NetNews Usenet Archive 1992 #27/NN_1992_27.iso/spool/rec/audio/car/4853

Text File | 1992-11-15 | 5.0 KB | 99 lines

Newsgroups: rec.audio.car
Path: sparky!uunet!pmafire!mica.inel.gov!ux1!news.byu.edu!eff!sol.ctr.columbia.edu!usc!zaphod.mps.ohio-state.edu!wupost!darwin.sura.net!eng.ufl.edu!crow.eel.ufl.edu!brian
From: [e-mail] (Brian Gentry)
Subject: Re: Capacitors, etc... CALMLY
Message-ID: <[e-mail]>
Sender: [e-mail] (Usenet Diskhog System)
Organization: UF EE Department (student)
References: <[e-mail]> <[e-mail]> <[e-mail]>
Date: Sun, 15 Nov 92 22:32:32 GMT
Lines: 87

I seem to have started this by asking why Andrew wanted to use a 10uF
tantalum in parallel with the big electrolytics used for "stiffening caps".
This discussion has prompted me to do some thinking on the subject and I
have a few comments.

In article <[e-mail]>, [e-mail] (Andrew Gilchrist) writes:
|> [e-mail] writes:
[...]
|> Usually not directly, true.
|> >the output devices by the amp's power supply.  Also, it is worth mentioning
|> >that the amp's power supply has its own filter caps.  These will tend to be
|>
|> On the load side of the power supply, yes.  Since this is not an off-line
|> switcher, caps on the supply side are not really necessary.  Anything put
|> there is to keep noise from getting onto the power system.  Any noise that
|> goes back onto the system will affect eveything else put there.
|> The point in this case, is to isolate the amp from the rest of the power
|> system.

Well, our larger amps at U.S.AMPS use filter capacitors on the power INPUT
side as well as the output side of the trasformer.  These are used primarily
to filter out ripple in the 12V power supply.

[...]
|> >smaller capacitances (hence lower inductance) than we'd typically see on the
|> >output of a linear power supply, since the amp's DC upverter typically runs at
|> >supersonic frequencies rather than 60 Hz.
|>
|> Oh yeah?  Do you think the power company picked 60 Hz by flipping a coin?
|> It is true that a switcher can run at supersonic frequencies, but you will find
|> that this is not usually desireable.  Still, your point about filter caps
|> is valid, since the frequencies used are generally higher than 60Hz.

What do you mean "Oh yeah"?  The majority of mobile audio amplfiers run at
frequencies greater than 20kHz.  Why would you say that it is "undesirable"
to do so?  There are many benifits to using a high frequency power supply.
Many computer power supplies are run at frequencies greater than 20kHz as
well.  They are quite common.

[...]
|> In any case, you are quite correct in pointing out the ability of a well
|> designed switcher to improve immunity to power rail fluctuations.
|>
|> However, the current has to come from somewhere, and it comes from
|> where everything else gets power.  If you play a loud transient (bass
|> drum, eg) the switcher will suck power out of the charging circuit
|> much faster than a linear regulator could hope to.  The load voltage will
|> drop, and while this may not affect the amplifier's power rails, since it is
|> supplied by the switcher, your head unit will feel it. And so will your
|> CD-changer or whatever.

This brings me back to the original point of using small caps with good
high frequency characteristics in parallel with large caps with poor high
frequency performance.  If you put a 10uF tantalum capacitor in parallel
with a .5F electrolytic (or any size electrolytic) you will only add:
E = .5 * C * V^2 = 0.5 * .00001 * 144 =  0.00072 Joules of energy storage
capacity to the system.  In TTL circuits where the transistor's switching
causes glitches in the power supply, this tiny increase in energy storage
would be enough to control the high frequency noise in the power supply.
However, we are talking about a high power car audio system that may demand
hundreds of amps from the car's power system.

So, I agree that high frequency components will be "injected" into the power
system by high energy transients, but I don't think that a small capacitor
will have much of an effect on the problem because of it's VERY small
contribution to the overall energy storage capacity of the system.

Think of it this way:  A large amplitude bass drum note begins.  The current
demanded by the amplifier increases rapidly.  The big electrolytic cap
can't change it's current that fast, so it adds almost nothing to the total
current going to the amp.  The little cap can change that fast and does.
The little cap gives up it's entire energy store.  This energy store is
so small that it will make almost no difference in the total current demanded
by the amplifier(s).  Remember, with a 10uF cap we're only talking about
720 MicroJoules of energy.  However, this condition is only true when the
base note first begins.  Once the bass note is acting as a sine wave (after
the voltage changes direction for the first time) then the current demand
is at a low frequency and the large cap can begin to deliver current to
the amplifier(s).

[...]
|>
|> Andrew

                 Brian L. Gentry
             ([e-mail])