Variable reluctance magnetic pickup signal conditionner

[Jacques St-Pierre]

I need HELP please!

I am trying to build a simple signal conditioner for a magnetic pickup and
it did not work well. I use the National Semiconductor application note
AN-64 figure 12 as model using a LM339 comparator. It work, but its much too
sensible, I obtain a lot too much false trig. Any sugestion can help.

Bye
Jacques

 

[Jim Thompson]

I need HELP please!

I am trying to build a simple signal conditioner for a magnetic pickup and
it did not work well. I use the National Semiconductor application note
AN-64 figure 12 as model using a LM339 comparator. It work, but its much too
sensible, I obtain a lot too much false trig. Any sugestion can help.

Bye
Jacques

Please **elaborate** on your application details.

...Jim Thompson

 

[Joerg]

Hi Jacques,

It would help if you could post a schematic. Then we all could see what
can be done. Maybe adding a hysteresis or something like that. Also, is
that magnetic loop located in a noisy environment?

Regards, Joerg

 

[Jacques St-Pierre]

I just need to turn digital a signal from a magnetic pickup coil. I need a
simple 8volts pulse when the pickup sence metal.

So I use the schematic that AN-74 describe, quite simple in fact:

The magnetic pickup is connect between input + & - of the LM339. The + pin
is set to 1/2 Vcc by a simple 10K ohms resistor network and a 10meg resistor
connect the + pin with the output of the LM339. That 10meg is suppose to
square the output, but it's not require for my application. The Vcc is set
to 8 volts and the output is pullup with a 470 ohms resistor in serie with a
led to indicate the output status.

The circuit was not stable, it was driving out high frequency pulse by
himself, so I add a 0.1uf cap between the + & - pin of the LM339. That help
a bit, now at least one of the two circuit I build work, but the other is
alway trun on, but still do a pulse when metal goes in front of the pickup.

So one circuit as an always HIGH output, while the other is always LOW.
Strange, but both are unstable and produce pulse for nothing. The circuit
pickup noise and amplified it.

Maybe it can simply be done with a transistor instead of the LM339, but I
need a good digital output signal 0/8 volts.

Jacques

 

[Jacques St-Pierre]

I did try to post the schematics as a JPEG, but I just got errors, so I send
it directly to Joerg

Bye
Jacques

 

[Fred Bartoli]

I just need to turn digital a signal from a magnetic pickup coil. I need a
simple 8volts pulse when the pickup sence metal.

So I use the schematic that AN-74 describe, quite simple in fact:

The magnetic pickup is connect between input + & - of the LM339. The + pin
is set to 1/2 Vcc by a simple 10K ohms resistor network and a 10meg resistor
connect the + pin with the output of the LM339. That 10meg is suppose to
square the output, but it's not require for my application. The Vcc is set
to 8 volts and the output is pullup with a 470 ohms resistor in serie with a
led to indicate the output status.

The circuit was not stable, it was driving out high frequency pulse by
himself, so I add a 0.1uf cap between the + & - pin of the LM339. That help
a bit, now at least one of the two circuit I build work, but the other is
alway trun on, but still do a pulse when metal goes in front of the pickup.

So one circuit as an always HIGH output, while the other is always LOW.
Strange, but both are unstable and produce pulse for nothing. The circuit
pickup noise and amplified it.

Maybe it can simply be done with a transistor instead of the LM339, but I
need a good digital output signal 0/8 volts.

Jacques

It can't work like this. You're missing at least a bypass cap on the minus
entry.
Depending on the frequency and your sensor inductance you can use from a few
nF to a few uF.
If frequency is low, you may also have to have another bias network on this
entry too.

 

[Fred Bartoli]

"Fred Bartoli"

It can't work like this. You're missing at least a bypass cap on the minus
entry.
Depending on the frequency and your sensor inductance you can use from a few
nF to a few uF.
If frequency is low, you may also have to have another bias network on this
entry too.

Or you could do better moving the bias network from the + entry to the -
entry.
Of course leave the feedback resistor on the + entry. Also 10M seems a high
value.

 

[Fred Bartoli]

"Fred Bartoli"

It can't work like this. You're missing at least a bypass cap on the minus
entry.
Depending on the frequency and your sensor inductance you can use from a few
nF to a few uF.
If frequency is low, you may also have to have another bias network on this
entry too.

Or you could do better moving the bias network from the + entry to the -
entry.
Of course leave the feedback resistor on the + entry. Also 10M seems a high
value.

 

[Fred Bartoli]

"Fred Bartoli"

It can't work like this. You're missing at least a bypass cap on the minus
entry.
Depending on the frequency and your sensor inductance you can use from a few
nF to a few uF.
If frequency is low, you may also have to have another bias network on this
entry too.

