How do caps work?

[mike]

In a recent thread here and many times in the past, I've seen discussion
about adding a BIG cap to the power supply of a car stereo.

Can someone explain to me how that works?


There are many ways to analyze this. I chose a time-domain approximation
because it makes the "back of the envelope" math easier.
Assumptions:

I used very round numbers to make the math easier. It's linear math, so
it should scale.

1500W = 100A at 15V (peak power)

What happens if we put a 100A transient load on our battery?
My battery is 8 years old, so measuring it won't help.
But for every 10 milliohms of resistance, internal resistance of the
battery plus all the wire and connection resistances,
we get a volt of transient drop at 100A.

I've also read that there's an electrochemical lag inside the battery,
but never seen any actual measured numbers.

So, the voltage at the amp is bouncing around with the load. What are some
typical numbers for this?

when we add the cap, we add additional series resistance in the form
of more wire and connections to the cap. That makes things worse,
even if we do know what we're doing and put the cap in the optimum wiring
configuration.

But the cap and the connections have internal resistance too. Every 10
mOhm gives us a drop
of 1V at 100A, then there's the decay of 1V every 10mS for a 1F cap.

Let's not forget that there's also a hundred amps (average) of AC coming
out of
the alternator into all these resistances.

The battery voltage is bouncing all around with engine speed and load.
The amplifier has a "common mode rejection ratio" that describes its
immunity to these variations. That same immunity mitigates the effect
of the cap.
One possible way of describing it is, "How much more undistorted power
output
can I get with the cap installed?"

Given a well-designed amp with good CMRR, the output should be
independent of
power supply voltage until the amp saturates at the lower power supply
voltage.

What's a typical speaker impedance? If it's one ohm, and you have a bridged
amp off of 12V, you can't get much more than 10A into a 1Ohm load.
That's 100W peak power. To get more than that, you have to up-convert
the 12V
to something higher. That up-conversion should make the system pretty
much immune
to what's going on on the 12V side???

Then, layered on top of all this is the liklihood that my stereo gets
installed
by a high-school dropout who's never heard of a volt or an amp. And the
installation
is more dependent on dealer profit, aesthetics and fixed postioning of
the automobile components
than it is on performance.

My questions are these:

1) In a well-installed car stereo, what are the typical resistances in the
wiring, connections, batteries etc.?

2) What is a typical speaker impedance for a high power system?

3) If I add a big cap, What improvement would I HEAR under what conditions?

4) If I hooked up an oscilloscope to the power supply at the amp and to
it's audio output,
what improvement would I SEE under what conditions if I added a cap?

5) Why would lights dim? IF the lights connect to the battery.
And the stereo connects to the battery.
And the sense point for the alternator/regulator connects to the battery.
And the alternator is big enough.
??

6) What percentage of people actually care about the technical merits of
their
car stero...as long as their ride is "fly" and gets 'em laid? ;-)

mike

[GregS]

In article <FXBih.5658$od6.1148@trnddc04>, wrote:
>In a recent thread here and many times in the past, I've seen discussion
>about adding a BIG cap to the power supply of a car stereo.
>
>Can someone explain to me how that works?
>
>
>There are many ways to analyze this. I chose a time-domain approximation
>because it makes the "back of the envelope" math easier.
>Assumptions:
>
>I used very round numbers to make the math easier. It's linear math, so
>it should scale.
>
>1500W = 100A at 15V (peak power)
>
>What happens if we put a 100A transient load on our battery?
>My battery is 8 years old, so measuring it won't help.
>But for every 10 milliohms of resistance, internal resistance of the
>battery plus all the wire and connection resistances,
>we get a volt of transient drop at 100A.
>
>I've also read that there's an electrochemical lag inside the battery,
>but never seen any actual measured numbers.
>
>So, the voltage at the amp is bouncing around with the load. What are some
>typical numbers for this?
>
>when we add the cap, we add additional series resistance in the form
>of more wire and connections to the cap. That makes things worse,
>even if we do know what we're doing and put the cap in the optimum wiring
>configuration.

