Sign in with
Sign up | Sign in
Your question

Audio LP filter

Last response: in Home Audio
Share
Anonymous
December 20, 2004 9:41:16 PM

Archived from groups: rec.audio.pro (More info?)

I require a .47uF capacitor rated at 100V for use in a LC filter at the
output stage of a class D power amp. I'm assuming this capacitor needs to be
bipolar, but most bipolar caps don't seem to be rated past 50V for this
capacitance. Is there anything wrong with using two 50V 1uF caps in series
instead?

Thanks,

Dave.

PS. Sorry if this is a double post. I've waited 3 hours and haven't seen my
post show up yet.

More about : audio filter

December 21, 2004 1:37:32 PM

Archived from groups: rec.audio.pro (More info?)

Paul Stamler wrote:
> "David Grant" <NO_SPAM_PLEASE_jmd_2003@msn.com> wrote in message
> news:vqJxd.1222$jT5.1209@read1.cgocable.net...
> > I require a .47uF capacitor rated at 100V for use in a LC filter at
the
> > output stage of a class D power amp. I'm assuming this capacitor
needs to
> be
> > bipolar, but most bipolar caps don't seem to be rated past 50V for
this
> > capacitance. Is there anything wrong with using two 50V 1uF caps in
series
> > instead?
>
> You don't want to be using a bipolar electrolytic for this, or any
kind of
> electrolytic. You want something like a good-quality polypropylene
with low
> inductance. A B&C MKP3 series 0.47uF 630V (Digi-Key Part # BC1617-ND)
should
> take you out to 330kHz before it turns inductive, and bypassing with
a
> smaller cap will extend this. If you don't mind adding 0.1uF to the
total
> capacitance, an 0.1uF 400V Panasonic ECQP(U) series (Digi-Key #
P3488-ND)
> will take you out to 810kHz. Otherwise, a 10nF 160V polystyrene
(you'll have
> to dig for that) will take you out to 1MHz or so, but with a somewhat
higher
> impedance minimum.
>
> Peace,
> Paul

You need to be carefull about putting caps in parallel. In the "old
days" it was common practice to put a 0.01uF cap across an electrolytic
for bypassing in digital logic and other high speed circuits. This
worked well becasue the electrolytic cap is relativle low Q. These
days, if you put two high Q caps in parallel to try to lower the Z at
high frequencies, you may actually end up raising the Z at some
frequncies. Two caps in parallel can form a parallel resonant
circuit from the parasitc inductance of the bigger cap in parallel with
the C of the smaller cap. A parralle resonant circuit has a high Z at
resonance. I'd look up some refrences about bypassing hi speed logic
circuits. Paralleling high Q caps can be a problem.

Mark
Anonymous
December 21, 2004 2:30:34 PM

Archived from groups: rec.audio.pro (More info?)

David Grant <NO_SPAM_PLEASE_jmd_2003@msn.com> wrote:
>I require a .47uF capacitor rated at 100V for use in a LC filter at the
>output stage of a class D power amp. I'm assuming this capacitor needs to be
>bipolar, but most bipolar caps don't seem to be rated past 50V for this
>capacitance. Is there anything wrong with using two 50V 1uF caps in series
>instead?

Why not just use a film cap? A .47 mylar is not very big.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Related resources
Anonymous
December 21, 2004 8:11:26 PM

Archived from groups: rec.audio.pro (More info?)

"David Grant" <NO_SPAM_PLEASE_jmd_2003@msn.com> wrote in message
news:vqJxd.1222$jT5.1209@read1.cgocable.net...
> I require a .47uF capacitor rated at 100V for use in a LC filter at the
> output stage of a class D power amp. I'm assuming this capacitor needs to
be
> bipolar, but most bipolar caps don't seem to be rated past 50V for this
> capacitance. Is there anything wrong with using two 50V 1uF caps in series
> instead?

