frequency detector

D

dileep

Guest
Hi,
In my application, i observe a 100Hz signal for a time period of
60ms...and this signal is very important for my application...i have
to build a ciruit which detects an audio frequency that exists for
only 60 milli sec..i would like to know whether there are any products
using which we can do the same..i am thankful to you if you can help
me.
regards
dileep
 
On 13 Mar 2004 23:28:45 -0800, dileepmaddukuri@yahoo.co.in (dileep)
wrote:

Hi,
In my application, i observe a 100Hz signal for a time period of
60ms...and this signal is very important for my application...i have
to build a ciruit which detects an audio frequency that exists for
only 60 milli sec..i would like to know whether there are any products
using which we can do the same..i am thankful to you if you can help
me.
---
If you feed the 100Hz signal into a comparator it will output six pulses
any number of which which you can count to determine when the signal is
there.
--
John Fields
 
"John Fields" <jfields@austininstruments.com> wrote in message
news:12p8505vkk00rnagee1upagfv5s7nc1kkc@4ax.com...
On 13 Mar 2004 23:28:45 -0800, dileepmaddukuri@yahoo.co.in (dileep)
wrote:

Hi,
In my application, i observe a 100Hz signal for a time period of
60ms...and this signal is very important for my application...i have
to build a ciruit which detects an audio frequency that exists for
only 60 milli sec..i would like to know whether there are any products
using which we can do the same..i am thankful to you if you can help
me.

---
If you feed the 100Hz signal into a comparator it will output six pulses
any number of which which you can count to determine when the signal is
there.
--
The problem here, is that if the 'gate' is asynchronous to the signal, there
are a lot of different combinations that will give the same result. A
frequency nearly 20% in error, that just happends to cross the comparator
threshold immediately after the gate opens, will still give six edges...
The 'solution', is to approach the problem from the other end, and have an
accurate zero crossing detector, and time the interval between successive
edges. There are examples of doing this as standard 'demos', for both PIC
and Atmel processors. If you count the pulses from (say) a 5MHz clock, then
there will be a count change, for a frequency change of just 0.002Hz.
The other alternative (depending on the repeat interval of the window), is
to lock a PLL, only pumping this during the window time, and then measuring
the frequency over a longer interval.

Best Wishes
 
On Sun, 14 Mar 2004 15:28:36 -0000, "Roger Hamlett"
<rogerspamignored@ttelmah.demon.co.uk> wrote:

"John Fields" <jfields@austininstruments.com> wrote in message
news:12p8505vkk00rnagee1upagfv5s7nc1kkc@4ax.com...
On 13 Mar 2004 23:28:45 -0800, dileepmaddukuri@yahoo.co.in (dileep)
wrote:

Hi,
In my application, i observe a 100Hz signal for a time period of
60ms...and this signal is very important for my application...i have
to build a ciruit which detects an audio frequency that exists for
only 60 milli sec..i would like to know whether there are any products
using which we can do the same..i am thankful to you if you can help
me.

---
If you feed the 100Hz signal into a comparator it will output six pulses
any number of which which you can count to determine when the signal is
there.
--
The problem here, is that if the 'gate' is asynchronous to the signal, there
are a lot of different combinations that will give the same result.
---
The point was not that a count is being made to determine frequency,
it's that a count is being made which can be used to determine the
presence or absence of a signal. The OP wasn't very specific about what
he wants to do, but the gist of his request seemed to be that he's only
interested in detecting a signal and he's (maybe) got a 60ms window in
which to make that determination.

--
John Fields
 
"John Fields" <jfields@austininstruments.com> wrote in message
news:po4950lc34p8e8osg0oolrr19mbjl6iuc4@4ax.com...
On Sun, 14 Mar 2004 15:28:36 -0000, "Roger Hamlett"
rogerspamignored@ttelmah.demon.co.uk> wrote:


"John Fields" <jfields@austininstruments.com> wrote in message
news:12p8505vkk00rnagee1upagfv5s7nc1kkc@4ax.com...
On 13 Mar 2004 23:28:45 -0800, dileepmaddukuri@yahoo.co.in (dileep)
wrote:

Hi,
In my application, i observe a 100Hz signal for a time period of
60ms...and this signal is very important for my application...i have
to build a ciruit which detects an audio frequency that exists for
only 60 milli sec..i would like to know whether there are any products
using which we can do the same..i am thankful to you if you can help
me.

---
If you feed the 100Hz signal into a comparator it will output six
pulses
any number of which which you can count to determine when the signal is
there.
--
The problem here, is that if the 'gate' is asynchronous to the signal,
there
are a lot of different combinations that will give the same result.

