Audio Delay Circuit

[George W]

I have a need to measure the relative delay between two audio signals
arriving at a receiver via different paths. The audio signals are standard
3 kHz audio bandwidths containing voice or music. One could arrive as much
as 1 msec before or after the other. I'd like to be able to measure the
difference in time of arrival between the two signals down to 1 usec if
that's possible.

Any suggestions? I'm thinking of using a 180 degree phase shifter and
adjusting a variable delay line for a null when the signals are summed.

Thanks.

George

 

[Mac]

I have a need to measure the relative delay between two audio signals
arriving at a receiver via different paths. The audio signals are standard
3 kHz audio bandwidths containing voice or music. One could arrive as much
as 1 msec before or after the other. I'd like to be able to measure the
difference in time of arrival between the two signals down to 1 usec if
that's possible.

Any suggestions? I'm thinking of using a 180 degree phase shifter and
adjusting a variable delay line for a null when the signals are summed.

Thanks.

George

Nowadays it might be cheap enough to do a DSP based solution. You might
not need an actual DSP. You might be able to just use programmable
logic, but the idea is to correlate the two signals and find out where
the maximum is of the correlation function.

Or something like that. I'd have to look into the details. ;-)

--Mac

 

[Tam/WB2TT]

I have a need to measure the relative delay between two audio signals
arriving at a receiver via different paths. The audio signals are
standard
3 kHz audio bandwidths containing voice or music. One could arrive as
much
as 1 msec before or after the other. I'd like to be able to measure the
difference in time of arrival between the two signals down to 1 usec if
that's possible.

Any suggestions? I'm thinking of using a 180 degree phase shifter and
adjusting a variable delay line for a null when the signals are summed.

Thanks.

George

How about limiting each signal, and using a counter in time interval mode to
measure A - B or B - A. Most counters would give 100 ns resolution. You
might also have to bandlimit each signal to BW < 1KHz on each channel prior
to the limiters.

Tam

Tam

 

[Robert]

Hell use a Scope, dual channel no problem.

 

[Tam/WB2TT]

Hell use a Scope, dual channel no problem.

Done that. I felt he wanted something more sophisticated.

Tam

 

[Ken Smith]

I have a need to measure the relative delay between two audio signals
arriving at a receiver via different paths. The audio signals are standard
3 kHz audio bandwidths containing voice or music. One could arrive as much
as 1 msec before or after the other. I'd like to be able to measure the
difference in time of arrival between the two signals down to 1 usec if
that's possible.

Any suggestions? I'm thinking of using a 180 degree phase shifter and
adjusting a variable delay line for a null when the signals are summed.

How long do you have to get the answer. If you've got a little time you
could do it with a microcontroller with a built in ADC.

If you were using a DSP, you might consider the FFT method of correlation.
In a microcontroller, you would be better off in time domain.

You are converting, lets say, the two inputs at a 50KHz rate. You need
enough ram to hold a little more than 50 numbers from which ever channel
you think is the faster and 50 correlation results. You start off
assuming that the offset is the full 1mS (50 samples), and correlate what
is coming in at that offset. After a while you step to the next delay and
so on until you've covered the whole +/-1mS span. You then use a best fit
peak finder to find the peak's location. Getting the location to within
1/20th of a point should be no trouble.

 

[Paul Hovnanian P.E.]

Hell use a Scope, dual channel no problem.

Hmm. A 3 kHz signal with -1 to +1 ms delay. At 3 kHz, that's 6 cycles
with no way of knowing which peaks correlate. For a modulated signal,
correlation might be easier to determine, but for something like audio
(speech for example), that might be difficult to follow on a scope.

Even a DSP correlator could match the wrong peaks if the modulation
didn't have enough frequency components to make the solution unique. A
longer sampling window will minimize this problem.

 

[Mac]

Hmm. A 3 kHz signal with -1 to +1 ms delay. At 3 kHz, that's 6 cycles
with no way of knowing which peaks correlate. For a modulated signal,
correlation might be easier to determine, but for something like audio
(speech for example), that might be difficult to follow on a scope.

Even a DSP correlator could match the wrong peaks if the modulation
didn't have enough frequency components to make the solution unique. A
longer sampling window will minimize this problem.

The OP said voice or music. I would not expect there to be any problem
correlating this in a DSP, although I have never done it. I actually think
that it would be easy on a scope, too (I have looked at music on a scope),
but if that's the kind of solution that was wanted, why ask on s.e.d?

--Mac

 

[Tam/WB2TT]

Hmm. A 3 kHz signal with -1 to +1 ms delay. At 3 kHz, that's 6 cycles
with no way of knowing which peaks correlate. For a modulated signal,
correlation might be easier to determine, but for something like audio
(speech for example), that might be difficult to follow on a scope.

That is why I suggested bandlimiting to less than 1 KHz.

Tam