Receiver bandpass filter questions.

Guest
Hello,

I have some questions about my first basic bandpass filter for an
ultrasonic receiver circuit. It's a Mupltiple feedback second order
bandpass II with Q=13.97, f0=40kHz, Apb=0.5dB, Bpb=1kHz and a gain of
32. All resistors are 1% with 2% ceramic caps. As long as R7 is exact
then the filters gain/frequency graph remains O.K. with the other
resistors as standard values.

I have put a schematic on the net at:
http://www.geocities.com/talionis.geo/Temp/temp.html

Questions:
Should I have the bandpass filter on both parts of the dual opamp? It
seems logical that with the filter on the first half the second half
wouldn't need one.

Have I connected it correctly to the second half - C4 and Pin#5?


In addition a problem I also have is without any load I can adjust the
trimmer TR1 down to 0.3V and still keep the comparators output high.
The comparator is attached to the trigger pin of a 555 astable timer
circuit which switches a load. As soon as I attach a load to the 555
things change. If for example I attach a 3V motor or even an LED via a
transistor switch then I have to readjust the trimmer or the load
won't turn off. The trimmer has to be adjusted to about 3.5V out for
the motor and 1.4V out for the LED. This reduces the sensitivity of
the receiver. My next step after the filter is to look into a stable
voltage reference for the LM393 but will that solve this particular
problem?

Also is it wise to use a pullup resistor on the comparators output?

Regards,

Andrew.
 
<Rubicon> wrote in message news:41219528.5520302@news.netaccess.co.nz...
Hello,

I have some questions about my first basic bandpass filter for an
ultrasonic receiver circuit. It's a Mupltiple feedback second order
bandpass II with Q=13.97, f0=40kHz, Apb=0.5dB, Bpb=1kHz and a gain of
32. All resistors are 1% with 2% ceramic caps. As long as R7 is exact
then the filters gain/frequency graph remains O.K. with the other
resistors as standard values.

I have put a schematic on the net at:
http://www.geocities.com/talionis.geo/Temp/temp.html

Questions:
Should I have the bandpass filter on both parts of the dual opamp? It
seems logical that with the filter on the first half the second half
wouldn't need one.

Have I connected it correctly to the second half - C4 and Pin#5?


In addition a problem I also have is without any load I can adjust the
trimmer TR1 down to 0.3V and still keep the comparators output high.
The comparator is attached to the trigger pin of a 555 astable timer
circuit which switches a load. As soon as I attach a load to the 555
things change. If for example I attach a 3V motor or even an LED via a
transistor switch then I have to readjust the trimmer or the load
won't turn off. The trimmer has to be adjusted to about 3.5V out for
the motor and 1.4V out for the LED. This reduces the sensitivity of
the receiver. My next step after the filter is to look into a stable
voltage reference for the LM393 but will that solve this particular
problem?

Also is it wise to use a pullup resistor on the comparators output?

Regards,

Andrew.
Click to expand...
Dunno about the filter stuff but most of the comparators that "Sink" their
output will need a pullup resistor if you need a :high".


--
Regards ........... Rheilly Phoull
 
Rubicon wrote:
Hello,

I have some questions about my first basic bandpass filter for an
ultrasonic receiver circuit. It's a Mupltiple feedback second order
bandpass II with Q=13.97, f0=40kHz, Apb=0.5dB, Bpb=1kHz and a gain of
32. All resistors are 1% with 2% ceramic caps. As long as R7 is exact
then the filters gain/frequency graph remains O.K. with the other
resistors as standard values.

I have put a schematic on the net at:
http://www.geocities.com/talionis.geo/Temp/temp.html

Questions:
Should I have the bandpass filter on both parts of the dual opamp? It
seems logical that with the filter on the first half the second half
wouldn't need one.

Have I connected it correctly to the second half - C4 and Pin#5?


In addition a problem I also have is without any load I can adjust the
trimmer TR1 down to 0.3V and still keep the comparators output high.
The comparator is attached to the trigger pin of a 555 astable timer
circuit which switches a load. As soon as I attach a load to the 555
things change. If for example I attach a 3V motor or even an LED via a
transistor switch then I have to readjust the trimmer or the load
won't turn off. The trimmer has to be adjusted to about 3.5V out for
the motor and 1.4V out for the LED. This reduces the sensitivity of
the receiver. My next step after the filter is to look into a stable
voltage reference for the LM393 but will that solve this particular
problem?

Also is it wise to use a pullup resistor on the comparators output?