---
The point was not that a count is being made to determine frequency,
it's that a count is being made which can be used to determine the
presence or absence of a signal. The OP wasn't very specific about what
he wants to do, but the gist of his request seemed to be that he's only
interested in detecting a signal and he's (maybe) got a 60ms window in
which to make that determination.
Fair enough. I read it as wanting to actually 'identify' the signal. :)

Best Wishes
 
On Sun, 14 Mar 2004 11:27:38 -0600, John Fields <jfields@austininstruments.com>
wrote:

On Sun, 14 Mar 2004 15:28:36 -0000, "Roger Hamlett"
rogerspamignored@ttelmah.demon.co.uk> wrote:


"John Fields" <jfields@austininstruments.com> wrote in message
news:12p8505vkk00rnagee1upagfv5s7nc1kkc@4ax.com...
On 13 Mar 2004 23:28:45 -0800, dileepmaddukuri@yahoo.co.in (dileep)
wrote:

Hi,
In my application, i observe a 100Hz signal for a time period of
60ms...and this signal is very important for my application...i have
to build a ciruit which detects an audio frequency that exists for
only 60 milli sec..i would like to know whether there are any products
using which we can do the same..i am thankful to you if you can help
me.

---
If you feed the 100Hz signal into a comparator it will output six pulses
any number of which which you can count to determine when the signal is
there.
--
The problem here, is that if the 'gate' is asynchronous to the signal, there
are a lot of different combinations that will give the same result.

---
The point was not that a count is being made to determine frequency,
it's that a count is being made which can be used to determine the
presence or absence of a signal. The OP wasn't very specific about what
he wants to do, but the gist of his request seemed to be that he's only
interested in detecting a signal and he's (maybe) got a 60ms window in
which to make that determination.
It is oh so common here that part problems are presented for a cure. We seem to
get sucked in to providing a solution - or usually many solutions - without
first being able to properly identify the problem.

For example, it wasn't indicated by the O/P whether the 100Hz is in the presence
of other signal components, or noise. If not, a raw detector is (obviously) the
simplest approach. We do not know if the solution needs to reject signals +/- n
Hz away.

Often in practice when I am presented with a problem couched in the O/P's terms,
I pursue detail. Usually my first question is "WHY (do you want/need ..)?"
 
"budgie" <me@privacy.net> wrote in message
news:6ofa505kjq3tj7hi0vpvirb1uallor97pq@4ax.com...
On Sun, 14 Mar 2004 11:27:38 -0600, John Fields
jfields@austininstruments.com
wrote:

On Sun, 14 Mar 2004 15:28:36 -0000, "Roger Hamlett"
rogerspamignored@ttelmah.demon.co.uk> wrote:


"John Fields" <jfields@austininstruments.com> wrote in
message
news:12p8505vkk00rnagee1upagfv5s7nc1kkc@4ax.com...
On 13 Mar 2004 23:28:45 -0800,
dileepmaddukuri@yahoo.co.in (dileep)
wrote:

Hi,
In my application, i observe a 100Hz signal for a
time period of
60ms...and this signal is very important for my
application...i have
to build a ciruit which detects an audio frequency
that exists for
only 60 milli sec..i would like to know whether there
are any products
using which we can do the same..i am thankful to you
if you can help
me.

---
If you feed the 100Hz signal into a comparator it will
output six pulses
any number of which which you can count to determine
when the signal is
there.
--
The problem here, is that if the 'gate' is asynchronous
to the signal, there
are a lot of different combinations that will give the
same result.

---
The point was not that a count is being made to determine
frequency,
it's that a count is being made which can be used to
determine the
presence or absence of a signal. The OP wasn't very
specific about what
he wants to do, but the gist of his request seemed to be
that he's only
interested in detecting a signal and he's (maybe) got a
60ms window in
which to make that determination.

It is oh so common here that part problems are presented
for a cure. We seem to
get sucked in to providing a solution - or usually many
solutions - without
first being able to properly identify the problem.

For example, it wasn't indicated by the O/P whether the
100Hz is in the presence
of other signal components, or noise. If not, a raw
detector is (obviously) the
simplest approach. We do not know if the solution needs
to reject signals +/- n
Hz away.

Often in practice when I am presented with a problem
couched in the O/P's terms,
I pursue detail. Usually my first question is "WHY (do
you want/need ..)?"

Notice my signature which I first started using in the early
80s when I provided the last level of hardware support for
IBM PCs in Australia.
--
John G

Wot's Your Real Problem?
 
On Mon, 15 Mar 2004 17:56:02 +1100, "John G" <Greentest@ozemail.com.au> wrote:


Notice my signature which I first started using in the early
80s when I provided the last level of hardware support for
IBM PCs in Australia.
Exactly ;-)
 

Welcome to EDABoard.com

Sponsor

Back
Top