You don't have to add more series resistance. If the cap is in parallel with
the supply it simply provides a kind of node at that point. When the cap draws
current, it does cut down on the available current out of the alternator,
but if the alt, starts to crap out, then the current comes back out of the
alternator. You could add 4 caps, all with their own wiring. In star
configuration, all the wires from each cap come to the same summing point.
The cap should be right at the power amp. Sometimes you see the bussing
of multiple caps. This is not the best method, but with heavy straps, the
resistance is minimal.


>But the cap and the connections have internal resistance too. Every 10
>mOhm gives us a drop
>of 1V at 100A, then there's the decay of 1V every 10mS for a 1F cap.
>
>Let's not forget that there's also a hundred amps (average) of AC coming
>out of
>the alternator into all these resistances.

Only if the RPM's are up. At 600 RPM the alt can only furnish a few
amps.

>The battery voltage is bouncing all around with engine speed and load.
>The amplifier has a "common mode rejection ratio" that describes its
>immunity to these variations. That same immunity mitigates the effect
>of the cap.
>One possible way of describing it is, "How much more undistorted power
>output
>can I get with the cap installed?"

it really about peak power. The cap will discharge and the average power
is not going to be any higher than with just the alternator.

>Given a well-designed amp with good CMRR, the output should be
>independent of
>power supply voltage until the amp saturates at the lower power supply
>voltage.
>
>What's a typical speaker impedance? If it's one ohm, and you have a bridged
>amp off of 12V, you can't get much more than 10A into a 1Ohm load.
>That's 100W peak power. To get more than that, you have to up-convert
>the 12V
>to something higher. That up-conversion should make the system pretty
>much immune
>to what's going on on the 12V side???

To a certain extent.

>Then, layered on top of all this is the liklihood that my stereo gets
>installed
>by a high-school dropout who's never heard of a volt or an amp. And the
>installation
>is more dependent on dealer profit, aesthetics and fixed postioning of
>the automobile components
>than it is on performance.
>
>My questions are these:
>
>1) In a well-installed car stereo, what are the typical resistances in the
>wiring, connections, batteries etc.?
>
>2) What is a typical speaker impedance for a high power system?
>
>3) If I add a big cap, What improvement would I HEAR under what conditions?
>
>4) If I hooked up an oscilloscope to the power supply at the amp and to
>it's audio output,
>what improvement would I SEE under what conditions if I added a cap?
>
>5) Why would lights dim? IF the lights connect to the battery.
>And the stereo connects to the battery.
>And the sense point for the alternator/regulator connects to the battery.
>And the alternator is big enough.
>??
>
>6) What percentage of people actually care about the technical merits of
>their
>car stero...as long as their ride is "fly" and gets 'em laid? ;-)

The caps provides better peak power, and also helps control lamp flicker.
That about it. How about the FAQ. There should be a good
explanation in there.

greg

[KU40]

what I can make educated guesses about:

1) resistance in wiring depends greatly on size (when delivering a
load) as well as the metal used to make the wire. Though typically
there isn't a lot. I'd say over an average 15 foot run of 4 guage
you'll lose .1-.2 volts. Connections........probably in the millivolt
range. batteries I don't know.

2) Generally interior speakers are 4 ohms, subs range from 1-4 ohms.
Most large subs these days are Dual 1, 2, or 4 ohms.

3) You will hear nothing. by the time you hear the bass note, the cap
will be drained and will be trying to recharge again.

4) This one I don't know. I would imagine/guess the result would show
a slightly higher voltage on both input and output immediately after the
signal is sent (relative to the non-cap system), but then after that it
would be the same. Would just take .01 seconds longer to get to the
sagged voltage.

5) It just has to do with lower voltage. Just because the regulator
sensor is there doesn't mean it's going to make the alternator put out
more power. If the RPMs aren't high enough the alternator won't put
out the power the regulator wants, no matter how hard it begs.

6) I do. And I've never seen nor heard of anybody getting laid
because of their stereo.

those are my ideas. feel free to correct/expand.