You don't want to be using a bipolar electrolytic for this, or any kind of
electrolytic. You want something like a good-quality polypropylene with low
inductance. A B&C MKP3 series 0.47uF 630V (Digi-Key Part # BC1617-ND) should
take you out to 330kHz before it turns inductive, and bypassing with a
smaller cap will extend this. If you don't mind adding 0.1uF to the total
capacitance, an 0.1uF 400V Panasonic ECQP(U) series (Digi-Key # P3488-ND)
will take you out to 810kHz. Otherwise, a 10nF 160V polystyrene (you'll have
to dig for that) will take you out to 1MHz or so, but with a somewhat higher
impedance minimum.

Peace,
Paul
Anonymous
December 22, 2004 11:31:44 AM

Archived from groups: rec.audio.pro (More info?)

"Mark" <makolber@yahoo.com> wrote in message
news:1103654252.149925.314980@z14g2000cwz.googlegroups.com...
>
> Paul Stamler wrote:
> > "David Grant" <NO_SPAM_PLEASE_jmd_2003@msn.com> wrote in message
> > news:vqJxd.1222$jT5.1209@read1.cgocable.net...
> > > I require a .47uF capacitor rated at 100V for use in a LC filter at
> the
> > > output stage of a class D power amp. I'm assuming this capacitor
> needs to
> > be
> > > bipolar, but most bipolar caps don't seem to be rated past 50V for
> this
> > > capacitance. Is there anything wrong with using two 50V 1uF caps in
> series
> > > instead?
> >
> > You don't want to be using a bipolar electrolytic for this, or any
> kind of
> > electrolytic. You want something like a good-quality polypropylene
> with low
> > inductance. A B&C MKP3 series 0.47uF 630V (Digi-Key Part # BC1617-ND)
> should
> > take you out to 330kHz before it turns inductive, and bypassing with
> a
> > smaller cap will extend this. If you don't mind adding 0.1uF to the
> total
> > capacitance, an 0.1uF 400V Panasonic ECQP(U) series (Digi-Key #
> P3488-ND)
> > will take you out to 810kHz. Otherwise, a 10nF 160V polystyrene
> (you'll have
> > to dig for that) will take you out to 1MHz or so, but with a somewhat
> higher
> > impedance minimum.
> >
> > Peace,
> > Paul
>
> You need to be carefull about putting caps in parallel. In the "old
> days" it was common practice to put a 0.01uF cap across an electrolytic
> for bypassing in digital logic and other high speed circuits. This
> worked well becasue the electrolytic cap is relativle low Q. These
> days, if you put two high Q caps in parallel to try to lower the Z at
> high frequencies, you may actually end up raising the Z at some
> frequncies. Two caps in parallel can form a parallel resonant
> circuit from the parasitc inductance of the bigger cap in parallel with
> the C of the smaller cap. A parralle resonant circuit has a high Z at
> resonance. I'd look up some refrences about bypassing hi speed logic
> circuits. Paralleling high Q caps can be a problem.

I did a simulation of the suggested 0.47 and 0.1 uF caps in parallel, and
found no problems up to 10MHz.

Peace,
Paul
December 22, 2004 1:15:04 PM

Archived from groups: rec.audio.pro (More info?)

>
> I did a simulation of the suggested 0.47 and 0.1 uF caps in parallel,
and
> found no problems up to 10MHz.
>
> Peace,
> Paul

Paul,

Did you include the parasitic inductance of the caps and the
connections in the simulation?

There will be a resonance that will cause an increase in the combined
Z.
The peak may be above 10 MHz depending upon the value of parasitic
inductance that you use in the simulation.

For example, 2 nH and 0.1 uF resonate at about 11 MHz.

Whats the total parasitic inducatance of the two caps and the
interconnections? It's easy to accumulate 2nH of inductance.

Mark
December 22, 2004 3:31:51 PM

Archived from groups: rec.audio.pro (More info?)

Thats about 1000nH or 1uH total. (seems high to me)
I agree you can ignore the interconnect inductance if the cap
inductance is really that high.
Is it really that high?