Or you could do better moving the bias network from the + entry to the -
entry.
Of course leave the feedback resistor on the + entry. Also 10M seems a high
value.

 

[CFoley1064]

Subject: Re: Variable reluctance magnetic pickup signal conditionner

From: "Jacques St-Pierre" [email protected]
Date: 8/27/2004 4:52 PM Central Daylight Time
Message-id: <g2OXc.66857$X12.5046@edtnps84>

I did try to post the schematics as a JPEG, but I just got errors, so I send
it directly to Joerg

Bye
Jacques

Nat. Semi. AN-74, Fig. 12
VCC VCC
+ +
| |
.-. |
| | .-.
10K | | | |Pullup
'-' | |
| '-'
| |\| |
..-----------------|-\ | Vout
| | | >----o-----o
C| .----o---o----|+/ |
C| | | | |/| |
C| | | | |
| | | | |
'---' .-. | ___ |
| | '----|___|---'
10K | | 20M
'-'
|
===
GND
created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

To get a good answer, you will need to characterize the inductance and
resistance of the coil. Sorry.

Good luck
Chris


Chris

 

[CFoley1064]

www.tech-chat.de>Subject: Re: Variable reluctance magnetic pickup signal
conditionner

From: "Jacques St-Pierre" [email protected]
Date: 8/27/2004 4:52 PM Central Daylight Time
Message-id: <g2OXc.66857$X12.5046@edtnps84>

I did try to post the schematics as a JPEG, but I just got errors, so I send
it directly to Joerg

Bye
Jacques

Nat. Semi. AN-74, Fig. 12
VCC VCC
+ +
| |
.-. |
| | .-.
10K | | | |Pullup
'-' | |
| '-'
| |\| |
..-----------------|-\ | Vout
| | | >----o-----o
C| .----o---o----|+/ |
C| | | | |/| |
C| | | | |
| | | | |
'---' .-. | ___ |
| | '----|___|---'
10K | | 20M
'-'
|
===
GND
created by Andy´s ASCII-Circuit v1.24.140803 Beta

Chris

 

[John Fields]

I just need to turn digital a signal from a magnetic pickup coil. I need a
simple 8volts pulse when the pickup sence metal.

So I use the schematic that AN-74 describe, quite simple in fact:

The magnetic pickup is connect between input + & - of the LM339. The + pin
is set to 1/2 Vcc by a simple 10K ohms resistor network and a 10meg resistor
connect the + pin with the output of the LM339. That 10meg is suppose to
square the output, but it's not require for my application. The Vcc is set
to 8 volts and the output is pullup with a 470 ohms resistor in serie with a
led to indicate the output status.

The circuit was not stable, it was driving out high frequency pulse by
himself, so I add a 0.1uf cap between the + & - pin of the LM339. That help
a bit, now at least one of the two circuit I build work, but the other is
alway trun on, but still do a pulse when metal goes in front of the pickup.

So one circuit as an always HIGH output, while the other is always LOW.
Strange, but both are unstable and produce pulse for nothing. The circuit
pickup noise and amplified it.

Maybe it can simply be done with a transistor instead of the LM339, but I
need a good digital output signal 0/8 volts.

---
Try this:



8V>--------+----------------------------------------+
| |
[10K] |
SENSOR | |
+--+ | [2K]
| |----+--------+----------[10K]----------|-\ |
| | | | >--+-+--->OUT
| |----+--[100]-|-------+--[10K]--------+-|+/ |
+--+ | | | | |
[1000pF] [1000pF] [10K] +---[2M]--+
| | |
GND>-------+--------+-------+----------------------------->GND

Not knowing more about the sensor it's difficult to say what exactly
the component values should be, but with the values shown and assuming
the sensor has a resistance of 100 ohms, the output should be
quiescently low and go high when the sensor senses a metallic object
and generates an output pulse of at least 100mV. If you need greater
sensitivity decrease the value of the 100 ohm resistor or, if you want
to make the sensitivity variable, make it a rheostat. The 2M
hysteresis resistor will give you about 10mV of hysteresis at the
output transitions and should (hopefully) eliminate chattering for
slow edges out of the sensor, if you have any.

 

[Fred Bloggs]

I just need to turn digital a signal from a magnetic pickup coil. I need a
simple 8volts pulse when the pickup sence metal.

