Maker Pro
Maker Pro

VR-sensor zero-crossing opamp input conditioner circuit

T

tekniq

Jan 1, 1970
0
Hi,

I'm working on a circuit to condition the AC signal from a variable
reluctance crank angle sensor.
The signal waveform amplitude varies from +- 100 mV to +- 50V depending on
RPM.
The output must be a 0V-5V square wave and switch on the negative going zero
crossing.

I'm using a standard TL082 opamp as a comparator with positive feedback
resistors R10 R11 to provide say 100mV hysterisis, I can set the trip-point
on the - input of the comp. with a trimpot.

This is the part i'm not too sure about.
As the opamp rail-rail voltage is 0-5V the imput signal must stay in
between.
I eliminate the negative half of the wave with a standard 1N4004 rectfier
diode and clamp it to max 5V with a zener diode and resistor.
I added another resistor R8 and capacitor C6 to filter the noise on the
input? but i doesn't look right to me. maybe i should put in an AC coupling
cap instead.
Also i don't know if the input resistor R8 conficts with the hysterisis
feedback.

Please examine the circuit on http://home.orange.nl/~tekniq/vr_input.jpg and
give suggestions to improve this cicruit and to determine component values.

Regards,

Martin
 
P

Phil Allison

Jan 1, 1970
0
"tekniq"
I'm using a standard TL082 opamp as a comparator with positive feedback
resistors R10 R11 to provide say 100mV hysterisis, I can set the
trip-point on the - input of the comp. with a trimpot.

This is the part i'm not too sure about.
As the opamp rail-rail voltage is 0-5V the imput signal must stay in
between.

http://home.orange.nl/~tekniq/vr_input.jpg


** The TL082 is not specified to work with such a low DC supply voltage.

IME, it may simply fail to operate.

The TL062 is a better bet with low rail voltages - the specs show some sort
of operation down to +/- 2 volts.

But then it can only swing about +/- 1 volt.

Better go for a type that is designed for single, 5 volt rail operation and
can swing rail to rail.



....... Phil
 
T

tekniq

Jan 1, 1970
0
I made a typo its the TL084 i want to use.
I looked in the datasheet but don't fully understand the specs.

-It states "Wide common-mode (up to VCC+) and differential voltage range"

In the electical characteristics table I find:
-Input common Mode Voltage Range: min +/-11V
-Output Voltage Swing: min 10 to 12V

Does that mean the imput has to be min +/- 11 V
and output min. +/- 10V?
 
P

Phil Allison

Jan 1, 1970
0
"tekniq"
"Phil Allison"

I made a typo its the TL084 i want to use.


** Same comments apply.

I looked in the datasheet but don't fully understand the specs.


** That I am convinced is true.

-It states "Wide common-mode (up to VCC+) and differential voltage range"

In the electical characteristics table I find:
-Input common Mode Voltage Range: min +/-11V
-Output Voltage Swing: min 10 to 12V

Does that mean the imput has to be min +/- 11 V
and output min. +/- 10V?


** All those specs are for /- 15 volt supply - right ?




....... Phil
 
J

Jan Panteltje

Jan 1, 1970
0
I'm using a standard TL082 opamp as a comparator with positive feedback
resistors R10 R11 to provide say 100mV hysterisis, I can set the trip-point
on the - input of the comp. with a trimpot.

This is the part i'm not too sure about.
As the opamp rail-rail voltage is 0-5V the imput signal must stay in
between.
I eliminate the negative half of the wave with a standard 1N4004 rectfier
diode and clamp it to max 5V with a zener diode and resistor.
I added another resistor R8 and capacitor C6 to filter the noise on the
input? but i doesn't look right to me. maybe i should put in an AC coupling
cap instead.
Also i don't know if the input resistor R8 conficts with the hysterisis
feedback.

Please examine the circuit on http://home.orange.nl/~tekniq/vr_input.jpg and
give suggestions to improve this cicruit and to determine component values.

Yes, use a cap on the input of the opamp to AC couple.
But more then that, use a cap in the INPUT, and short out R9, and remove D9.
The zener will then cause a -0.7 to + zener voltage.