The ESR also seems VERY high.

What kind of cap are you talking about?
Maybe I've become accostomed to the small SMT caps.
Anyway, 1uH and 0.1uF resonante at about 0.5 MHz.

See my point?

Mark
Anonymous
December 22, 2004 9:42:03 PM

Archived from groups: rec.audio.pro (More info?)

"Mark" <makolber@yahoo.com> wrote in message
news:1103739304.880167.205830@f14g2000cwb.googlegroups.com...
>
> >
> > I did a simulation of the suggested 0.47 and 0.1 uF caps in parallel,
> and
> > found no problems up to 10MHz.

> Did you include the parasitic inductance of the caps and the
> connections in the simulation?
>
> There will be a resonance that will cause an increase in the combined
> Z.
> The peak may be above 10 MHz depending upon the value of parasitic
> inductance that you use in the simulation.
>
> For example, 2 nH and 0.1 uF resonate at about 11 MHz.
>
> Whats the total parasitic inducatance of the two caps and the
> interconnections? It's easy to accumulate 2nH of inductance.

I included the inductances of the caps but not, I confess, the
interconnections. The 0.47uF cap has an inductance of about 600nH and an ESR
of 4.45 ohms. The 0.1uF has 386nH and 4.8 ohms. (Error bars: 5% for
inductance, 0.15 ohms for ESR.) Since those inductances appear in series
with the interconnect inductances, I would expect the latter to be swamped.

Peace,
Paul
Anonymous
December 23, 2004 10:25:34 AM

Archived from groups: rec.audio.pro (More info?)

"Mark" <makolber@yahoo.com> wrote in message
news:1103747511.789586.159920@z14g2000cwz.googlegroups.com...
> Thats about 1000nH or 1uH total. (seems high to me)

The caps are in parallel, so the inductances would also be in parallel. But
onward.

> I agree you can ignore the interconnect inductance if the cap
> inductance is really that high.
> Is it really that high?

The resonance point on the 0.47uF cap by itself is 300kHz. On the 0.1uF cap,
810kHz. Unless I've misplaced the decimal point someplace, that works out to
599nH and 386nH respectively.

> The ESR also seems VERY high.

I agree; I was quite surprised to find that. All of the small (<1uF) film
caps I tested had ESRs ranging from about 4 to 6.4 ohms. I repeated the
tests quite a few times because I couldn't believe the results either. But
there they were. In the same tests, with the meter on the same scale, I
measured ESRs from electrolytics in the 0.16 - 0.5 ohm range. (0.15 was
about the noise limit of the measuring setup, so the lower numbers are
somewhat dubious, but the upper ones aren't.)

> What kind of cap are you talking about?

Panasonic ECQP(U) series in both cases -- 400V polypropylene. Digi-Key part
numbers P3496-ND and P3488-ND respectively.

> Maybe I've become accostomed to the small SMT caps.
> Anyway, 1uH and 0.1uF resonante at about 0.5 MHz.
>
> See my point?

Nope. *All* capacitors have a resonant point. But here's the difference with
and without the bypass cap:

0.47uF only:
Resonance: 300kHz
Z at 1 MHz: 5.54 ohms
Z at 10 MHz: 36.73 ohms

0.47uF || 0.1uF:
Resonance: ~630kHz
Z at 1 MHz: 2.67 ohms
Z at 10 MHz: 14.73 ohms

Peace,
Paul
December 23, 2004 12:44:48 PM

Archived from groups: rec.audio.pro (More info?)

I suggest you repeat the simulation with an ESR value of about 0.1 Ohm
or less.
With the lower ESR, the parallel resonance becomes more prominant.

The inductances ADD for the parrallel resonant circuit that is formed.

Are you measuring ESR or Z? They are not the same thing.
The ESR of those caps is probably well below 0.1 Ohm.


To Tony P, the OP is talking about a class D amp where the switching
harmonics can go above 1 MHz. I'm not saying you will HEAR that but
its nice to filter it off so your amp and speaker wires doesn't become
a radio transmitter.

