective.

Go away and die.

Is that the real Graham Stevenson or a forgery ? The Path: looks
genuine to me.

It's genuine.

Anyone who thinks you use ceramic caps as reservoir caps needs their
brain examined.

Are there no other situations where ripple current matters? Perhaps unusual
power HF systems?

Can you be more specific ? At medium HF, polypropylene is a good dielectric to
use and since values tend to be lower at HF, not a problem size wise. I've put
8-10A though a relatively small polyprop cap.

Graham

 

[Eeyore]

Lostgallifreyan wrote:

"Phil Allison" <philallison@tpg.com.au> wrote

** The " ripple current" rating matters whenever ANY capacitor is
passing significant amounts of current - such that its ESR causes it to
self heat.

Agreed. Small mica caps seemed to me to be at risk but the smallness of their
values just seemed to imply that HF would be needed to make enough current
pass to put them at risk.

Perhaps you could be more specific about your concerns, e.g. a specific
application ?

Graham

 

[Lostgallifreyan]

Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote in
news:49005465.44E68389@hotmail.com:

Lostgallifreyan wrote:

"Phil Allison" <philallison@tpg.com.au> wrote

** The " ripple current" rating matters whenever ANY capacitor is
passing significant amounts of current - such that its ESR causes it
to self heat.

Agreed. Small mica caps seemed to me to be at risk but the smallness of
their values just seemed to imply that HF would be needed to make
enough current pass to put them at risk.

Perhaps you could be more specific about your concerns, e.g. a specific
application ?

Graham

No, I couldn't.
The point is that caps are all able to be damaged by excess AC currrent and
that if you use a high enough frequency it's easy to push such currents
through them. That this is true invalidates limiting to specifics as a means
to question the general case. If you find proof of an exception it's up to
you to assert it.

 

[Lostgallifreyan]

Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote in
news:490053AE.990915F6@hotmail.com:

Are there no other situations where ripple current matters? Perhaps
unusual power HF systems?

Can you be more specific ? At medium HF, polypropylene is a good
dielectric to use and since values tend to be lower at HF, not a problem
size wise. I've put 8-10A though a relatively small polyprop cap.

Again, no. I was asking. It wasn't a loaded question. I wasn't suggesting
that I know something specific. By HF, maybe I should have said RF because
that's what I meant. One possible example occurs to me: an unstable audio amp
might oscillate at RF if designed wrong, and that might be one way a
capacitor might get damaged by a high current at high frequency. Fixing the
cap isn't the answer in this case but it would prove the cap's limits if it
happened.

 

[Phil Allison]

"Lostgallifreyan"

One possible example occurs to me: an unstable audio amp
might oscillate at RF if designed wrong, and that might be one way a
capacitor might get damaged by a high current at high frequency.

** That is no imaginary scenario - it ACTUALLY happens.

But it don't take a major design error, just misuse on the part of the
owner - since almost any power amp will break into full level, HF
oscillation of there is capacitive coupling from output back to input.

A few pF will do the trick, in most cases.




..... Phil

 

[Lostgallifreyan]

"Phil Allison" <philallison@tpg.com.au> wrote in
news:6mb6coFg2c4vU1@mid.individual.net:

"Lostgallifreyan"


One possible example occurs to me: an unstable audio amp
might oscillate at RF if designed wrong, and that might be one way a
capacitor might get damaged by a high current at high frequency.


** That is no imaginary scenario - it ACTUALLY happens.

Cool. I've seen burnt resistors in a few amps but rarely caps that weren't in
the PSU.

But it don't take a major design error, just misuse on the part of the
owner - since almost any power amp will break into full level, HF
oscillation of there is capacitive coupling from output back to input.

A few pF will do the trick, in most cases.

Interesting. I can see why, but wouldn't a high pass filter be useful on the
input? Even a 1 pole RC type might roll off enough HF to reduce risk to
capacitors. Capacitor bypass in the loops of ap-amps is another way.

 

[Phil Allison]

"Lostgallifreyan"

"Phil Allison"
"Lostgallifreyan"


One possible example occurs to me: an unstable audio amp
might oscillate at RF if designed wrong, and that might be one way a
capacitor might get damaged by a high current at high frequency.


** That is no imaginary scenario - it ACTUALLY happens.


Cool. I've seen burnt resistors in a few amps but rarely caps that weren't
in
the PSU.

But it don't take a major design error, just misuse on the part of the
owner - since almost any power amp will break into full level, HF
oscillation of there is capacitive coupling from output back to input.

A few pF will do the trick, in most cases.


Interesting. I can see why, but wouldn't a high pass filter be useful on
the
input? Even a 1 pole RC type might roll off enough HF to reduce risk to
capacitors. Capacitor bypass in the loops of ap-amps is another way.

** You have to do all the math - but suffice to say that there will be
a cap value in the tens of pF that WILL cause full power oscillation in
any power amp that meets the usual ( audio band ) input impedance value and
bandwidth spec of such amps.



..... Phil

 

[Eeyore]

Lostgallifreyan wrote:

Eeyore wrote in
Lostgallifreyan wrote:
"Phil Allison" wrote

** The " ripple current" rating matters whenever ANY capacitor is
passing significant amounts of current - such that its ESR causes it
to self heat.