Regards,

Andrew.
Click to expand...
Your bandpass filter will latch at negative rail if you keep the resistors
R8 and R7. C1 is useless and can be omitted. We have posted the same
bandpass here under ugly transfer function. Look at the design formula and
determine your Q, it is not 14.
Maybe you put the part-B stage in front of the bandpass. I doubt that this
bandpass filter is needed at all. If you have the transmitting signal at
hand it will be better to make a synchonous rectifier with the delayed
transmitter signal. What do you want to measure?
--
ciao Ban
Bordighera, Italy
 
<Rubicon> wrote in message
news:41219528.5520302@news.netaccess.co.nz...
Hello,

I have some questions about my first basic bandpass filter for an
ultrasonic receiver circuit.
Click to expand...
I don't know much about filter design, but I've had really good luck
using LM567 chips in 40kHz ultrasonic applications such as this. Its a
'tone decoder', basically a PLL with a logic output when the thing
locks. It can be a bit persnickety to get going, but does a very good
job once its tuned up. It is VERY sensitive.

LM567s are out of production, at least at national, but you can
occasionally get them surplus. Also, I think NTE still sells them.

However, I'm guessing you could use any PLL, such as a 4046, using the
phase comparator output as an indication of lock. Here is a lock
indication circuit I lifted from a Forest Mims book:


Vcc ------------------------------------+----------------+
| |
V RED V GRN
- LED - LED
| Unlocked | LCKD
.-. .-.
|1| |1|
4001 4001 |k| 4001 |k|
__ 100k __ '-' __ '-'
PC1-----|>=| ___ +-|>=| | +-|>=| |
|1 |o--+--|___|--+-----+ |1 |o----+----+ |1 |o-----+
PC2-----|__| | | +-|__| +-|__|
'--->|----+
1N4148 |
|
--- 0.05 uF
---
|
GND

created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

Its meant to indicate lock of a 4046; PC1 goes to pin 1, and PC2 goes
to pin 2. Thus, it works no matter what phase comparator you are
using. The PC1 or PC2 input goes high, and the little circuit between
the first two NOR gates acts like a filter, filtering out the
transistions caused by the phase comparators. I'm guessing that the
4046 + digital output is similar to an LM567. The 4046 + 4001 probably
uses far less power, if thats an issue.

Regards,
Bob Monsen
 
<Rubicon> wrote in message
news:4122edcc.13580699@news.netaccess.co.nz...
On Mon, 16 Aug 2004 20:15:19 GMT, "Robert C Monsen"
rcsurname@comcast.net> wrote:


Rubicon> wrote in message
news:41219528.5520302@news.netaccess.co.nz...

Hello,

I have some questions about my first basic bandpass filter for an
ultrasonic receiver circuit.

I don't know much about filter design, but I've had really good
luck
using LM567 chips in 40kHz ultrasonic applications such as this.
Its a
'tone decoder', basically a PLL with a logic output when the thing
locks. It can be a bit persnickety to get going, but does a very
good
job once its tuned up. It is VERY sensitive.

LM567s are out of production, at least at national, but you can
occasionally get them surplus. Also, I think NTE still sells them.

However, I'm guessing you could use any PLL, such as a 4046, using
the
phase comparator output as an indication of lock. Here is a lock
indication circuit I lifted from a Forest Mims book:


Vcc ------------------------------------+----------------+
| |
V RED V
GRN
- LED -
LED
| Unlocked |
LCKD
.-. .-.
|1| |1|
4001 4001 |k| 4001 |k|
__ 100k __ '-' __ '-'
PC1-----|>=| ___ +-|>=| | +-|>=| |
|1 |o--+--|___|--+-----+ |1 |o----+----+ |1 |o-----+
PC2-----|__| | | +-|__| +-|__|
'--->|----+
1N4148 |
|
--- 0.05 uF
---
|
GND

created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

Its meant to indicate lock of a 4046; PC1 goes to pin 1, and PC2
goes
to pin 2. Thus, it works no matter what phase comparator you are
using. The PC1 or PC2 input goes high, and the little circuit
between
the first two NOR gates acts like a filter, filtering out the
transistions caused by the phase comparators. I'm guessing that the
4046 + digital output is similar to an LM567. The 4046 + 4001
probably
uses far less power, if thats an issue.

Regards,
Bob Monsen




Bob,

Thankyou for all the info there. Phase comparators are components I
haven't yet had to deal with but I'll look into them now.

Regards,

Andrew.
Click to expand...
Andrew, A phase comparator is part of a phase-locked loop (PLL), which
is what a 4046 is called.

A PLL consists of a voltage controlled oscillator (VCO), and a phase
comparator circuit. The phase comparator takes as input the input
frequency and the output of the VCO, and outputs a voltage. That
voltage is used to control the VCO. Thus, you have a feedback system
where the input frequency is tracked by the VCO. The circuit above is
used to indicate whether the VCO is tracking the input freqency. The
first LED is on if its not, and the second LED is on if it is...