--
KU40

[mike]

> 6) I do. And I've never seen nor heard of anybody getting laid
> because of their stereo.

You should get out more.
In my universe, virtually all animal competition and display of prowess
is directly related to getting laid. Peacock, bullfrog, little guy with
the pimped out ride...ain't that why they call it pimped??? You really
think "SHE" would be in the car if it were a '75 Pinto with one
headlight out?

Back in the 60's I had a homebrew stereo record player in my car.
Sounded GREAT compared to the alternatives at the time.
Problem was that it was small and the 12" vinyl hung way over the sides.
Only thing you could do was park and sit it on the front seat.
Now, there's nowhere to sit....except maybe the back seat...hand me that
Johnny Mathis record....sometimes it's good when the record skips ;-)

I REALLY miss the good old days...
mike

[mike]

> The caps provides better peak power, and also helps control lamp flicker.
> That about it. How about the FAQ. There should be a good
> explanation in there.
>
> greg
Well, DUH! Thanks for reminding me to read the FAQ.
But it didn't help much.
It did raise another question.
I never see any mention of folded horn loaded speakers for car use.
You can get big volume with low power. My Klipsch are 105dB/W,
but they won't fit in the pickup.
If you're in the business
of selling amps and caps and other power stuff, horns would be
counterproductive. Sounds like a natural application for the
small slot between the seat and rear of the cab.
Are there technical reasons for not using horns?

I'm having visions of removing the window and using the whole
pickup bed for the horn.
mike

[bassman07]

i wouldnt add a cap. just do batteries. a cap is puts more strain and is
more for your alternator to charge. it would rather charge a battery at
constant.


--
bassman07

[D.Kreft]

On Dec 21, 4:44 pm, bassman07
> wrote:

> i wouldnt add a cap. just do batteries. a cap is puts more strain and is
> more for your alternator to charge.

And batteries are *not* a load on the alternator?????

-dan

[KU40]

when you say horns, are you talking for full range use or in a sub
enclosure? I've seen and heard of people doing similar sub boxes in
their vehicles. full range wouldn't be so good though. The big shots
say to have no full range speakers behind you because that messes up
the imaging, which is "supposed" to sound like a live event or whatnot,
with the music in front of you.

and that's why I said stereo in relation to not helping get laid.
Ladies don't care about stereos. they care about the car itself a
million times more.


--
KU40

[mike]

KU40 wrote:
> when you say horns, are you talking for full range use or in a sub
> enclosure? I've seen and heard of people doing similar sub boxes in
> their vehicles. full range wouldn't be so good though. The big shots
> say to have no full range speakers behind you because that messes up
> the imaging, which is "supposed" to sound like a live event or whatnot,
> with the music in front of you.
>
> and that's why I said stereo in relation to not helping get laid.
> Ladies don't care about stereos. they care about the car itself a
> million times more.
>
>
I never understood the whole thing about imaging and sound stage.
At tolerable listening levels in a pickup truck, the signal to noise
ratio has gotta be 20 dB or so. Where the sound is coming from is a
minor concern. I would like to hear some of the bass tho.

Problem with a full range horn is that you can't fold it.
If you can't fold it, it won't fit.

Somewhere around here I've got an article on horny math.
mike

[Matt Ion]

mike wrote:
>
>> The caps provides better peak power, and also helps control lamp
>> flicker. That about it. How about the FAQ. There should be a good
>> explanation in there.
>>
>> greg
>
> Well, DUH! Thanks for reminding me to read the FAQ.
> But it didn't help much.
> It did raise another question.
> I never see any mention of folded horn loaded speakers for car use.
> You can get big volume with low power. My Klipsch are 105dB/W,
> but they won't fit in the pickup.
> If you're in the business
> of selling amps and caps and other power stuff, horns would be
> counterproductive. Sounds like a natural application for the
> small slot between the seat and rear of the cab.
> Are there technical reasons for not using horns?
>
> I'm having visions of removing the window and using the whole
> pickup bed for the horn.

