will a dc offset affect a speaker?

[panfilero]

ok... here's my question: If I have a millivolt range audio signal
and I add a dc offset to it, and then I take my new signal and send it
to a speaker with a built in amplifer, will this be the same as
sending the audio signal alone to the amplifier?

What I think: I think yes... because the way speakers work is by
vibrating along with the fluctuations of the audio signal, and since a
dc signal is flat, this wont affect the output of the audio
amplifer....

thanks

 

[Steve]

How much DC are you sending and is it deliberate or
a side effect of an amplifier stage? Sending DC to a speaker is not a
really a good thing I would think as it will push the cone out of its centre
range. But I would guess that the built-in amplifier is AC-coupled so the
DC might never make it to the speaker as the first stage of the amp blocks
it.

 

[panfilero]

How much DC are you sending and is it deliberate or
a side effect of an amplifier stage? Sending DC to a speaker is not a
really a good thing I would think as it will push the cone out of its centre
range. But I would guess that the built-in amplifier is AC-coupled so the
DC might never make it to the speaker as the first stage of the amp blocks
it.

2.5V, delibrate

 

[Tom Biasi]

How much DC are you sending and is it deliberate or
a side effect of an amplifier stage? Sending DC to a speaker is not a
really a good thing I would think as it will push the cone out of its
centre
range. But I would guess that the built-in amplifier is AC-coupled so the
DC might never make it to the speaker as the first stage of the amp blocks
it.

2.5V, delibrate

Why?

Tom

 

[panfilero]

2.5V, delibrate

Why?

Tom

Because I'm passing an audio signal through a 4066 analog switch which
can not handle the negative swing from an audio signal, so I'd like to
include a DC offset of 1/2 the Vcc of the chip in order to avoid this.

 

[Spehro Pefhany]

ok... here's my question: If I have a millivolt range audio signal
and I add a dc offset to it, and then I take my new signal and send it
to a speaker with a built in amplifer, will this be the same as
sending the audio signal alone to the amplifier?

Usually you'll want to AC-couple the output, although there is
doubtless some AC coupling in an external amplfier.

What I think: I think yes... because the way speakers work is by
vibrating along with the fluctuations of the audio signal, and since a
dc signal is flat, this wont affect the output of the audio
amplifer....

thanks

You might get a wee bit of an audible click* at turn-on (and
turn-off).

* possibly to the point of blowing out the speakers.


Best regards,
Spehro Pefhany

 

[panfilero]

---
Why not use a 4051 instead?

Just set Vee to a voltage enough more negative than your input signal in
order to provide the needed footroom and you should be OK.

JF

The 4051 looks pretty good, except its got a fairly high Ron
resistance that might throw me off.... so.... there's not much of a
difference between a MUX and an analog switch is there?

 

[Spehro Pefhany]

---
Why not use a 4051 instead?

Just set Vee to a voltage enough more negative than your input signal in
order to provide the needed footroom and you should be OK.

JF

Or a 74HC4316, which is like a 4066 but has level translation like the
4051. You can make -V from +V with a capacitor charge pump such as a
7660 or an inductor-based switching supply.


Best regards,
Spehro Pefhany

 

[Archimedes' Lever]

2.5V, delibrate

Why?

Tom

A DC offset will affect a speaker, if it is direct coupled.

Speakers use a standing magnetic field so ANY voltage on the coil will
affect its position. That coil is suspended in the speaker assembly at a
specific central (to its motion) rest point. A DC offset will change
that "rest point", which will subsequently affect the speaker's
operation. It will be compressed response for half the wave for every
cycle it carries. For signals less than the offset voltage, the signal
would "disappear" for that half of each cycle.

 

[Le Chaud Lapin]

ok... here's my question:  If I have a millivolt range audio signal
and I add a dc offset to it, and then I take my new signal and send it
to a speaker with a built in amplifer, will this be the same as
sending the audio signal alone to the amplifier?

What I think: I think yes... because the way speakers work is by
vibrating along with the fluctuations of the audio signal, and since a
dc signal is flat, this wont affect the output of the audio
amplifer....

Hmm..., I must comment, since this vaguely implies that a speaker will
"auto-AC-couple itself" in presence of signal with DC offset. It
won't. Without a blocking capacitor, DC-component will choke the
speaker, since, after all, a speaker is little more than a solenoid
with membrane attached to the dynamic part.

An analogy would be getting a neck massage from a masseuse who
applies, in addition the nice, gentle, AC undulating motion, a
potentially suffocating DC grip that is always present.

-Le Chaud Lapin-

 

[[email protected]]

---
Actually, there is.

An analog switch is a bidirectional contrivance which can transfer
analog (or digital) signals back and forth, while a conventional
multiplexer is unidirectional.

Correct, but only if you are talking about a digital multiplexer,
which panfilero obviously isn't.

