FPGA Central

Hi

Can anyone explain the principle of an anti-Larsen circuit or give
references.
I know that it is used to prevent howling due to an acoustical path from
loudspeaker to microphone, but I do not know which technique is used.

Thanks in advance
Jan Hansen

Jan Rørgård Hansen wrote:

> Hi
>
> Can anyone explain the principle of an anti-Larsen circuit or give
> references.
> I know that it is used to prevent howling due to an acoustical path from
> loudspeaker to microphone, but I do not know which technique is used.
>
> Thanks in advance
> Jan Hansen
>
>
Did you run down http://www.google.com/search?q=anti-Larsen%20circuit ?

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

"Jerry Avins" <[email protected]> wrote in message
news:bn68vh$tb$[email protected]..
> Jan Rørgård Hansen wrote:
>
> > Hi
> >
> > Can anyone explain the principle of an anti-Larsen circuit or give
> > references.
> > I know that it is used to prevent howling due to an acoustical path from
> > loudspeaker to microphone, but I do not know which technique is used.
> >
> > Thanks in advance
> > Jan Hansen
> >
> >
> Did you run down http://www.google.com/search?q=anti-Larsen%20circuit ?

I didn't find an explanation the first time, but OK I tried again and:
I understand that an anti-Larsen circuit decreases the gain in the speaker
amplifier dependent on how much signal is received by the microphone.
I can see how this would work in a telephone with the speaker enabled. I
have heard that it should have been used in PA-systems to avoid howling,
this I don't understand.

Any explanations?

Jan

Jan Rørgård Hansen wrote>
> Hi
>
> Can anyone explain the principle of an anti-Larsen circuit or give
> references.

Gee, I hope Keith doesn't take this personally!

Jan Rørgård Hansen wrote:

> "Jerry Avins" <[email protected]> wrote in message
> news:bn68vh$tb$[email protected]..
>
>>Jan Rørgård Hansen wrote:
>>
>>
>>>Hi
>>>
>>>Can anyone explain the principle of an anti-Larsen circuit or give
>>>references.
>>>I know that it is used to prevent howling due to an acoustical path from
>>>loudspeaker to microphone, but I do not know which technique is used.
>>>
>>>Thanks in advance
>>>Jan Hansen
>>>
>>>
>>
>>Did you run down http://www.google.com/search?q=anti-Larsen%20circuit ?
>
>
> I didn't find an explanation the first time, but OK I tried again and:
> I understand that an anti-Larsen circuit decreases the gain in the speaker
> amplifier dependent on how much signal is received by the microphone.
> I can see how this would work in a telephone with the speaker enabled. I
> have heard that it should have been used in PA-systems to avoid howling,
> this I don't understand.
>
> Any explanations?
>
> Jan
>
>
No, just guesses. An adaptive filter seems more reasonable to me. The
few systems I've seen described in detail put the filter in what I think
is the wrong place -- assuming I understood. I hope you're familiar with
two-wire full-duplex telephony. One way to look at it is that feedback
from the the transmitter prevents the receiver from seeing local
signals. The scheme breaks down when the delay and attenuation between
local transmission and reception isn't matched by the feedback. The
adaptive filter is an element in the feedback.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Andor wrote:

> Jan Rørgård Hansen wrote>
>
>>Hi
>>
>>Can anyone explain the principle of an anti-Larsen circuit or give
>>references.
>
>
> Gee, I hope Keith doesn't take this personally!
>
>

That's "Larson" He's in the clear.
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

"Jan Rørgård Hansen" <jrh(ingenSpAm)@NOSPAMPLEASE!!person.dk> wrote in
message news:3f9774d9$0$29575$[email protected] ...
>
> "Jerry Avins" <[email protected]> wrote in message
> news:bn68vh$tb$[email protected]..
> > Jan Rørgård Hansen wrote:
> >
> > > Hi
> > >
> > > Can anyone explain the principle of an anti-Larsen circuit or give
> > > references.
> > > I know that it is used to prevent howling due to an acoustical path
from
> > > loudspeaker to microphone, but I do not know which technique is used.
> > >
> > > Thanks in advance
> > > Jan Hansen
> > >
> > >
> > Did you run down http://www.google.com/search?q=anti-Larsen%20circuit ?
>
> I didn't find an explanation the first time, but OK I tried again and:
> I understand that an anti-Larsen circuit decreases the gain in the speaker
> amplifier dependent on how much signal is received by the microphone.
> I can see how this would work in a telephone with the speaker enabled. I
> have heard that it should have been used in PA-systems to avoid howling,
> this I don't understand.
>
> Any explanations?
>
> Jan

In a PA system, howling occurs when the gain of a system is too great.
Imagine a single mic in proximity to a single speaker. As the gain of the
system is increased (i.e. mic volume is turned up), the mic "hears" more and
more of what the speaker is putting out. Eventually, when the loop gain >
1, the speaker outputs a sound, which the mic hears, which is amplified,
which the speaker puts out again, which the mic hears now at an even louder
level...howling. That's just a rough layman's explanation.

So one way to combat that is to simply turn down the volume until the loop
gain is < 1. A sound operator will often do just that. Then the howling
goes away. In fact, this could be done automatically with a conventional
compressor/limiter which decreases gain when the signal is above a
threshold. However, often times one wants to achieve maximum gain without
howling (e.g. a soft talker). In that case, turning down the level is
contrary to what you are trying to achieve, though it does at least prevent
howling.

Another approach is with filtering. Feedback often occurs at a single
frequency, or at least a fairly narrow range of frequencies. This is due to
the fact that the frequency response of the system is never flat due to many
variables. By cutting just the frequency that is most likely to feedback,
you can get more gain before feedback. Commercial anti-feedback devices do
just that with adaptive filters. So do experienced sound operators with
parametric or graphic EQs--it's often called ringing out a room.*

I've never heard of a anti-Larsen circuit before, but looking at it briefly,
it looks like it simply cuts the gain based on some simple criteria such as
plugging in a microphone. It doesn't sound to me like this is broadly
applicable to PA systems.

*Often times, after you cut one frequency, another will pop up after just a
slight bit more gain is added. So this can be a iterative process.
Anti-feedback devices often have multiple filters (8-24) to deal with
howling at multiple frequencies.

Right... Not me.

> That's "Larson" He's in the clear.

+------------------------------------------+
|Keith Larson |
|Member Group Technical Staff |
|Texas Instruments Incorporated |
| |
| 281-274-3288 |
| [email protected] |
|------------------------------------------+
| TMS320C3x/C4x/VC33 Applications |
| |
| $150 TMS320VC33 DSK's ARE AVAILABLE NOW |
| |
| TMS320VC33 |
| The lowest cost and lowest power |
| floating point DSP on the planet! |
| 500uw/Mflop |
+------------------------------------------+