So I use the schematic that AN-74 describe, quite simple in fact:

The magnetic pickup is connect between input + & - of the LM339. The + pin
is set to 1/2 Vcc by a simple 10K ohms resistor network and a 10meg resistor
connect the + pin with the output of the LM339. That 10meg is suppose to
square the output, but it's not require for my application. The Vcc is set
to 8 volts and the output is pullup with a 470 ohms resistor in serie with a
led to indicate the output status.

The circuit was not stable, it was driving out high frequency pulse by
himself, so I add a 0.1uf cap between the + & - pin of the LM339. That help
a bit, now at least one of the two circuit I build work, but the other is
alway trun on, but still do a pulse when metal goes in front of the pickup.

So one circuit as an always HIGH output, while the other is always LOW.
Strange, but both are unstable and produce pulse for nothing. The circuit
pickup noise and amplified it.

Maybe it can simply be done with a transistor instead of the LM339, but I
need a good digital output signal 0/8 volts.

Jacques

No one can answer your question because you are apparently clueless
about what you want to do. The incoherent statement of purpose, "I just
need to turn digital a signal from a magnetic pickup coil" says it all-
total confusion. That circuit from AN-74 is intended for tachometer
applications where the comparator is right on top of the pickup and both
are enclosed in a housing shielded from stray fields. The sensitivity is
very high, something like 200V/mV, and the comparator output polarity is
indeterministic and unimportant because its job is to produce a pulse
train of *frequency* proportional to rotational speed. The circuit is a
*speed-to-frequency* converter and not a presence detector, it is
unsuitable for detecting presence of a metal object in a general situation.

 

[Joerg]

Hi Jacques,

Didn't receive the schematic but looked up NS AN-74 figure 12. If noise
is a problem you may want to increase the hysteresis via R3. Also, you
could lowpass filter the coil signal if you suspect RF spikes.

One frequent problem with these "differential" setups is that they
aren't always differential enough and the comparator gets hit with way
too much common mode noise if the two wires to the coil are long and not
shielded. But even then a lowpass, maybe just a simple differential RC
filter right before the comparator can help.

BTW, you can't post schematics on this news server but you can on
alt.binaries, schematics. electronic . Or use a web space somewhere.

Regards, Joerg

 

[Jacques St-Pierre]

The application is exactly what you describe, a speed-to-frequency. It will
be use to pick up a diesel motor speed from a rotating gear. So, you are
right, it is not very important that the circuit output a high or low, I
just mention it because it what I saw from both circuit. The major problem
is the fact it pick up noise and generate pulse at the output at high
frequency without any metal in from of the pickup coil.

I have not the luxury of putting the signal conditioner near the pickup
coil. I have to work with what I have, a pickup coil with 3 feet of wire (no
shield) and the signal conditioner at the end.

I did not have time today to implement all your suggestion, but I will
monday morning and let you know what append. I will put some bypass
capacitor (RC filter) at the input and put it in a metal case. I will let
you know what append next.

Bye
Jacques

 

[John Woodgate]

I read in sci.electronics.design that Jacques St-Pierre
about 'Variable reluctance magnetic pickup signal conditionner', on Sat,
28 Aug 2004:

I have to work with what I have, a pickup coil with 3 feet of wire
(no shield) and the signal conditioner at the end.

Twist the two conductors together tightly. That approaches the effect of
a shielded wire.

 

[Joerg]

Hi John,

Try this:



8V>--------+----------------------------------------+
| |
[10K] |
SENSOR | |
+--+ | [2K]
| |----+--------+----------[10K]----------|-\ |
| | | | >--+-+--->OUT
| |----+--[100]-|-------+--[10K]--------+-|+/ |
+--+ | | | | |
[1000pF] [1000pF] [10K] +---[2M]--+
| | |
GND>-------+--------+-------+----------------------------->GND

Not knowing more about the sensor it's difficult to say what exactly
the component values should be, but with the values shown and assuming
the sensor has a resistance of 100 ohms, the output should be
quiescently low and go high when the sensor senses a metallic object
and generates an output pulse of at least 100mV. If you need greater
sensitivity decrease the value of the 100 ohm resistor or, if you want
to make the sensitivity variable, make it a rheostat. The 2M
hysteresis resistor will give you about 10mV of hysteresis at the
output transitions and should (hopefully) eliminate chattering for
slow edges out of the sensor, if you have any.

I'd place a resistor between coil and 1000pF caps on each leg. Or at
least a couple of inductors. He is running the non-shielded cable along
an engine and that can be pretty noisy. The 2M could be taylored to
whatever signal the coil generates so he has max hysteresis protection
against spikes. Maybe to 1/4th of signal swing or so.

