[EE] Audio muting - (was "Looking for N-ch MOSFET TO-92")
Josh Koffman
> Its a long story, but I'm using it to mute a low-level audio
> signal. MOSFETs work very well in this type of application, so long
> as the negative peak audio level does not approach the turn-on
> voltage of the intrinsic back-body diode in the MOSFET. That is the
> case for this project.
Dwayne, would you mind talking a bit more about how you're muting the
signal? Does it click when you mute? I made a little box about a year
ago to mute a microphone signal from a headset mounted boom mic
(dynamic mic). I'm not great with analog, but I managed to find a
circuit that used a relay, a resistor, and a cap to make the mute
"clickless". Once I made it I discovered that it did still click, but
it wasn't super loud so it wasn't a big deal. What is a bit annoying
is that it isn't a complete mute, more like a huge damper so some
sound still sneaks through if it's loud enough. I'd love to update the
circuit but I haven't had time yet. I'd love suggestions on making it
solid state.
Just out of curiousity, would your solution unmute if power was lost?
That was one of the things I liked about the relay - if my circuit
ever lost power communication could still occur.
Thanks!
Josh
--
A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Dwayne Reid
>Dwayne, would you mind talking a bit more about how you're muting
>the signal? Does it click when you mute?
It can. There are really only two ways to ensure that a full mute is
noiseless: delay turning the mute switch to the mute position until
you reach zero-crossing *or* 'fade' the mute in.
For example, I build Ref Mic switches for the sports-broadcast
industry. These are worn by the referees and go in-line between the
microphone and the body-pack transmitter. These 'fade' from full
audio to full mute over a period of about 150ms. They are completely
free of any click or other artifacts.
In this case, though, I don't care about click. This is partly
because I'm 'dimming' the audio (attenuating it) rather than a full
mute. This is also happening at the same time as DTMF signals are
being generated and any click would be lost in the DTMF tone.
>I made a little box about a year ago to mute a microphone signal
>from a headset mounted boom mic (dynamic mic). I'm not great with
>analog, but I managed to find a circuit that used a relay, a
>resistor, and a cap to make the mute "clickless". Once I made it I
>discovered that it did still click, but it wasn't super loud so it
>wasn't a big deal. What is a bit annoying is that it isn't a
>complete mute, more like a huge damper so some sound still sneaks
>through if it's loud enough. I'd love to update the circuit but I
>haven't had time yet. I'd love suggestions on making it solid state.
Dynamic mics are pretty easy to deal with. You might try setting
your relay up so that it puts a short across the mic terminals when
you want mute. Dynamic mic capsules are typically 200R impedance and
will run into a dead short without damage. Be aware, though, that
you can still get clicks. That's because you are not controlling
WHEN the relay closes and puts the short across the mic. If the mic
happens to be putting out a signal that is not near ground, the
sudden change from whatever that voltage was to 0V sounds like a
'click' (because it is). Its the sharp fall time that gets you.
I can perhaps send you one of my mute PCB's. It will be potted in
epoxy, I'm afraid. Although simple, it is truly non-obvious in the
way it works and the person for whom I build these is afraid that the
design could become widely known and he will then lose one small part
of his competitive advantage. As it is, I build these pretty much
only for him - he then supplies them to sports broadcasters in both
the US and Canada.
I'd have to do a slightly custom version for you - most of these
units that I build are for electret mic capsules. However, making it
work with a dynamic mic is pretty easy. It could be powered from a
small lithium cell.
>Just out of curiousity, would your solution unmute if power was
>lost? That was one of the things I liked about the relay - if my
>circuit ever lost power communication could still occur.
The Ref Mic mute circuit that I mentioned above is powered from the
bias supply used for the electret mic capsule. It won't work if
there is no bias - but neither does the mic capsule. Its a fair
trade-off <grin>.
The circuit that I needed these N-channel MOSFETs for is set up such
that the audio is NOT affected should the mute circuit lose
power. However, that's just the way this particular setup works - it
wasn't specifically designed that way.