TheCD4051 (like the 4052 and 4053) are all analog multiplexers and
perfectly bi-directional.

 

[[email protected]]

ok... here's my question:  If I have a millivolt range audio signal
and I add a dc offset to it, and then I take my new signal and send it
to a speaker with a built in amplifer, will this be the same as
sending the audio signal alone to the amplifier?

What I think: I think yes... because the way speakers work is by
vibrating along with the fluctuations of the audio signal, and since a
dc signal is flat, this wont affect the output of the audio
amplifer....

As has been pointed out repeatedly in this thread, this is wrong. The
DC signal will move the speaker coil away from it normal resting
position. A small DC offset may not matter too much, but 2.5V is a
lot.

One complicating factor, not mentioned in the thread so far, is that
the stiffness/spring constant of many speakers is mostly determined by
the compressibility of the air in the cabinet. The cabinet is never
completely sealed - because if it were changes in atmospheric pressure
would move the coil off-centre - so the stiffness of the speaker is
much lower at DC than at music frequencies, and even a small DC offset
can jam the speaker at one or other end of its travel, giving very
nasty distortion if you try and get music out if it.

 

[John Devereux]

As has been pointed out repeatedly in this thread, this is wrong. The
DC signal will move the speaker coil away from it normal resting
position. A small DC offset may not matter too much, but 2.5V is a
lot.

One complicating factor, not mentioned in the thread so far, is that
the stiffness/spring constant of many speakers is mostly determined by
the compressibility of the air in the cabinet. The cabinet is never
completely sealed - because if it were changes in atmospheric pressure
would move the coil off-centre - so the stiffness of the speaker is
much lower at DC than at music frequencies, and even a small DC offset
can jam the speaker at one or other end of its travel, giving very
nasty distortion if you try and get music out if it.

So what you are saying is that DC causes sidebands?

 

[[email protected]]

   No surprise there.

Mike Terrill making his usual sophisticated contribution to the
discussion.

I don't know where we'd be without his wisdom and insight.

 

[[email protected]]

This seems unlikely.

My use of "conventional multiplexer" was intended to refer to digital
multiplexers since analog multiplexers are more properly called
"multiplexer/demultiplexer":

"Conventional" is much less specific than "digital" - you were simply
careless in your choice of words.

http://focus-webapps.ti.com/general/docs/sitesearch/searchdevice.tsp;...

and, I believe, are used less frequently than digital multiplexers and
demultiplexers, making them, in my view, less conventional.

"Conventional" is an odd word to use in this context - one doesn't
choose to use a particular sort of multiplexer because there are
socially accepted conventions on what you need in a particular
circuit, you use something that will work in that circuit.

It may seem obvious to you now, but you didn't bother making this less-
than-helpful distinction when you originally corrected Panfilero. The
analog/digital distinction captures more of the idea you were trying -
rather ineptly - to get across. Of course, there are internally
buffered analog multiplexers, which aren't bidirectional either, and
"multiplexer/demultipler" does exclude these parts but you need to
prime an unsophisticated audience fairly carefully (as I've done
here) to make the title meaningful.

 

[[email protected]]

So what you are saying is that DC causes sidebands?

Enough DC bias on the right loudspeaker could persuade the loudspeaker
to generate acoustic side-bands of the AC content of the electrical
input to the speaker. Guy Macon discusses the mechanism in more
detail.

 

[Eeyore]

ok... here's my question: If I have a millivolt range audio signal
and I add a dc offset to it, and then I take my new signal and send it
to a speaker with a built in amplifer, will this be the same as
sending the audio signal alone to the amplifier?

Unless the speaker/amplifier was designed by cretins, there will be a 'DC
blocking' capacitor at the input so it won't make any difference at all.

What I think: I think yes... because the way speakers work is by
vibrating along with the fluctuations of the audio signal, and since a
dc signal is flat, this wont affect the output of the audio
amplifer....

What you think is totally wrong. Apply a significant DC offset *directly*
to the speaker and you'll get a very nasty effect.

Graham

 

[Eeyore]

How much DC are you sending and is it deliberate or
a side effect of an amplifier stage? Sending DC to a speaker

He said a "speaker/amplifier".

Read the damn post FFS !

Graham

 

[Eeyore]

Because I'm passing an audio signal through a 4066 analog switch which
can not handle the negative swing from an audio signal, so I'd like to
include a DC offset of 1/2 the Vcc of the chip in order to avoid this.

You can run a 4066 off +/- 7.5V and offset the logic inputs vial level
translators. NO DC added.

Graham

 

[Eeyore]

---
Why not use a 4051 instead?

Just set Vee to a voltage enough more negative than your input signal in
order to provide the needed footroom and you should be OK.

The 4053 ? is also very popular in this kind of task. But you can make a 4016 or
4066 behave equivalently too with split supplies.

Graham