Someone a few years ago (coulda been Phoenix Gold) made a set of compression
horn drivers for the upper-mids-and-above designed to be mounted under the dash.
In addition to covering pretty much the full 400Hz-22kHz range, they helped
imaging by increasing the total path length to the drivers, thus reducing the
DIFFERENCE in distance between left and right sound paths. Dunno if anyone is
still making such setups for cars...

Of course, you've hit on the main problem with using horn-loaded SUBS: the sheer
size. Phoenix Gold also marketed a couple subs based on the Servodrive design
several years ago, with some limited success....

[GregS]

In article <yNEih.3950$Iy5.3581@trnddc01>, wrote:
>
>> The caps provides better peak power, and also helps control lamp flicker.
>> That about it. How about the FAQ. There should be a good
>> explanation in there.
>>
>> greg
>Well, DUH! Thanks for reminding me to read the FAQ.
>But it didn't help much.
>It did raise another question.
>I never see any mention of folded horn loaded speakers for car use.
>You can get big volume with low power. My Klipsch are 105dB/W,
>but they won't fit in the pickup.
>If you're in the business
>of selling amps and caps and other power stuff, horns would be
>counterproductive. Sounds like a natural application for the
>small slot between the seat and rear of the cab.
>Are there technical reasons for not using horns?
>
>I'm having visions of removing the window and using the whole
>pickup bed for the horn.
>mike
>

its just too big to fit in the cab. You would be inside the horn, and being
inside I really don't know the results. If you can't do low bass, then there
isn't much point. Horns rapidly cut off below 40 Hz, and even that
usually requires the right driver, a 16 foot length, and a mouth about
12 foot square. Horns automatically add 10 dB to a drivers sensitivity,
playing twice as loud to the ear, and requiring on 1/10 the power
to play at the same volume. I think the old AR3 speaker played at 100 watts
what the Klipshorn would do with 1 watt.

greg

[Matt Ion]

GregS wrote:
> In article <yNEih.3950$Iy5.3581@trnddc01>, wrote:
>
>>>The caps provides better peak power, and also helps control lamp flicker.
>>>That about it. How about the FAQ. There should be a good
>>>explanation in there.
>>>
>>>greg
>>
>>Well, DUH! Thanks for reminding me to read the FAQ.
>>But it didn't help much.
>>It did raise another question.
>>I never see any mention of folded horn loaded speakers for car use.
>>You can get big volume with low power. My Klipsch are 105dB/W,
>>but they won't fit in the pickup.
>>If you're in the business
>>of selling amps and caps and other power stuff, horns would be
>>counterproductive. Sounds like a natural application for the
>>small slot between the seat and rear of the cab.
>>Are there technical reasons for not using horns?
>>
>>I'm having visions of removing the window and using the whole
>>pickup bed for the horn.
>>mike
>>
>
>
> its just too big to fit in the cab. You would be inside the horn, and being
> inside I really don't know the results. If you can't do low bass, then there
> isn't much point. Horns rapidly cut off below 40 Hz, and even that
> usually requires the right driver, a 16 foot length, and a mouth about
> 12 foot square. Horns automatically add 10 dB to a drivers sensitivity,
> playing twice as loud to the ear, and requiring on 1/10 the power
> to play at the same volume. I think the old AR3 speaker played at 100 watts
> what the Klipshorn would do with 1 watt.

Funny thing is, speaking of being "inside the horn", some car designs themselve
ARE horn-like and do affect the bass. My single 10" JL sub thumps my Accord
hatchback far more when it's placed all the way to the rear than anywhere else
in the car.

Back in the day, the high school burnouts with the '81 Trans Ams got
(relatively) large amounts of bass out of a simple pair of 6x9s in their rear
deck, because the curved rear window effectively horn-loaded them.