Mark
December 23, 2004 9:14:27 PM

Archived from groups: rec.audio.pro (More info?)

The measured ESRs are at least an order of magnitude higher than I at
least would have expected, and because of that, the Qs are well below
1, and the expected impedance peak between the two resonances just
doesn't happen. I like to know more about why PP caps would behave so
badly. I still believe that with "good" caps, there WILL be a
potentially harmful impedance peak in between.

On Thu, 23 Dec 2004 07:25:34 GMT, "Paul Stamler"
<pstamlerhell@pobox.com> wrote:

>
>"Mark" <makolber@yahoo.com> wrote in message
>news:1103747511.789586.159920@z14g2000cwz.googlegroups.com...
>> Thats about 1000nH or 1uH total. (seems high to me)
>
>The caps are in parallel, so the inductances would also be in parallel. But
>onward.
>
>> I agree you can ignore the interconnect inductance if the cap
>> inductance is really that high.
>> Is it really that high?
>
>The resonance point on the 0.47uF cap by itself is 300kHz. On the 0.1uF cap,
>810kHz. Unless I've misplaced the decimal point someplace, that works out to
>599nH and 386nH respectively.
>
>> The ESR also seems VERY high.
>
>I agree; I was quite surprised to find that. All of the small (<1uF) film
>caps I tested had ESRs ranging from about 4 to 6.4 ohms. I repeated the
>tests quite a few times because I couldn't believe the results either. But
>there they were. In the same tests, with the meter on the same scale, I
>measured ESRs from electrolytics in the 0.16 - 0.5 ohm range. (0.15 was
>about the noise limit of the measuring setup, so the lower numbers are
>somewhat dubious, but the upper ones aren't.)
>
>> What kind of cap are you talking about?
>
>Panasonic ECQP(U) series in both cases -- 400V polypropylene. Digi-Key part
>numbers P3496-ND and P3488-ND respectively.
>
>> Maybe I've become accostomed to the small SMT caps.
>> Anyway, 1uH and 0.1uF resonante at about 0.5 MHz.
>>
>> See my point?
>
>Nope. *All* capacitors have a resonant point. But here's the difference with
>and without the bypass cap:
>
>0.47uF only:
>Resonance: 300kHz
>Z at 1 MHz: 5.54 ohms
>Z at 10 MHz: 36.73 ohms
>
>0.47uF || 0.1uF:
>Resonance: ~630kHz
>Z at 1 MHz: 2.67 ohms
>Z at 10 MHz: 14.73 ohms
>
>Peace,
>Paul
>

Tony (remove the "_" to reply by email)
Anonymous
December 23, 2004 9:14:28 PM

Archived from groups: rec.audio.pro (More info?)

"Tony" <tony_roe@tpg.com.au> wrote in message
news:t4vks059lvreto28p1484v20t5qig24j0s@4ax.com...
> The measured ESRs are at least an order of magnitude higher than I at
> least would have expected, and because of that, the Qs are well below
> 1, and the expected impedance peak between the two resonances just
> doesn't happen. I like to know more about why PP caps would behave so
> badly. I still believe that with "good" caps, there WILL be a
> potentially harmful impedance peak in between.

I did a simulation with ESRs an order of magnitude smaller, and yes, the
peak shows up right where it should. But I have to say that I measured a
bunch of small film capacitors, ranging from the Panasonics mentioned to
some B&Cs to a MultiCap to a generic polystyrene to a stacked-film Wima
polycarbonate, and they all showed high ESRs. The only film caps that didn't
were large tubulars such as the Solen Fastcap series, and they had high
enough inductances for their resonances to fall in the 50-100 range.

Either my measurement technique is somehow faulty (always possible, of
course), or film caps show behavior different from what we all expected.

Peace,
Paul
Anonymous
December 23, 2004 10:32:58 PM

Archived from groups: rec.audio.pro (More info?)

