amplifying signal from crank position sensor

[kell]

My 1989 Jeep Cherokee started having an intermittent problem where it
dies sometimes when I'm driving and is getting harder to start..
After reading this page:
http://www.lunghd.com/Tech_Articles/Engine/Basic_Sensors_Diagnostics.htm
I tested the crank position sensor (cps) with a voltmeter. It showed
about .4 volts ac or a little less when cranking. That's below the
acceptable voltage range of .5 to .8 volts stated in the article.
The cps is on the bell housing, very difficult to get to. It's
recommended that you lower the transmission a little. You need a
floor jack, which I don't have.
I got the idea of building a circuit to amplify the signal from the
cps.
Some basic facts... there are only two wires coming from the cps.
With the car turned off, the cps is supposed to measure about 275 ohms
(mine measures 230). As I indicated before, the cps generates a small
signal (detectable with a voltmeter set on ac) when the flywheel is
turning and the slots are flying past the sensor.
While I had the connector unplugged, I also tested the pins on the
wiring harness side of the plug. With the ignition key turned on both
pins show 3.37 volts with respect to ground and zero volts across the
pins.
I need some kind of fix, even a temporary one, just to get the car
running. I park on the street and have to move it periodically.
I thought of using a non-inverting op amp circuit, powering it
independently (just a couple of nine volt batteries for now) because
I'm not sure how it would work using the vehicle's electrical system
to supply the power. Can you use a single supply if the op amp is
amplifying a small ac voltage several volts above ground? There
wouldn't really be a fixed ground for the op amp to refer to.

 

[kell]

My 1989 Jeep Cherokee started having an intermittent problem where it
dies sometimes when I'm driving and is getting harder to start..
After reading this page:http://www.lunghd.com/Tech_Articles/Engine/Basic_Sensors_Diagnostics.htm
I tested the crank position sensor (cps) with a voltmeter. It showed
about .4 volts ac or a little less when cranking. That's below the
acceptable voltage range of .5 to .8 volts stated in the article.
The cps is on the bell housing, very difficult to get to. It's
recommended that you lower the transmission a little. You need a
floor jack, which I don't have.
I got the idea of building a circuit to amplify the signal from the
cps.
Some basic facts... there are only two wires coming from the cps.
With the car turned off, the cps is supposed to measure about 275 ohms
(mine measures 230). As I indicated before, the cps generates a small
signal (detectable with a voltmeter set on ac) when the flywheel is
turning and the slots are flying past the sensor.
While I had the connector unplugged, I also tested the pins on the
wiring harness side of the plug. With the ignition key turned on both
pins show 3.37 volts with respect to ground and zero volts across the
pins.
I need some kind of fix, even a temporary one, just to get the car
running. I park on the street and have to move it periodically.
I thought of using a non-inverting op amp circuit, powering it
independently (just a couple of nine volt batteries for now) because
I'm not sure how it would work using the vehicle's electrical system
to supply the power. Can you use a single supply if the op amp is
amplifying a small ac voltage several volts above ground? There
wouldn't really be a fixed ground for the op amp to refer to.

Now that I think of it, the 3.37 volts is ground for the op amp. So
maybe it would work to run it off the vehicle's battery...
Should I do this?

 

[Jim Thompson]

Now that I think of it, the 3.37 volts is ground for the op amp. So
maybe it would work to run it off the vehicle's battery...
Should I do this?

Conduct another test first. While cranking are you getting a spark?

Pull the _ignition_module_ and take it to a shop that can test it.

...Jim Thompson

 

[kell]

Conduct another test first. While cranking are you getting a spark?

Pull the _ignition_module_ and take it to a shop that can test it.

...Jim Thompson
--

Assuming the ignition module is that thing with the coil clipped into
it. I pulled that out. Then I took the coil out of the ignition
module, clipped a cap across the primary terminals and touched 12
volts to it, got a phat spark. So the coil is ok. But there are
apparently some electronics in the module, because it has more than
just two wires going in and there's a cavity under it filled with some
potting material. So now I have to go and find somebody who can test
this module? Sheesh, why can't they make cars a guy can FIX.