[mike]

GregS wrote:
> In article <yNEih.3950$Iy5.3581@trnddc01>, wrote:
>
>>>The caps provides better peak power, and also helps control lamp flicker.
>>>That about it. How about the FAQ. There should be a good
>>>explanation in there.
>>>
>>>greg
>>
>>Well, DUH! Thanks for reminding me to read the FAQ.
>>But it didn't help much.
>>It did raise another question.
>>I never see any mention of folded horn loaded speakers for car use.
>>You can get big volume with low power. My Klipsch are 105dB/W,
>>but they won't fit in the pickup.
>>If you're in the business
>>of selling amps and caps and other power stuff, horns would be
>>counterproductive. Sounds like a natural application for the
>>small slot between the seat and rear of the cab.
>>Are there technical reasons for not using horns?
>>
>>I'm having visions of removing the window and using the whole
>>pickup bed for the horn.
>>mike
>>
>
>
> its just too big to fit in the cab. You would be inside the horn, and being
> inside I really don't know the results. If you can't do low bass, then there
> isn't much point.

Maybe you have to open the windows ;-)
I haven't looked at the math since 1967, but as I recall, if you take
whatever driver size you have and run an exponential out as far as you
have room, don't you still get coupling improvement over the bare
driver? At cutof, the driver will unload, so you still have to have
some volume behind it and some active compensation. Would that be any
worse than having the same driver without the horn? I just don't have
room for anything thicker than about 5" behind the seat.

Horns rapidly cut off below 40 Hz, and even that
> usually requires the right driver, a 16 foot length, and a mouth about
> 12 foot square. Horns automatically add 10 dB to a drivers sensitivity,
> playing twice as loud to the ear, and requiring on 1/10 the power
> to play at the same volume. I think the old AR3 speaker played at 100 watts
> what the Klipshorn would do with 1 watt.

I picked up a pair of Magneplanar ribbon speakers at a garage sale.
Was a $20 pair of "room dividers".
I really liked the way they sounded, but I had to put them far away from
the wall or they sounded like crap. So they actually took up more
living space than my K-horns. And I didn't like the smell that came off
my poor little 100W amplifier at even moderate volume. Went back to the
horns.
mike
>
> greg
>

[Captain_Howdy]

High school burnouts with the '81 Trans Ams? You must of been the poor kid at
school or does this remark go a little deeper then that?

>
>Back in the day, the high school burnouts with the '81 Trans Ams got
>(relatively) large amounts of bass out of a simple pair of 6x9s in their rear
>deck, because the curved rear window effectively horn-loaded them.

[bob wald]

caps are a load on your alt. LMAO....
thats about the funnyest thing i ever heard....
if ya alt. cant supply a tiny bit of power every 3-4seconds....
do you think ya cap runs down to nothing n then goes ok ill charge up
now..lol
i'd guess under a extra 4-5amps will power most caps. 1-2 farad.
thats continious charging.
as the caps discharging.its charging back.
only discharges when more powers needed.
i'd guess 90% of the time caps discharge less that 33%.
but i'm no expert like all you guys. laughing here.
if you discharging more. you power system suckssss..you finished anyway.

[Atomic Fusion]

GregS Wrote:
>
> its just too big to fit in the cab. You would be inside the horn, and
> being
> inside I really don't know the results. If you can't do low bass, then
> there
> isn't much point. Horns rapidly cut off below 40 Hz, and even that
> usually requires the right driver, a 16 foot length, and a mouth about
> 12 foot square. Horns automatically add 10 dB to a drivers
> sensitivity,
> playing twice as loud to the ear, and requiring on 1/10 the power
> to play at the same volume. I think the old AR3 speaker played at 100
> watts
> what the Klipshorn would do with 1 watt.
>
> greg

I just need to correct some of this information, as most of it is
generalized or false.

A) Horns do not "automatically add 10 db" to a driver's sensitivity.
The increased effeciency varies *widely* depending on the type of flare
used, along with other design factors. It is also not uniform over the
frequency spectrum. Depending on horn design, you can get anywhere from
3-15db increase in sensitivity, with varying improvement over different
frequencies.

B) Horns do not rapidly cut below 40 Hz unless they are designed to do
so. A horn with a throat 16 feet long could play to below 20 Hz.