Agreed. Small mica caps seemed to me to be at risk but the smallness of
their values just seemed to imply that HF would be needed to make
enough current pass to put them at risk.

Perhaps you could be more specific about your concerns, e.g. a specific
application ?

No, I couldn't.
The point is that caps are all able to be damaged by excess AC currrent and
that if you use a high enough frequency it's easy to push such currents
through them. That this is true invalidates limiting to specifics as a means
to question the general case. If you find proof of an exception it's up to
you to assert it.

Given sufficient resources you can 'blow up' just about anything. Not sure what
that proves though.

Graham

 

[Eeyore]

Lostgallifreyan wrote:

Eeyore wrote

Are there no other situations where ripple current matters? Perhaps
unusual power HF systems?

Can you be more specific ? At medium HF, polypropylene is a good
dielectric to use and since values tend to be lower at HF, not a problem
size wise. I've put 8-10A though a relatively small polyprop cap.

Again, no. I was asking. It wasn't a loaded question. I wasn't suggesting
that I know something specific. By HF, maybe I should have said RF because
that's what I meant. One possible example occurs to me: an unstable audio amp
might oscillate at RF if designed wrong, and that might be one way a
capacitor might get damaged by a high current at high frequency. Fixing the
cap isn't the answer in this case but it would prove the cap's limits if it
happened.

Far more likely to burn out some resistors and transistors actually.

Graham

 

[Eeyore]

Lostgallifreyan wrote:

but wouldn't a high pass filter be useful on the input?

I presume you mean low pass ? But people like amps to repsond way beyong the
audio spectrum these days.

Balanced inputs fix it of course as in professional gear.

Graham

 

[Lostgallifreyan]

Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote in
news:49007F1B.57A9E0C2@hotmail.com:

but wouldn't a high pass filter be useful on the input?

I presume you mean low pass ? But people like amps to repsond way beyong
the audio spectrum these days.

High pass to ground. Decoupling. Even though people like more bandwidth there
could be some use of that as prevention. If the caps are chosen to withstand
up to 100 KHz in a condition where that is the frequency of full-bore
feedback in an error condition then you pass anything above 50 KHz to ground,
you'd still have twice the bandwidth that we can hear before it fell off at
6db with a 1 pole RC filter, and the caps will be protected against worst
case feedback at any frequency. Phil says that any audio power amp with
standard specs is vulnerable though, and I'll have to think about that to
understand why, based on my assumption of 20 HZ (or DC) to 20 KHz, with input
impedance of 10K, maybe 100K. For the moment, I can't see why decoupling
unwanted HF range on the input or in op-amp gain stages won't protect as I
suggest above if the cap in that filter is able to handle that kind of
current at HF and even low RF.

 

[Phil Allison]

"Eeysore the Congenital Fuckwit "

Lostgallifreyan wrote:

but wouldn't a high pass filter be useful on the input?

I presume you mean low pass ? But people like amps to repsond way beyong
the
audio spectrum these days.

Balanced inputs fix it of course as in professional gear.

** More asinine drivel from the know nothing Stuffed Donkey.

Just makes it up as he goes along.

Then poops it out all over usenet.

Dats what they all call him "poopie" .

Don't they, Graham.





..... Phil

 

[Eeyore]

Lostgallifreyan wrote:

Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote

but wouldn't a high pass filter be useful on the input?

I presume you mean low pass ? But people like amps to repsond way beyong
the audio spectrum these days.


High pass to ground. Decoupling. Even though people like more bandwidth there
could be some use of that as prevention. If the caps are chosen to withstand
up to 100 KHz in a condition where that is the frequency of full-bore
feedback in an error condition then you pass anything above 50 KHz to ground,
you'd still have twice the bandwidth that we can hear before it fell off at
6db with a 1 pole RC filter, and the caps will be protected against worst
case feedback at any frequency. Phil says that any audio power amp with
standard specs is vulnerable though, and I'll have to think about that to
understand why, based on my assumption of 20 HZ (or DC) to 20 KHz, with input
impedance of 10K, maybe 100K. For the moment, I can't see why decoupling
unwanted HF range on the input or in op-amp gain stages won't protect as I
suggest above if the cap in that filter is able to handle that kind of
current at HF and even low RF.

Consider yourself unemployable in the pro-audio industry.

Graham

 

[Eeyore]

Phil Allison wrote:

"Eeysore the Congenital Fuckwit
Lostgallifreyan wrote:

but wouldn't a high pass filter be useful on the input?

I presume you mean low pass ? But people like amps to repsond way beyong
the
audio spectrum these days.

Balanced inputs fix it of course as in professional gear.

** More asinine drivel from the know nothing Stuffed Donkey.

I notice an absence of reasoning.

Probably because it's beyond you.

Graham

 

[Lostgallifreyan]

Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote in
news:49009DE9.ECC1E7C7@hotmail.com:

Consider yourself unemployable in the pro-audio industry.

I already do. Wouldn't have it any other way. What is your point?