Thus, the way you set it up is to configure the 4046 to lock near
40kHz (using information in the datasheet), and to feed the amplified
output of your receiver transducer into the PLL input. When the input
is near 40kHz, it'll lock, and the green LED will turn on.

You can use this to turn on your switch.

The advantage of the LM567 mentioned before is that its designed with
this application in mind, so you don't need the output filter circuit
given above. Here is a link where you can get 5 of these for 99 cents
+ shipping.

http://sales.goldmine-elec.com/prodinfo.asp?prodid=8036

Regards,
Bob Monsen
 
Thanyou for your reply.

I'm not trying to actually measure anything but create a sensitive and
reliable battery powered ultrasonic switch and I'm having trouble
doing it.

Cheers,

Andrew.




On Mon, 16 Aug 2004 14:18:27 GMT, "Ban" <bansuri@web.de> wrote:

Rubicon wrote:
Hello,

I have some questions about my first basic bandpass filter for an
ultrasonic receiver circuit. It's a Mupltiple feedback second order
bandpass II with Q=13.97, f0=40kHz, Apb=0.5dB, Bpb=1kHz and a gain of
32. All resistors are 1% with 2% ceramic caps. As long as R7 is exact
then the filters gain/frequency graph remains O.K. with the other
resistors as standard values.

I have put a schematic on the net at:
http://www.geocities.com/talionis.geo/Temp/temp.html

Questions:
Should I have the bandpass filter on both parts of the dual opamp? It
seems logical that with the filter on the first half the second half
wouldn't need one.

Have I connected it correctly to the second half - C4 and Pin#5?


In addition a problem I also have is without any load I can adjust the
trimmer TR1 down to 0.3V and still keep the comparators output high.
The comparator is attached to the trigger pin of a 555 astable timer
circuit which switches a load. As soon as I attach a load to the 555
things change. If for example I attach a 3V motor or even an LED via a
transistor switch then I have to readjust the trimmer or the load
won't turn off. The trimmer has to be adjusted to about 3.5V out for
the motor and 1.4V out for the LED. This reduces the sensitivity of
the receiver. My next step after the filter is to look into a stable
voltage reference for the LM393 but will that solve this particular
problem?

Also is it wise to use a pullup resistor on the comparators output?

Regards,

Andrew.

Your bandpass filter will latch at negative rail if you keep the resistors
R8 and R7. C1 is useless and can be omitted. We have posted the same
bandpass here under ugly transfer function. Look at the design formula and
determine your Q, it is not 14.
Maybe you put the part-B stage in front of the bandpass. I doubt that this
bandpass filter is needed at all. If you have the transmitting signal at
hand it will be better to make a synchonous rectifier with the delayed
transmitter signal. What do you want to measure?
--
ciao Ban
Bordighera, Italy
Click to expand...
 
Bob,

Thankyou for all the info there. Phase comparators are components I
haven't yet had to deal with but I'll look into them now.

Regards,

Andrew.


On Mon, 16 Aug 2004 20:15:19 GMT, "Robert C Monsen"
<rcsurname@comcast.net> wrote:

Rubicon> wrote in message
news:41219528.5520302@news.netaccess.co.nz...

Hello,

I have some questions about my first basic bandpass filter for an
ultrasonic receiver circuit.

I don't know much about filter design, but I've had really good luck
using LM567 chips in 40kHz ultrasonic applications such as this. Its a
'tone decoder', basically a PLL with a logic output when the thing
locks. It can be a bit persnickety to get going, but does a very good
job once its tuned up. It is VERY sensitive.

LM567s are out of production, at least at national, but you can
occasionally get them surplus. Also, I think NTE still sells them.

However, I'm guessing you could use any PLL, such as a 4046, using the
phase comparator output as an indication of lock. Here is a lock
indication circuit I lifted from a Forest Mims book:


Vcc ------------------------------------+----------------+
| |
V RED V GRN
- LED - LED
| Unlocked | LCKD
.-. .-.
|1| |1|
4001 4001 |k| 4001 |k|
__ 100k __ '-' __ '-'
PC1-----|>=| ___ +-|>=| | +-|>=| |
|1 |o--+--|___|--+-----+ |1 |o----+----+ |1 |o-----+
PC2-----|__| | | +-|__| +-|__|
'--->|----+
1N4148 |
|
--- 0.05 uF
---
|
GND

created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

Its meant to indicate lock of a 4046; PC1 goes to pin 1, and PC2 goes
to pin 2. Thus, it works no matter what phase comparator you are
using. The PC1 or PC2 input goes high, and the little circuit between
the first two NOR gates acts like a filter, filtering out the
transistions caused by the phase comparators. I'm guessing that the
4046 + digital output is similar to an LM567. The 4046 + 4001 probably
uses far less power, if thats an issue.

Regards,
Bob Monsen
Click to expand...
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top