Help with audio amplification

[[email protected]]

I hope you'll forgive a simple query.

I have an audio amplifier module that takes 12v and an audio input, and
in turn drives a loudspeaker. Specifically, it is the module shown
on...

http://www.kemo-electronic.de/en/module/m031/index.htm

....and for what it's worth, I've wired ina 10k log potentiometer as
suggested.

I'm using this to amplify the audio output from a Garmin Quest ---
a small in-car GPS navigation unit. The audio output I'm using is
intended for use with an earphone. Using a multimeter I see that for
the duration of the announcements, the DC voltage across the output
rises from 0v to about 1v.

The problem I have is that <1s before and after the GPS makes an
announcement, the louspeaker makes a loud and irritating pop, and
I'd really appreciate some advice on what to do. By experimenting I
discovered that putting a 10 or a 100 micro-Farad capacitor across
the audio input to the amplifier removes the final pop and most of
the first one. It also reduces the volume. Since I don't really know
what I'm doing I'd be grateful if anyone could let me know the best
way to approach this.

Thanks and regards,

Jim

 

[Chris Jones]

I hope you'll forgive a simple query.

I have an audio amplifier module that takes 12v and an audio input, and
in turn drives a loudspeaker. Specifically, it is the module shown
on...

http://www.kemo-electronic.de/en/module/m031/index.htm

...and for what it's worth, I've wired ina 10k log potentiometer as
suggested.

I'm using this to amplify the audio output from a Garmin Quest ---
a small in-car GPS navigation unit. The audio output I'm using is
intended for use with an earphone. Using a multimeter I see that for
the duration of the announcements, the DC voltage across the output
rises from 0v to about 1v.

The problem I have is that <1s before and after the GPS makes an
announcement, the louspeaker makes a loud and irritating pop, and
I'd really appreciate some advice on what to do. By experimenting I
discovered that putting a 10 or a 100 micro-Farad capacitor across
the audio input to the amplifier removes the final pop and most of
the first one. It also reduces the volume. Since I don't really know
what I'm doing I'd be grateful if anyone could let me know the best
way to approach this.

Thanks and regards,

Jim

Probably you will have more success if you put a capacitor in series with
the audio input, between the GPS and the pot. This would form a high-pass
filter. You would still hear a click, but with less bass. You could
adjust the cutoff frequency, fc=1/(6.28 * R * C) where R is roughly the 10k
of your pot, and C is the capacitor you put in series. Try maybe 100nF or
1uF to start with.

Do you hear the click when the earphone is connected instead? If there is
no click with the earphone, then you might have one of those capacitorless
headphone amplifiers which are so trendy these days. They work by biasing
the "ground" of the headphone jack up to some DC voltage so that the other
terminals of the headphone socket can swing above or below the "ground" of
the headphone socket, without the expense of a 5 cent capacitor. If it is
one of these chips then you could use an audio isolation transformer.
(should cost $1 or so as a component, or more with connectors already
attached, but you might be able to salvage one from an old transistor radio
or similar) You could also use a differential amplifier but that would be
more work.

Chris

 

[Mac]

I hope you'll forgive a simple query.

I have an audio amplifier module that takes 12v and an audio input, and
in turn drives a loudspeaker. Specifically, it is the module shown
on...

http://www.kemo-electronic.de/en/module/m031/index.htm

...and for what it's worth, I've wired ina 10k log potentiometer as
suggested.

I'm using this to amplify the audio output from a Garmin Quest ---
a small in-car GPS navigation unit. The audio output I'm using is
intended for use with an earphone. Using a multimeter I see that for
the duration of the announcements, the DC voltage across the output
rises from 0v to about 1v.

The problem I have is that <1s before and after the GPS makes an
announcement, the louspeaker makes a loud and irritating pop, and
I'd really appreciate some advice on what to do. By experimenting I
discovered that putting a 10 or a 100 micro-Farad capacitor across
the audio input to the amplifier removes the final pop and most of
the first one. It also reduces the volume. Since I don't really know
what I'm doing I'd be grateful if anyone could let me know the best
way to approach this.