Regards, Joerg

 

[Spehro Pefhany]

Hi John,

Try this:



8V>--------+----------------------------------------+
| |
[10K] |
SENSOR | |
+--+ | [2K]
| |----+--------+----------[10K]----------|-\ |
| | | | >--+-+--->OUT
| |----+--[100]-|-------+--[10K]--------+-|+/ |
+--+ | | | | |
[1000pF] [1000pF] [10K] +---[2M]--+
| | |
GND>-------+--------+-------+----------------------------->GND

Not knowing more about the sensor it's difficult to say what exactly
the component values should be, but with the values shown and assuming
the sensor has a resistance of 100 ohms, the output should be
quiescently low and go high when the sensor senses a metallic object
and generates an output pulse of at least 100mV. If you need greater
sensitivity decrease the value of the 100 ohm resistor or, if you want
to make the sensitivity variable, make it a rheostat. The 2M
hysteresis resistor will give you about 10mV of hysteresis at the
output transitions and should (hopefully) eliminate chattering for
slow edges out of the sensor, if you have any.

I'd place a resistor between coil and 1000pF caps on each leg. Or at
least a couple of inductors. He is running the non-shielded cable along
an engine and that can be pretty noisy. The 2M could be taylored to
whatever signal the coil generates so he has max hysteresis protection
against spikes. Maybe to 1/4th of signal swing or so.

Regards, Joerg

IIRC, the output signal of a variable-reluctance sensor is
proportional to speed, so you'd want the hysteresis to be higher at
higher signal levels.

Best regards,
Spehro Pefhany

 

[Fred Bloggs]

I read in sci.electronics.design that Jacques St-Pierre
about 'Variable reluctance magnetic pickup signal conditionner', on Sat,
28 Aug 2004:



Twist the two conductors together tightly. That approaches the effect of
a shielded wire.

Another interesting aspect of this circuit is that the 20M resistor is
not for hysteresis like everyone thinks- its current is strictly
common-mode and produces no hysteresis whatsoever. So what does it do?
Leave it to the practical applications experts at National to use this
resistor to swamp the output transition dV/dt induced current through
the inevitable Cstray to each input- after all, that input is
differential open loop, relatively high impedance, with a gain of
200,000 reacting to mV signal levels -and that output is transitioning
12Vp-p or whatever the original app was. So 20M was chosen so that
d(i(+))=d(Vout)/20M to In(+) swamps d/dt(Vout) x Cstray to In(-). By
similar reasoning, he will need shielded twisted pair to the pickup.
This also happens to be one application where input offset saves the
day, because it is the offset that overdrives the noise in the absence
of pickup signal. Therefore, once he tames pickup noise, he should also
consider applying a deliberate offset to that In(+) terminal through an
arrangement like this:

View in a fixed-width font such as Courier.



+----------------+------------+--<8V
| | |
| / /
/ 390K R
10K / /p
/ \ \
\ | 20M |
| +---/\/\-----+
| | |
| 1K | |\ |
+--------+-/\/\--+----|+\ |
| | | \ |
---\/\/\/\/--|--------+ | >---+---OUT
PU | | /
---/\/\/\/\--|---------------------|-/
| |/
/
10K
/
\
|
|
---

 

[Fred Bloggs]

Hi John,

Try this:



8V>--------+----------------------------------------+
| |
[10K] |
SENSOR | |
+--+ | [2K]
| |----+--------+----------[10K]----------|-\ |
| | | | >--+-+--->OUT
| |----+--[100]-|-------+--[10K]--------+-|+/ |
+--+ | | | | |
[1000pF] [1000pF] [10K] +---[2M]--+
| | |
GND>-------+--------+-------+----------------------------->GND

Not knowing more about the sensor it's difficult to say what exactly
the component values should be, but with the values shown and assuming
the sensor has a resistance of 100 ohms, the output should be
quiescently low and go high when the sensor senses a metallic object
and generates an output pulse of at least 100mV. If you need greater
sensitivity decrease the value of the 100 ohm resistor or, if you want
to make the sensitivity variable, make it a rheostat. The 2M
hysteresis resistor will give you about 10mV of hysteresis at the
output transitions and should (hopefully) eliminate chattering for
slow edges out of the sensor, if you have any.

I'd place a resistor between coil and 1000pF caps on each leg. Or at
least a couple of inductors. He is running the non-shielded cable along
an engine and that can be pretty noisy. The 2M could be taylored to
whatever signal the coil generates so he has max hysteresis protection
against spikes. Maybe to 1/4th of signal swing or so.

Regards, Joerg

IIRC, the output signal of a variable-reluctance sensor is
proportional to speed, so you'd want the hysteresis to be higher at
higher signal levels.

Best regards,
Spehro Pefhany

Why would you expect the noise to be proportional to signal speed?