A
sensor ---||------------- R ---------||--- R11
|k |
zener === rest same
|a |
------------------------------------------


For low signals A carries the full sensor signal, AC coupled to R11
For very high signals (speeds if inductive sensor) the voltage in A is
limited by the zener.

There may be other and better ways to do this.
Is the sensor grounded on one side? if not go differential to reduce noise,
with offset compensation.

-------- R5-----
limiter | |
| | \ |
-----------------------------R1 -----------|+ \ |
| | | | \ |
| k a | \________|___
sensor diode diode | /
| a k | /
| | | | /
----------------------------R2-------------|- /
| | | /
R3 R4
| |
+ ///


R2 should equal R1,
R3 and R4 should each be 2 x R5.

This right?
ehhh
Oh well.
<disclaimer>
 
T

tekniq

Jan 1, 1970
0
Thanks for pointing this out to me using opamps is relativly new for me.

Would the LM324 / LM358 be suitable, it can have single supply min 3V, input
can go rail to rail, and has no min diff. input voltage in the datasheet.

Regards,

Martin
 
P

Phil Allison

Jan 1, 1970
0
"tekniq

** Please stop Top Posting !!
Thanks for pointing this out to me using opamps is relativly new for me.

Would the LM324 / LM358 be suitable, it can have single supply min 3V,
input can go rail to rail, and has no min diff. input voltage in the
datasheet.


** Maybe.

Depends on the max frequency the pulses will be are arriving.




....... Phil
 
T

tekniq

Jan 1, 1970
0
Jan Panteltje said:
Yes, use a cap on the input of the opamp to AC couple.
But more then that, use a cap in the INPUT, and short out R9, and remove
D9.
The zener will then cause a -0.7 to + zener voltage.




A
sensor ---||------------- R ---------||--- R11
|k |
zener === rest same
|a |
------------------------------------------


For low signals A carries the full sensor signal, AC coupled to R11
For very high signals (speeds if inductive sensor) the voltage in A is
limited by the zener.

There may be other and better ways to do this.
Is the sensor grounded on one side? if not go differential to reduce
noise,
with offset compensation.

-------- R5-----
limiter | |
| | \ |
-----------------------------R1 -----------|+ \ |
| | | | \ |
| k a | \________|___
sensor diode diode | /
| a k | /
| | | | /
----------------------------R2-------------|- /
| | | /
R3 R4
| |
+ ///


R2 should equal R1,
R3 and R4 should each be 2 x R5.

This right?
ehhh
Oh well.
<disclaimer>

Thanks for the reply's

I have corrected the schematic. I haven't got much exp. with opamps circuits
yet, so ur help is very apreciated.

I replaced the opamp with a LM324, it has a supply voltage of min 3V, input
and output can swing rail to rail, which should be suitable.

I added an AC-coupling cap and corectly reversed the D9 diode, also i
removed the C6 capacitor, this should remove the negative half and clamp the
positive voltage to the 5.1 zener level, right?

Below is the new schematic, i included the estmated component values, is
this about right?

One side of the pickup is connected to ground the other goes into the
circuit input.

The frequenty of the waveform is expected to vary between 220 and 5000 Hz.
Below is a picture of the waveform which i captured with my laptop.

I don't realy understand what is involved with designing the signal input
side of the circuit, maybe someone can explain this a bit?

-I only have to clamp the positive rail with the zener because D9 gets rid
of the negative?

-The coupling cap blocks DC voltage and acts as a lowpass filter?
-Do resistor r8 and C6 also function as a filter?
-Then what is the use for R8?

-By removing the current limiting resistor from the zener doesn't it get too
much current?

-What components do i need to include and omit?

http://home.wanadoo.nl/~tekniq/vr_input2.jpg
http://home.wanadoo.nl/~tekniq/vr_waveform.jpg
 
R

Rich Grise

Jan 1, 1970
0
[crossposted to s.e.basics, where the question really belongs]


I'd lose D9; just put C6 directly to R8. The zener will work; my
personal preference would be diode clamps to the rails:

+5
|
| K
[diode]
|
input ----[C6]----[R8]----+--+----[R11]---etc.
|
| K
[diode]
|
[GND]

Something like 1N4148 or 1N914 would be faster switching, but, lessee -
with 50V peak at the end of a 1K, that's like 50 mA, which might be
a bit much ... Oh, sure. The 1N4148 is good for 150 mA:
http://www.rectron.com/data_sheets/1n4148.pdf

I'd put the diodes "after" R8, because I'd want the signal to see a
resistance before the diode shorts it out.