I hope this helps!
dwayne
--
Dwayne Reid <dwa...@planet.eon.net>
Trinity Electronics Systems Ltd Edmonton, AB, CANADA
(780) 489-3199 voice (780) 487-6397 fax
www.trinity-electronics.com
Custom Electronics Design and Manufacturing
Josh Koffman
> It can. There are really only two ways to ensure that a full mute is
> noiseless: delay turning the mute switch to the mute position until
> you reach zero-crossing *or* 'fade' the mute in.
I had initially considered trying to find a digital pot or something
that would allow me to vary the resistance. I didn't end up going this
route because I was worried about impedances and I wasn't sure if I'd
have to add buffering to the signals.
> For example, I build Ref Mic switches for the sports-broadcast
> industry. These are worn by the referees and go in-line between the
> microphone and the body-pack transmitter. These 'fade' from full
> audio to full mute over a period of about 150ms. They are completely
> free of any click or other artifacts.
I believe I've probably seen your work on TV then!
> In this case, though, I don't care about click. This is partly
> because I'm 'dimming' the audio (attenuating it) rather than a full
> mute. This is also happening at the same time as DTMF signals are
> being generated and any click would be lost in the DTMF tone.
>
Just to clarify, you end up at a full mute, right? Your previous
paragraph mentions going from full audio to full mute.
> Dynamic mics are pretty easy to deal with. You might try setting
> your relay up so that it puts a short across the mic terminals when
> you want mute. Dynamic mic capsules are typically 200R impedance and
> will run into a dead short without damage. Be aware, though, that
> you can still get clicks. That's because you are not controlling
> WHEN the relay closes and puts the short across the mic. If the mic
> happens to be putting out a signal that is not near ground, the
> sudden change from whatever that voltage was to 0V sounds like a
> 'click' (because it is). Its the sharp fall time that gets you.
I don't have a schematic handy, but the basic idea was that there was
a resistor and a largish electrolytic cap (100uF springs to mind but I
can't really remember) across the mic lines. The relay shorted out the
resistor. It certainly is a lot less obtrusive then a dead short
across the lines, I've heard that and my clicks are less noticeable.
> I can perhaps send you one of my mute PCB's. It will be potted in
> epoxy, I'm afraid. Although simple, it is truly non-obvious in the
> way it works and the person for whom I build these is afraid that the
> design could become widely known and he will then lose one small part
> of his competitive advantage. As it is, I build these pretty much
> only for him - he then supplies them to sports broadcasters in both
> the US and Canada.
I appreciate the offer, but I'm happy to discuss it abstractly for a
while. I don't plan on making these commercially, but it'd be nice to
figure out a way to do it that didn't have to use a "black box"
circuit. If we get too close to proprietary info, please just tell me.
> I'd have to do a slightly custom version for you - most of these
> units that I build are for electret mic capsules. However, making it
> work with a dynamic mic is pretty easy. It could be powered from a
> small lithium cell.
Power isn't an issue, my circuit has a PIC in it. Yes, it's a bit
overkill for just switching a relay, but I have it set up to debounce
the two switches are able to trigger the mute, and also to offer
additional "modes" of operation (push to talk, push to mute, push to
toggle, etc).
> The Ref Mic mute circuit that I mentioned above is powered from the
> bias supply used for the electret mic capsule. It won't work if
> there is no bias - but neither does the mic capsule. Its a fair
> trade-off <grin>.
Nice!
> The circuit that I needed these N-channel MOSFETs for is set up such
> that the audio is NOT affected should the mute circuit lose
> power. However, that's just the way this particular setup works - it
> wasn't specifically designed that way.
>
> I hope this helps!