"Paul Stamler" <pstamlerhell@pobox.com> wrote in message
news:2PYxd.1135999$Gx4.440003@bgtnsc04-news.ops.worldnet.att.net...
> A B&C MKP3 series 0.47uF 630V (Digi-Key Part # BC1617-ND) should
> take you out to 330kHz before it turns inductive, and bypassing with a
> smaller cap will extend this. If you don't mind adding 0.1uF to the total
> capacitance, an 0.1uF 400V Panasonic ECQP(U) series (Digi-Key # P3488-ND)
> will take you out to 810kHz. Otherwise, a 10nF 160V polystyrene (you'll
have
> to dig for that) will take you out to 1MHz or so, but with a somewhat
higher
> impedance minimum.

I guess you missed the "Audio" part of the header.

TonyP.
Anonymous
December 23, 2004 10:32:59 PM

Archived from groups: rec.audio.pro (More info?)

"TonyP" <TonyP@optus.net.com.au> wrote in message
news:41ca82e8$0$4534$afc38c87@news.optusnet.com.au...
>
> "Paul Stamler" <pstamlerhell@pobox.com> wrote in message
> news:2PYxd.1135999$Gx4.440003@bgtnsc04-news.ops.worldnet.att.net...
> > A B&C MKP3 series 0.47uF 630V (Digi-Key Part # BC1617-ND) should
> > take you out to 330kHz before it turns inductive, and bypassing with a
> > smaller cap will extend this. If you don't mind adding 0.1uF to the
total
> > capacitance, an 0.1uF 400V Panasonic ECQP(U) series (Digi-Key #
P3488-ND)
> > will take you out to 810kHz. Otherwise, a 10nF 160V polystyrene (you'll
> have
> > to dig for that) will take you out to 1MHz or so, but with a somewhat
> higher
> > impedance minimum.
>
> I guess you missed the "Audio" part of the header.

No; did you miss the original poster's note that this is a low-pass filter
for a switching amplifier? In that context, behavior at high frequencies
matters.

Peace,
Paul
Anonymous
December 24, 2004 10:18:02 AM

Archived from groups: rec.audio.pro (More info?)

"Mark" <makolber@yahoo.com> wrote in message
news:1103823888.019532.110680@c13g2000cwb.googlegroups.com...
> I suggest you repeat the simulation with an ESR value of about 0.1 Ohm
> or less.
> With the lower ESR, the parallel resonance becomes more prominant.
>
> The inductances ADD for the parrallel resonant circuit that is formed.
>
> Are you measuring ESR or Z? They are not the same thing.
> The ESR of those caps is probably well below 0.1 Ohm.

At resonance, for a single cap, they ought to be, no? Or am I wrong? In any
case, yes, I'm measuring the minimum Z and calling that ESR. What other
impedtance is there in a single not-very-leaky cap besides the L, C and ESR?

Peace,
Paul
December 24, 2004 11:24:08 AM

Archived from groups: rec.audio.pro (More info?)

Paul Stamler wrote:
> "Mark" <makolber@yahoo.com> wrote in message
> news:1103823888.019532.110680@c13g2000cwb.googlegroups.com...
> > I suggest you repeat the simulation with an ESR value of about 0.1
Ohm
> > or less.
> > With the lower ESR, the parallel resonance becomes more prominant.
> >
> > The inductances ADD for the parrallel resonant circuit that is
formed.
> >
> > Are you measuring ESR or Z? They are not the same thing.
> > The ESR of those caps is probably well below 0.1 Ohm.
>
> At resonance, for a single cap, they ought to be, no? Or am I wrong?
In any
> case, yes, I'm measuring the minimum Z and calling that ESR. What
other
> impedtance is there in a single not-very-leaky cap besides the L, C
and ESR?
>
> Peace,
> Paul


Well thats a good quesiton...

Since you are getting much higher than expected results for ESR, I will
have to say the min Z at series resonance is not the same as the ESR.
A good poly cap should have a very low ESR. I imagine at 0.1uF it may
be difficult to measure ESR since the Z of the cap is relativley high.
I don't know exactly what the error is in your measument method but the
results do not agree with what one expects for poly caps.