Also, put a 1K or 2.2K pullup directly from IC5A-1 (the output) to
Vcc. The LM324 is a comparator, with an open-collector output.

Of course, your supply is protected from load dumps and all that,
right?

Good Luck!
Rich
 
T

tekniq

Jan 1, 1970
0
Jan Panteltje said:
Yes, use a cap on the input of the opamp to AC couple.
But more then that, use a cap in the INPUT, and short out R9, and remove
D9.
The zener will then cause a -0.7 to + zener voltage.




A
sensor ---||------------- R ---------||--- R11
|k |
zener === rest same
|a |
------------------------------------------


For low signals A carries the full sensor signal, AC coupled to R11
For very high signals (speeds if inductive sensor) the voltage in A is
limited by the zener.

There may be other and better ways to do this.
Is the sensor grounded on one side? if not go differential to reduce
noise,
with offset compensation.

-------- R5-----
limiter | |
| | \ |
-----------------------------R1 -----------|+ \ |
| | | | \ |
| k a | \________|___
sensor diode diode | /
| a k | /
| | | | /
----------------------------R2-------------|- /
| | | /
R3 R4
| |
+ ///


R2 should equal R1,
R3 and R4 should each be 2 x R5.

This right?
ehhh
Oh well.
<disclaimer>

Thanks for the reply's

I have corrected the schematic. I haven't got much exp. with opamps circuits
yet, so ur help is very apreciated.

I replaced the opamp with a LM324, it has a supply voltage of min 3V, input
and output can swing rail to rail, which should be suitable.

I added an AC-coupling cap and corectly reversed the D9 diode, also i
removed the C6 capacitor, this should remove the negative half and clamp the
positive voltage to the 5.1 zener level, right?

Below is the new schematic, i included the estmated component values, is
this about right?

One side of the pickup is connected to ground the other goes into the
circuit input.

The frequenty of the waveform is expected to vary between 220 and 5000 Hz.
Below is a picture of the waveform which i captured with my laptop.

I don't realy understand what is involved with designing the signal input
side of the circuit, maybe someone can explain this a bit?

-I only have to clamp the positive rail with the zener because D9 gets rid
of the negative?

-The coupling cap blocks DC voltage and acts as a lowpass filter?
-Do resistor r8 and C6 also function as a filter?
-Then what is the use for R8?

-By removing the current limiting resistor from the zener doesn't it get too
much current?

-What components do i need to include and omit?

http://home.wanadoo.nl/~tekniq/vr_input2.jpg
http://home.wanadoo.nl/~tekniq/vr_waveform.jpg


Rich Grise said:
[crossposted to s.e.basics, where the question really belongs]


I'd lose D9; just put C6 directly to R8. The zener will work; my
personal preference would be diode clamps to the rails:

+5
|
| K
[diode]
|
input ----[C6]----[R8]----+--+----[R11]---etc.
|
| K
[diode]
|
[GND]

Something like 1N4148 or 1N914 would be faster switching, but, lessee -
with 50V peak at the end of a 1K, that's like 50 mA, which might be
a bit much ... Oh, sure. The 1N4148 is good for 150 mA:
http://www.rectron.com/data_sheets/1n4148.pdf

I'd put the diodes "after" R8, because I'd want the signal to see a
resistance before the diode shorts it out.

Also, put a 1K or 2.2K pullup directly from IC5A-1 (the output) to
Vcc. The LM324 is a comparator, with an open-collector output.

Of course, your supply is protected from load dumps and all that,
right?

Good Luck!
Rich

Thanks

I'll continue my posts here in basics where i still belong (sigh :))
I find your comments very useful, i thought by restricting the signal to
+0.6 V would be a bit small and more susceptible to noise, but as the
comparator swiches at very low signal levels this should make no difference.
As for the current tru the diodes, i could make the 1K resistor a lot bigger
(i have just estamated some component values) but i thought this would
interfere with the hysterisis feedback loop?