C) The mouth area required by a horn to play to a given frequency
depends completely on the space which it is driving. A horn with a 30
Hz flare, operating in the middle of an open field, would need
something in the neihbourhood of 30 square feet of mouth area. To get
the same performance in the corner of a room, you'd need around 7-8
square feet of mouth area. In the volume of a car, it could drop (in
theory) to 3-4 square feet, although the real world results would not
be easily predictable.

Adam


--
Atomic Fusion

[eezip]

mike;661966 Wrote:
>
> My questions are these:
>
> 1) In a well-installed car stereo, what are the typical resistances in
> the
> wiring, connections, batteries etc.?
>
> 2) What is a typical speaker impedance for a high power system?
>
> 3) If I add a big cap, What improvement would I HEAR under what
> conditions?
>
> 4) If I hooked up an oscilloscope to the power supply at the amp and
> to
> it's audio output,
> what improvement would I SEE under what conditions if I added a cap?
>
> 5) Why would lights dim? IF the lights connect to the battery.
> And the stereo connects to the battery.
> And the sense point for the alternator/regulator connects to the
> battery.
> And the alternator is big enough.
> ??
>
> 6) What percentage of people actually care about the technical merits
> of
> their
> car stero...as long as their ride is "fly" and gets 'em laid? ;-)
>
> mike

Mike,

You didn't post any time-domain approximation.

CMRR doesn't affect the power supply in any way. CMRR is the rejection
of a signal that is common to 2 signals, for instance, noise picked up
in the ground and hot of an RCA is rejected by the CMRR of a
differential amplifier. The term you mean is PSRR, or power supply
rejection ratio.

The 'voltage upconversion' you're talking about is the power supply.
All car amps I've seen use a push-pull SMPS. Push-pull supplies can be
regulated or unregulated. If they're regulated, the output voltage is
largely unaffected by fluctuations of the input voltage - unregulated
supplies have a fixed boost and the output moves just as the input
moves or the load changes. Most car audio amplifier supplies I've seen
were unregulated for reasons that hopefully will be more clear later in
my response. Regulated supplies also have a specific loop gain which
affects the power supplies' ability to react to changes in input or
output. With regulated supplies this is also a factor.

1. I think 'this chart'
(http://www.interfacebus.com/Copper_Wire_AWG_SIze.html) will help you
out. Keep in mind these values are for a specific temperature and type
of cable jacket. This can vary with different winding techniques of the
strands in the cable as well.

2. As above, 1-4ohm is typical for a car subwoofer.

3. You won't hear any difference. Adding more capacitance will help
continuous power a bit, because CV will be greater, but the voltage
rails won't change so peak power will remain the same. Audio has high
crest factors, and thus peak power is actually relevant and meaningful
for audio. RMS power tells you more about how robust the amp is, but
peak power won't change if you add more capacitance. And that is what
will tell you more about audible difference between amplifiers.

4. Same as above. You'll see high peaks before clipping if the input
signal is a sine wave or something with a low crest factor. With audio
or pink noise, you wouldn't see any tangible difference. The capacitor
will create a LPF which may decrease any HF noise on the power rails,
but this ultrasonic noise is of little consequence for audio.

5. Lights dim when voltage falls. This is because the alternator cannot
output enough current to the system. Because audio amplifiers have SMPS,
they are constant power; this means if input voltage decreases, input
current increases. The voltage in the entire system decreases when the
amp pulls a lot of current. Adding a capacitor won't help much because
capacitors can't do much about transients, like audio, and help mostly
with continuous signals. Think of each circuit [lights, stereo, ECU,
etc.] being in parallel. Since each load looks like a resistor [with
likely some reactance], the voltages are equal. So if the stereo drags
the voltage down, the voltage in the entire system goes down.