Mark
December 24, 2004 6:59:25 PM

Archived from groups: rec.audio.pro (More info?)

Paul Stamler wrote:
> "Mark" <makolber@yahoo.com> wrote in message
> news:1103905448.942917.278620@c13g2000cwb.googlegroups.com...
> >
> > Paul Stamler wrote:
> > > "Mark" <makolber@yahoo.com> wrote in message
> > > news:1103823888.019532.110680@c13g2000cwb.googlegroups.com...
> > > > I suggest you repeat the simulation with an ESR value of about
0.1
> > Ohm
> > > > or less.
> > > > With the lower ESR, the parallel resonance becomes more
prominant.
> > > >
> > > > The inductances ADD for the parrallel resonant circuit that is
> > formed.
> > > >
> > > > Are you measuring ESR or Z? They are not the same thing.
> > > > The ESR of those caps is probably well below 0.1 Ohm.
> > >
> > > At resonance, for a single cap, they ought to be, no? Or am I
wrong?
> > In any
> > > case, yes, I'm measuring the minimum Z and calling that ESR. What
> > other
> > > impedtance is there in a single not-very-leaky cap besides the L,
C
> > and ESR?
> > >
> > Well thats a good quesiton...
> >
> > Since you are getting much higher than expected results for ESR, I
will
> > have to say the min Z at series resonance is not the same as the
ESR.
> > A good poly cap should have a very low ESR. I imagine at 0.1uF it
may
> > be difficult to measure ESR since the Z of the cap is relativley
high.
> > I don't know exactly what the error is in your measument method but
the
> > results do not agree with what one expects for poly caps.
>
> Well, I measure much lower ESRs on some large electrolytics, so the
setup
> seems capable of measuring <1 ohm if it's there to be measured. I
confess
> I'm quite surprised by my results, and have been wondering about
> experimental error, but I go back and forth between the film caps and
the
> electrolytics, and the setup works just fine on the latter.
>
> Peace,
> Paul

That may be because they are a larger value and therefore a lower Z
also. It's hard to measure a low ESR in the series with a high Z.
It's easier to measure a low Z if it is in series with a larger low Z
cap.

A poly cap will typically have a lower ESR compared to an electrolytic
but at the same time the poly will have a higher Z becasue it is
(typically) a smaller C value.
What method are you using to measure ESR?

Mark
Anonymous
December 24, 2004 8:38:28 PM

Archived from groups: rec.audio.pro (More info?)

"Paul Stamler" <pstamlerhell@pobox.com> wrote in message
news:KoPyd.26415$uM5.2348@bgtnsc05-news.ops.worldnet.att.net...
>
> "Mark" <makolber@yahoo.com> wrote in message
> news:1103823888.019532.110680@c13g2000cwb.googlegroups.com...
> > I suggest you repeat the simulation with an ESR value of about 0.1 Ohm
> > or less.
> > With the lower ESR, the parallel resonance becomes more prominant.
> >
> > The inductances ADD for the parrallel resonant circuit that is formed.
> >
> > Are you measuring ESR or Z? They are not the same thing.
> > The ESR of those caps is probably well below 0.1 Ohm.
>
> At resonance, for a single cap, they ought to be, no? Or am I wrong? In
any
> case, yes, I'm measuring the minimum Z and calling that ESR. What other
> impedtance is there in a single not-very-leaky cap besides the L, C and
ESR?

Following up on which, I ran a simulation of the 0.47uF cap with its ESR set
to 0.1 ohm, and the Z dipped at the resonance to 0.1 ohm. The real capacitor
did not, however, when measured, leading me to suspect that I am measuring
the real ESR, and that it's high.

Peace,
Paul
Anonymous
December 24, 2004 8:41:59 PM

Archived from groups: rec.audio.pro (More info?)