The 1N4001 i choose because i still have a lot of them, I would like to know
the swiching speed of these but none of the datasheets states them( guess
its that bad)
So faster diodes it is.

I'm still a bit in the dark about the input filters and how and when to use
what. As i see it al i have left now is a coupling cap and a input resistor.
What are the functions of the input resistor here? It limits current trough
the diodes but otherwise it seems redundent.
Is there still something missing here? some improvements maybe?

My supply is bypassed with a 10uF elco and a 100n cap, does it need more
protection against load dumps from what?

Very curious about these types of circuit, would like to know more.

Regards,

Mrtin
 
T

tekniq

Jan 1, 1970
0
tekniq said:
Jan Panteltje said:
Yes, use a cap on the input of the opamp to AC couple.
But more then that, use a cap in the INPUT, and short out R9, and remove
D9.
The zener will then cause a -0.7 to + zener voltage.




A
sensor ---||------------- R ---------||--- R11
|k |
zener === rest same
|a |
------------------------------------------


For low signals A carries the full sensor signal, AC coupled to R11
For very high signals (speeds if inductive sensor) the voltage in A is
limited by the zener.

There may be other and better ways to do this.
Is the sensor grounded on one side? if not go differential to reduce
noise,
with offset compensation.

-------- R5-----
limiter | |
| | \ |
-----------------------------R1 -----------|+ \ |
| | | | \ |
| k a | \________|___
sensor diode diode | /
| a k | /
| | | | /
----------------------------R2-------------|- /
| | | /
R3 R4
| |
+ ///


R2 should equal R1,
R3 and R4 should each be 2 x R5.

This right?
ehhh
Oh well.
<disclaimer>

Thanks for the reply's

I have corrected the schematic. I haven't got much exp. with opamps
circuits
yet, so ur help is very apreciated.

I replaced the opamp with a LM324, it has a supply voltage of min 3V,
input
and output can swing rail to rail, which should be suitable.

I added an AC-coupling cap and corectly reversed the D9 diode, also i
removed the C6 capacitor, this should remove the negative half and clamp
the
positive voltage to the 5.1 zener level, right?

Below is the new schematic, i included the estmated component values, is
this about right?

One side of the pickup is connected to ground the other goes into the
circuit input.

The frequenty of the waveform is expected to vary between 220 and 5000 Hz.
Below is a picture of the waveform which i captured with my laptop.

I don't realy understand what is involved with designing the signal input
side of the circuit, maybe someone can explain this a bit?

-I only have to clamp the positive rail with the zener because D9 gets rid
of the negative?

-The coupling cap blocks DC voltage and acts as a lowpass filter?
-Do resistor r8 and C6 also function as a filter?
-Then what is the use for R8?

-By removing the current limiting resistor from the zener doesn't it get
too
much current?

-What components do i need to include and omit?

http://home.wanadoo.nl/~tekniq/vr_input2.jpg
http://home.wanadoo.nl/~tekniq/vr_waveform.jpg


Rich Grise said:
[crossposted to s.e.basics, where the question really belongs]


I'd lose D9; just put C6 directly to R8. The zener will work; my
personal preference would be diode clamps to the rails:

+5
|
| K
[diode]
|
input ----[C6]----[R8]----+--+----[R11]---etc.
|
| K
[diode]
|
[GND]

Something like 1N4148 or 1N914 would be faster switching, but, lessee -
with 50V peak at the end of a 1K, that's like 50 mA, which might be
a bit much ... Oh, sure. The 1N4148 is good for 150 mA:
http://www.rectron.com/data_sheets/1n4148.pdf

I'd put the diodes "after" R8, because I'd want the signal to see a
resistance before the diode shorts it out.

Also, put a 1K or 2.2K pullup directly from IC5A-1 (the output) to
Vcc. The LM324 is a comparator, with an open-collector output.

Of course, your supply is protected from load dumps and all that,
right?

Good Luck!
Rich

Thanks

I'll continue my posts here in basics where i still belong (sigh :))
I find your comments very useful, i thought by restricting the signal to
+0.6 V would be a bit small and more susceptible to noise, but as the
comparator swiches at very low signal levels this should make no
difference.
As for the current tru the diodes, i could make the 1K resistor a lot
bigger (i have just estamated some component values) but i thought this
would interfere with the hysterisis feedback loop?