Well it's certainly got me thinking! The only time I've used MOSFETs
recently was for a PWM DC lamp dimmer. Not really applicable here. If
I found a logic level FET with the right specs (not that I know what
the right specs are, but the one you indicated might be good), how
would I connect it to the audio lines? I guess I would need a common
ground between one line of the mic signal and the drain perhaps? Then
connect source to the other line? I'm not familiar enough with MOSFETs
to know if that would even work. I'm also unsure how that would effect
the audio signal. The ground of my circuit would be varying in
relation to the audio ground as the mic diaphragm moves in and out. I
guess in retrospect one nice thing about the relay is that
power/ground-wise my circuit is isolated from the mic and connected
equipment.
However, assuming electrically this could be made to work, would I be
able to PWM the gate and ramp it up and down to achieve a soft mute?
Or would I need a smoother voltage ramp, something a bit more A2D?
Thanks for getting me thinking about this!
Josh
--
A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams
--
Olin Lathrop
> I had initially considered trying to find a digital pot or something
> that would allow me to vary the resistance. I didn't end up going this
> route because I was worried about impedances and I wasn't sure if I'd
> have to add buffering to the signals.
This may not be applicable to your situation, but a long time ago I did a
"electronic volume control" circuit using LEDs and CdS cells. This was
before digital pots and the like were available. Make a resistor divider
from two CdS cells and light the LEDs opposite from each other. Since these
are inherently isolated, you can use them directly as attenuators, or even
in the feedback loop of gain stages. CdS cells are rather slow, which in
this case was used to advantage to eliminate popping. They are fast enough
to respond as a volume control on a human time scale.
RussellMc
>> I had initially considered trying to find a digital pot or something
>> that would allow me to vary the resistance. I didn't end up going this
>> route because I was worried about impedances and I wasn't sure if I'd
>> have to add buffering to the signals.
Olin replied:
> This may not be applicable to your situation, but a long time ago I did a
> "electronic volume control" circuit using LEDs and CdS cells.
Yes. An entirely practical solution.
CdS sensors are still available (I bought some within the last year)
BUT ROHS may frown on the Cadmium content.
Various photo transistors may be used in the same manner.
Even an optocoupler may be usable.
Bias issues may mean DC levels need to be managed.
For extra isolation ratio a series and shunt cell could be used.
Back to back optos may allow better bidirectional response - as long
as peak off voltage do not break down reverse CE junctions.
Any of the above are liable to be horrendously non linear and are
liable to need linearisation or feedback for sensible use. The digital
pot starts sounding attractive :-).
This of course addresses the 'digital pot' suggestion but not the
current TO92 one.
R
ivp
>> did a "electronic volume control" circuit using LEDs and CdS cells
The early musical instrument tremolos used a bulb / CdS. They were
followed by LED / CdS or photo-transistors and then VCA ICs like
the MFC6040 (odd little thing that was) and the MC3340. An
interesting application is the Univibe, one of Hendrix's signature effects,
which is part-tremolo, part-phaser shifter, part-Leslie simulator. It
causes his "wobbly" sound, as heard throughout the Rainbow Bridge
album
http://www.univox.org/pics/schematics/univibe.gif
alan.b...@stfc.ac.uk
> >> I had initially considered trying to find a digital pot or
something
> >> that would allow me to vary the resistance. I didn't end up going
> this
> >> route because I was worried about impedances and I wasn't sure if
I'd
> >> have to add buffering to the signals.
>
> Olin replied:
> > This may not be applicable to your situation, but a long time ago I
> did a
> > "electronic volume control" circuit using LEDs and CdS cells.
>
> Yes. An entirely practical solution.
> CdS sensors are still available (I bought some within the last year)
> BUT ROHS may frown on the Cadmium content.
> Various photo transistors may be used in the same manner.
> Even an optocoupler may be usable.
You can get optocouplers that use an FET rather than a bipolar
transistor. I don't know how linear the transfer function would be in
this instance, but probably worth investigating.
--
Scanned by iCritical.
Josh Koffman
bit of searching on using a FET to mute audio. I found this page with
a bit of searching:
http://www.pic101.com/audiosw/index.html
The FET circuit seems simple enough to implement, but I've seen some
chatter on a forum that this circuit won't attenuate enough with a low
impedance. My source is low impedance, but it's also balanced.