"Mark" <makolber@yahoo.com> wrote in message
news:1103905448.942917.278620@c13g2000cwb.googlegroups.com...
>
> Paul Stamler wrote:
> > "Mark" <makolber@yahoo.com> wrote in message
> > news:1103823888.019532.110680@c13g2000cwb.googlegroups.com...
> > > I suggest you repeat the simulation with an ESR value of about 0.1
> Ohm
> > > or less.
> > > With the lower ESR, the parallel resonance becomes more prominant.
> > >
> > > The inductances ADD for the parrallel resonant circuit that is
> formed.
> > >
> > > Are you measuring ESR or Z? They are not the same thing.
> > > The ESR of those caps is probably well below 0.1 Ohm.
> >
> > At resonance, for a single cap, they ought to be, no? Or am I wrong?
> In any
> > case, yes, I'm measuring the minimum Z and calling that ESR. What
> other
> > impedtance is there in a single not-very-leaky cap besides the L, C
> and ESR?
> >
> Well thats a good quesiton...
>
> Since you are getting much higher than expected results for ESR, I will
> have to say the min Z at series resonance is not the same as the ESR.
> A good poly cap should have a very low ESR. I imagine at 0.1uF it may
> be difficult to measure ESR since the Z of the cap is relativley high.
> I don't know exactly what the error is in your measument method but the
> results do not agree with what one expects for poly caps.

Well, I measure much lower ESRs on some large electrolytics, so the setup
seems capable of measuring <1 ohm if it's there to be measured. I confess
I'm quite surprised by my results, and have been wondering about
experimental error, but I go back and forth between the film caps and the
electrolytics, and the setup works just fine on the latter.

Peace,
Paul
Anonymous
December 24, 2004 8:42:00 PM

Archived from groups: rec.audio.pro (More info?)

Paul Stamler <pstamlerhell@pobox.com> wrote:
>
>Well, I measure much lower ESRs on some large electrolytics, so the setup
>seems capable of measuring <1 ohm if it's there to be measured. I confess
>I'm quite surprised by my results, and have been wondering about
>experimental error, but I go back and forth between the film caps and the
>electrolytics, and the setup works just fine on the latter.

At the risk of making that Aussie guy explode again, may I suggest that
you try the Dick Smith Electronics ESR meter kit? It's not very expensive
and it will give you solid and repeatable numbers for caps on the bench.

What is more interesting is that it will give good numbers for caps in circuit
which makes it a great diagnostic tool.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Anonymous
December 25, 2004 3:11:01 AM

Archived from groups: rec.audio.pro (More info?)

"Paul Stamler" <pstamlerhell@pobox.com> wrote in message
news:3uDyd.1148841$Gx4.63831@bgtnsc04-news.ops.worldnet.att.net...
> No; did you miss the original poster's note that this is a low-pass filter
> for a switching amplifier? In that context, behavior at high frequencies
> matters.

Yes I missed it, and you are correct.

TonyP.
Anonymous
December 25, 2004 10:50:02 AM

Archived from groups: rec.audio.pro (More info?)

> That may be because they are a larger value and therefore a lower Z
> also. It's hard to measure a low ESR in the series with a high Z.
> It's easier to measure a low Z if it is in series with a larger low Z
> cap.

Mmmm...not sure that's the case. As I said in a separate post, running a
simulation of a poly cap with the same inductance as the one I measured but
the ESR set to 0.1 ohm, the impedance should dip below 1 ohm from about 200
kHz to about 450 kHz, something which should be easily measurable. But
that's not anything like what happened; instead, I saw an impedance that
clearly bottomed out in the 6 ohm range.

> A poly cap will typically have a lower ESR compared to an electrolytic
> but at the same time the poly will have a higher Z becasue it is
> (typically) a smaller C value.

> What method are you using to measure ESR?

If this ASCII art makes it through okay:

Oscillator----resistor---|
| | ----------|
| Cap AC VTVM
| | ----------|
Gnd--------------------|

Peace,
Paul
!