The 1N4001 i choose because i still have a lot of them, I would like to
know the swiching speed of these but none of the datasheets states them(
guess its that bad)
So faster diodes it is.

I'm still a bit in the dark about the input filters and how and when to
use what. As i see it al i have left now is a coupling cap and a input
resistor.
What are the functions of the input resistor here? It limits current
trough the diodes but otherwise it seems redundent.
Is there still something missing here? some improvements maybe?

My supply is bypassed with a 10uF elco and a 100n cap, does it need more
protection against load dumps from what?

Very curious about these types of circuit, would like to know more.

Regards,

Mrtin

Hi

Another thing when i come to think about it; by putting the diodes rail to
rail the signal would still be -0.6v <-> +0.6v which is below the negative
(GND) supply rail of the comparator this is unacceptable according the
datasheets, the same happens with the zeners. I thought the diodes would
take care of this, but know?

+5
|
| K
[diode]
|
input ----[C6]----[R8]----+-------[R11]---etc.
|
| K
[diode]
|
[GND]
 
R

Rich Grise

Jan 1, 1970
0
[followups-to is now only s.e.basics]
.
Another thing when i come to think about it; by putting the diodes rail
to rail the signal would still be -0.6v <-> +0.6v which is below the
negative (GND) supply rail of the comparator this is unacceptable
according the datasheets, the same happens with the zeners. I thought
the diodes would take care of this, but know?

+5
|
| K
[diode]
|
input ----[C6]----[R8]----+-------[R11]---etc.
|
| K
[diode]
|
[GND]

It's always worked for me, but I've never really subjected an input
to such drastic swings.

The logical choice, then, would be Schottky diodes, but someone else
would have to come up with a part number. A google search on Schottky
diodes might be fruitful.

Good Luck!
Rich
 
J

Jan Panteltje

Jan 1, 1970
0
On a sunny day (Tue, 27 Feb 2007 22:15:07 +0100) it happened "tekniq"

C1 A
One side of the pickup is connected to ground the other goes into the
circuit input.
-By removing the current limiting resistor from the zener doesn't it get too
much current?

The peak current is limited by the inductance of the pickup coil, and its resistance,
in series with the impedance of C1.

-What components do i need to include and omit?

Just try above circuit :)

Are those VOLTS? In that case you have only 200mV pp, and the whole issue changes.

Then I would perhaps do it like this:

-- R3 ----------
DC block | |
CD block lowpass C3 | |
sensor ---||------------- R1 ---------||--------------- + |
C1 |a |k | | opamp -------
diode diode === C2 bias R2
|k |a | |---- -
GND - /// /// /// |
|
1/2 V (from you trim pot)

Added R2, to make sure the + input of the opamp is closer then 100mV to 1/2V too,
else it wont trigger.
If you make R2 say 10 kOhm, then R3 should be very very high,
because if the output flips to +12 (supply) thene you have a divider R3 / (R2 + R3) that
should yield less then 100 mV, so R3 should be at least 60 x R2, or better 1 MOhm.
 
P

Peter Bennett

Jan 1, 1970
0
Hi,

I'm working on a circuit to condition the AC signal from a variable
reluctance crank angle sensor.
The signal waveform amplitude varies from +- 100 mV to +- 50V depending on
RPM.
The output must be a 0V-5V square wave and switch on the negative going zero
crossing.

I'm using a standard TL082 opamp as a comparator with positive feedback
resistors R10 R11 to provide say 100mV hysterisis, I can set the trip-point
on the - input of the comp. with a trimpot.

If you want a comparator, USE A BLOODY COMPARATOR!!

You can find comparators that are designed to work on 5 volts, and to
give a logic level output.

--
Peter Bennett, VE7CEI Vancouver BC, Canada
peterbb4 (at) interchange.ubc.ca
new newsgroup users info : http://vancouver-webpages.com/nnq
GPS and NMEA info: http://vancouver-webpages.com/peter
Vancouver Power Squadron: http://vancouver.powersquadron.ca
 
Top