Thanks and regards,

Jim

You can probably get away with a series resistor and a shunt capacitor.
This forms a low-pass filter. By tweaking the capacitor and resistor
values, you can probably reduce the pop to an almost inaudible level
without attenuating the voice too much.

You want your cutoff frequency to be around 5 kHz, I guess. That should
let the voice through. You have a 10 k pot, so you want the R to be around
100 Ohms. Now we can just ignore the pot.

C = 1/(2 * pi * 5kHz * 100 Ohms) = 318 nF

So use the next lowest value, 220 nF. This will move the cutoff up from 5
kHz, but the exact value isn't critical, so this should be OK.

I am infamous for making arithmetic mistakes in this forum, so double
check my arithmetic. Hopefully I got equation right. ;-)

Oh, since you say you don't really know what you are doing, I'll draw an
ASCII art schematic (view with courier or other fixed-width font):


100
signal---/\/\/-----+-----------------+
| |
| \
| 10 k /
= 220 nF pot \<--- amp input
| /
| \
| |
+-----------------+
|

 

[Mac]

Probably you will have more success if you put a capacitor in series with
the audio input, between the GPS and the pot. This would form a high-pass
filter. You would still hear a click, but with less bass. You could
adjust the cutoff frequency, fc=1/(6.28 * R * C) where R is roughly the 10k
of your pot, and C is the capacitor you put in series. Try maybe 100nF or
1uF to start with.

[snip]

Heh. I just finished telling him to try a low-pass filter. Would that not
work?

--Mac

 

[John Woodgate]

I am infamous for making arithmetic mistakes in this forum, so double
check my arithmetic. Hopefully I got equation right. ;-)

Not quite.

Oh, since you say you don't really know what you are doing, I'll draw
an ASCII art schematic (view with courier or other fixed-width font):

You have proposed a low-pass filter. Ideal for getting rid of the signal
and leaving the pop.

 

[Pooh Bear]

Probably you will have more success if you put a capacitor in series with
the audio input, between the GPS and the pot. This would form a high-pass
filter. You would still hear a click, but with less bass. You could
adjust the cutoff frequency, fc=1/(6.28 * R * C) where R is roughly the 10k
of your pot, and C is the capacitor you put in series. Try maybe 100nF or
1uF to start with.

[snip]

Heh. I just finished telling him to try a low-pass filter. Would that not
work?

He needs a high pass filter.

Having looked at the product description though, I wonder if the amp isn't
simply defective. I'd expect the amp module to be already ac coupled.

Graham

 

[Mac]

Not quite.

You have proposed a low-pass filter. Ideal for getting rid of the signal
and leaving the pop.

So I have now gathered. For some reason I thought the pop was mostly high
frequency.

Oh well.

Nice thing about usenet is that bad information will usually get corrected
by followups.

--Mac

 

[Mark]

A HIGH PASS filter might change the pop into a click.
If you want to get rid of the click too , you'll need something more
complicated, like a gated attenuator.

Mark

 

[GPG]

Terminate the GPS output with a resistor equal to the normal headphone
load.

 

[[email protected]]

Thanks very much for all the replies. If nothing else I have
a lot of ideas to keep me busy with!

It was interesting to see the dissent regarding
high/low pass filters. Having a physics background,
I had assumed that the click was essentially a square
pulse reflecting the bias of the GPS output. Thinking
of the corresponding Fourier series, I would have thought
that blocking high frequencies would remove the
sharp corner of the square pulse and round it out more.

Thanks once again,

Jim

 

[Rich Grise]

Thanks very much for all the replies. If nothing else I have
a lot of ideas to keep me busy with!

It was interesting to see the dissent regarding
high/low pass filters. Having a physics background,
I had assumed that the click was essentially a square
pulse reflecting the bias of the GPS output. Thinking
of the corresponding Fourier series, I would have thought
that blocking high frequencies would remove the
sharp corner of the square pulse and round it out more.

Actually, that would turn a 'click' into a 'thump'. What you
need to do is provide a leakage path from the amplifier side
of the coupling capacitor, to drain off the DC bias from the
cap. Or mute it completely, which is another project. ;-)

Good Luck!
Rich