So...I'm a little confused if I could make this work. Would I have to
use two of these circuits (one on each of the mic lines)? Would it
make sense to power my circuit off an isolate DC/DC converter to avoid
ground hum issues? I think it's safe to assume that my circuit will be
off a different power source (and possibly different ground) from the
one that the mic is connected to.
I'm fairly confused about this.
Josh
--
A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams
Dwayne Reid
>Sorry to reincarnate this thread, but I finally had some time to do a
>bit of searching on using a FET to mute audio. I found this page with
>a bit of searching:
>http://www.pic101.com/audiosw/index.html
>
>The FET circuit seems simple enough to implement, but I've seen some
>chatter on a forum that this circuit won't attenuate enough with a low
>impedance. My source is low impedance, but it's also balanced.
>
>So...I'm a little confused if I could make this work. Would I have to
>use two of these circuits (one on each of the mic lines)? Would it
>make sense to power my circuit off an isolate DC/DC converter to avoid
>ground hum issues? I think it's safe to assume that my circuit will be
>off a different power source (and possibly different ground) from the
>one that the mic is connected to.
Ouch. Using series FET switches on a balanced line means that you
need to ensure that the switches track identically. But: its not the
end of the world.
Best way is to try it. Build 2 of the first circuit shown but for
now, eliminate the Q2 transistor. Use a SPDT switch wired up to a 9V
battery, with the battery (-) going to gnd and the switch armature
going to both FET gates.
Operate the switch and see what it sounds like. Armature at gnd =
audio ON; armature at +9V = audio off.
Then we can go on from there.
dwayne
--
Dwayne Reid <dwa...@planet.eon.net>
Trinity Electronics Systems Ltd Edmonton, AB, CANADA
(780) 489-3199 voice (780) 487-6397 fax
www.trinity-electronics.com
Custom Electronics Design and Manufacturing
--
Harold Hallikainen
muting balanced line level audio. They work well. We also use the NJU7313
SPI controlled analog switch for low level audio switching (source
selection, gain setting, etc.). Another nice SPI controlled analog switch
is the ADG714.
Harold
--
FCC Rules Updated Daily at http://www.hallikainen.com - Advertising
opportunities available!
Josh Koffman
> Ouch. Using series FET switches on a balanced line means that you
> need to ensure that the switches track identically. But: its not the
> end of the world.
I would connect them to the same PIC pin so they'd switch at the same
time (in theory).
> Best way is to try it. Build 2 of the first circuit shown but for
> now, eliminate the Q2 transistor. Use a SPDT switch wired up to a 9V
> battery, with the battery (-) going to gnd and the switch armature
> going to both FET gates.
>
> Operate the switch and see what it sounds like. Armature at gnd =
> audio ON; armature at +9V = audio off.
>
> Then we can go on from there.
Sounds good. I will try to find some P-channel FETs locally and wire it up.
Any thoughts on the grounding/voltage supply issue?
Thanks!
Josh
--
A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams
--
Josh Koffman
<har...@hallikainen.org> wrote:
> I'm not sure of your entire application, but we use DG452s for audio
> muting balanced line level audio. They work well. We also use the NJU7313
> SPI controlled analog switch for low level audio switching (source
> selection, gain setting, etc.). Another nice SPI controlled analog switch
> is the ADG714.
Thanks for the info Harold. Do you use one switch to short the high
and low lines together? Or do you use two switches to disconnect the
lines?
Do you have any grounding issues or do you have everything connected
to a common ground?
Thanks!
Josh
--
A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams
Harold Hallikainen
We're mostly using the NJU7313 to select between low impedance sources, so
the open switch resistance is very large compared to the source
resistance.
There is a separate ground system for the analog circuitry, which is tied
to the digital at one point. The switches get their power and ground from
the digital side of things. This does not throw any signal into the analog
being carried by the switches.
Good luck!
Harold
--
FCC Rules Updated Daily at http://www.hallikainen.com - Advertising
opportunities available!