6. I care about the technical merits of audio, but then again I get
paid to :)


--
eezip

[mike]

eezip wrote:
> mike;661966 Wrote:
>
>>My questions are these:
>>
>>1) In a well-installed car stereo, what are the typical resistances in
>>the
>>wiring, connections, batteries etc.?
>>
>>2) What is a typical speaker impedance for a high power system?
>>
>>3) If I add a big cap, What improvement would I HEAR under what
>>conditions?
>>
>>4) If I hooked up an oscilloscope to the power supply at the amp and
>>to
>>it's audio output,
>>what improvement would I SEE under what conditions if I added a cap?
>>
>>5) Why would lights dim? IF the lights connect to the battery.
>>And the stereo connects to the battery.
>>And the sense point for the alternator/regulator connects to the
>>battery.
>>And the alternator is big enough.
>>??
>>
>>6) What percentage of people actually care about the technical merits
>>of
>>their
>>car stero...as long as their ride is "fly" and gets 'em laid? ;-)
>>
>>mike
>
>
> Mike,
>
> You didn't post any time-domain approximation.
>
> CMRR doesn't affect the power supply in any way. CMRR is the rejection
> of a signal that is common to 2 signals, for instance, noise picked up
> in the ground and hot of an RCA is rejected by the CMRR of a
> differential amplifier. The term you mean is PSRR, or power supply
> rejection ratio.

My bad.
>
> The 'voltage upconversion' you're talking about is the power supply.
> All car amps I've seen use a push-pull SMPS. Push-pull supplies can be
> regulated or unregulated. If they're regulated, the output voltage is
> largely unaffected by fluctuations of the input voltage - unregulated
> supplies have a fixed boost and the output moves just as the input
> moves or the load changes. Most car audio amplifier supplies I've seen
> were unregulated for reasons that hopefully will be more clear later in
> my response. Regulated supplies also have a specific loop gain which
> affects the power supplies' ability to react to changes in input or
> output. With regulated supplies this is also a factor.
>
> 1. I think 'this chart'
> (http://www.interfacebus.com/Copper_Wire_AWG_SIze.html) will help you
> out. Keep in mind these values are for a specific temperature and type
> of cable jacket. This can vary with different winding techniques of the
> strands in the cable as well.
>
> 2. As above, 1-4ohm is typical for a car subwoofer.
>
> 3. You won't hear any difference. Adding more capacitance will help
> continuous power a bit, because CV will be greater, but the voltage
> rails won't change so peak power will remain the same. Audio has high
> crest factors, and thus peak power is actually relevant and meaningful
> for audio. RMS power tells you more about how robust the amp is, but
> peak power won't change if you add more capacitance. And that is what
> will tell you more about audible difference between amplifiers.
>
> 4. Same as above. You'll see high peaks before clipping if the input
> signal is a sine wave or something with a low crest factor. With audio
> or pink noise, you wouldn't see any tangible difference. The capacitor
> will create a LPF which may decrease any HF noise on the power rails,
> but this ultrasonic noise is of little consequence for audio.
>
> 5. Lights dim when voltage falls. This is because the alternator cannot
> output enough current to the system. Because audio amplifiers have SMPS,
> they are constant power; this means if input voltage decreases, input
> current increases. The voltage in the entire system decreases when the
> amp pulls a lot of current. Adding a capacitor won't help much because
> capacitors can't do much about transients, like audio, and help mostly
> with continuous signals.

I was agreeing with you until you got here.
Transients (or in the frequency domain, higher frequency perturbations)
are the ONLY things capacitors (parallel on the power rail) can do
anything about???

Think of each circuit [lights, stereo, ECU,
> etc.] being in parallel. Since each load looks like a resistor [with
> likely some reactance], the voltages are equal. So if the stereo drags
> the voltage down, the voltage in the entire system goes down.
>
> 6. I care about the technical merits of audio, but then again I get
> paid to :)
>
>

[Matt Ion]

eezip wrote:

Good technical answers overall, nicely done :)

> The 'voltage upconversion' you're talking about is the power supply.
> All car amps I've seen use a push-pull SMPS. Push-pull supplies can be
> regulated or unregulated. If they're regulated, the output voltage is
> largely unaffected by fluctuations of the input voltage - unregulated
> supplies have a fixed boost and the output moves just as the input
> moves or the load changes. Most car audio amplifier supplies I've seen
> were unregulated for reasons that hopefully will be more clear later in
> my response. Regulated supplies also have a specific loop gain which
> affects the power supplies' ability to react to changes in input or
> output. With regulated supplies this is also a factor.

One thing to remember too: since there is little in the way of 'standards' for
measuring and rating car-audio amp output, a lot of manufacturers will use the
more voltage = more power factor to "fudge" their output ratings. If, for
example, they bench-test their amp with a full 14.5VDC supply, they'll get a
certain power rating across a specified load... but as the voltage decreases
(generally to around 13.5VDC with engine at idle, 12VDC or even down to 11VDC
when running from battery), so does the actual output power.

14.5V down to 12V is close to a 15% reduction in voltage, which if I remember my
Ohm's Law correctly, means as much as a 30% drop from the "rated" power output
of the amp (forgive me if I'm off, I'm just home from a marvellous steak dinner
and several Irish coffees, so my brain isn't quite working at peak efficiency).

When you look at the output specs for an amp with a regulated supply, you know
you're getting that output pretty much across the range of supply voltage. With
an unregulated supply, unless the manufacturer states otherwise, you can
probably assume that their rated output is available only at a high idle without
a bunch of other accessories turned on.

> 1. I think 'this chart'
> (http://www.interfacebus.com/Copper_Wire_AWG_SIze.html) will help you
> out. Keep in mind these values are for a specific temperature and type
> of cable jacket. This can vary with different winding techniques of the
> strands in the cable as well.

In the real world, of course, as long as your power wires are of sufficient
guage, you're probably not to to notice a significant difference from one to the
next.

> 3. You won't hear any difference. Adding more capacitance will help
> continuous power a bit, because CV will be greater, but the voltage
> rails won't change so peak power will remain the same. Audio has high
> crest factors, and thus peak power is actually relevant and meaningful
> for audio. RMS power tells you more about how robust the amp is, but
> peak power won't change if you add more capacitance. And that is what
> will tell you more about audible difference between amplifiers.
>
> 4. Same as above. You'll see high peaks before clipping if the input
> signal is a sine wave or something with a low crest factor. With audio
> or pink noise, you wouldn't see any tangible difference. The capacitor
> will create a LPF which may decrease any HF noise on the power rails,
> but this ultrasonic noise is of little consequence for audio.

To my understanding, the main reasoning behind the "stiffening" capacitor is
this: Given the resistance of the supply wires and relatively slow reaction time
of the alternator to increased current loads (notice even when turning on your
headlights, other lights will often dim briefly), the electrical system may not
be able to respond fast enough to very sudden, very brief, very-high current
demands (like a really hard bass hit). The cap is therefor added very close to
the amp, where the interconnecting wire length, and thus the resistance of that
wire, will be minimal, allowing the cap to respond to those instantaneous
high-current demands.

> 5. Lights dim when voltage falls. This is because the alternator cannot
> output enough current to the system. Because audio amplifiers have SMPS,
> they are constant power; this means if input voltage decreases, input
> current increases. The voltage in the entire system decreases when the
> amp pulls a lot of current. Adding a capacitor won't help much because
> capacitors can't do much about transients, like audio, and help mostly
> with continuous signals. Think of each circuit [lights, stereo, ECU,
> etc.] being in parallel. Since each load looks like a resistor [with
> likely some reactance], the voltages are equal. So if the stereo drags
> the voltage down, the voltage in the entire system goes down.

Indeed.

Thus a cap is NOT a replacement for a better alternator, since it doesn't CREATE
current. It should help the dimming lights A BIT, because the amp's immediate
current demands are fed by the cap before pulling them from the rest of the
system, giving the alternator time to respond to the increased need, but the
current to recharge the cap after that hit still has to come from somewhere, and
an alt that can't supply the current will still be overworked and the lights
will still dim... maybe just not quite as much, and very slightly time-delayed :)

> 6. I care about the technical merits of audio, but then again I get
> paid to :)

Damn, I wish I got paid for caring too!