 

[kell]

Assuming the ignition module is that thing with the coil clipped into
it. I pulled that out. Then I took the coil out of the ignition
module, clipped a cap across the primary terminals and touched 12
volts to it, got a phat spark. So the coil is ok. But there are
apparently some electronics in the module, because it has more than
just two wires going in and there's a cavity under it filled with some
potting material. So now I have to go and find somebody who can test
this module? Sheesh, why can't they make cars a guy can FIX.- Hide quoted text -

Forgot to mention, no spark during crank.

 

[Jamie]

My 1989 Jeep Cherokee started having an intermittent problem where it
dies sometimes when I'm driving and is getting harder to start..
After reading this page:
http://www.lunghd.com/Tech_Articles/Engine/Basic_Sensors_Diagnostics.htm
I tested the crank position sensor (cps) with a voltmeter. It showed
about .4 volts ac or a little less when cranking. That's below the
acceptable voltage range of .5 to .8 volts stated in the article.
The cps is on the bell housing, very difficult to get to. It's
recommended that you lower the transmission a little. You need a
floor jack, which I don't have.
I got the idea of building a circuit to amplify the signal from the
cps.
Some basic facts... there are only two wires coming from the cps.
With the car turned off, the cps is supposed to measure about 275 ohms
(mine measures 230). As I indicated before, the cps generates a small
signal (detectable with a voltmeter set on ac) when the flywheel is
turning and the slots are flying past the sensor.
While I had the connector unplugged, I also tested the pins on the
wiring harness side of the plug. With the ignition key turned on both
pins show 3.37 volts with respect to ground and zero volts across the
pins.
I need some kind of fix, even a temporary one, just to get the car
running. I park on the street and have to move it periodically.
I thought of using a non-inverting op amp circuit, powering it
independently (just a couple of nine volt batteries for now) because
I'm not sure how it would work using the vehicle's electrical system
to supply the power. Can you use a single supply if the op amp is
amplifying a small ac voltage several volts above ground? There
wouldn't really be a fixed ground for the op amp to refer to.

Hmm, by looking at the specs i'm guess it's a reluctant type of
pick up?
since the ohm reading is low on yours, it can be assumed that
a couple of turns are shorted there by reducing output voltage..
how about using a step up transformer.
something in the audio class should do it.
you can use an pot on one side to shunt out over voltage.

Using something like a BuckBoost configuration..
Radio Shaft has an audio output transformer you could use
by putting the primary leads across the sensor, then, connect
one of the leads of the secondary to one of the primary leads.
the other secondary lead will be a lead going to the computer.
and remaining lead of sensor which is also connected to the
primary of the transformer will go to the computer..
in the old days, Picture tube boosters were like this.
you will have to play with the phase polarity to get it to
boost.
The ratio of this transformer should be close enough to give you
a slight raise in voltage.
the primary has a CT so you can play a bit..
You could also use the 1:1 audio xformer they have with the same
config, that would double it. use a shunt R on one side to drain
out over voltage.
so in your case. 0.4 volts would be 0.8 in theory how ever, roughly
thinking about loss from the sensor due to extra load and all it most
likely will end up to be around .6 volts/

 

[kell]

Hmm, by looking at the specs i'm guess it's a reluctant type of
pick up?
since the ohm reading is low on yours, it can be assumed that
a couple of turns are shorted there by reducing output voltage..
how about using a step up transformer.
something in the audio class should do it.
you can use an pot on one side to shunt out over voltage.

Using something like a BuckBoost configuration..
Radio Shaft has an audio output transformer you could use
by putting the primary leads across the sensor, then, connect
one of the leads of the secondary to one of the primary leads.
the other secondary lead will be a lead going to the computer.
and remaining lead of sensor which is also connected to the
primary of the transformer will go to the computer..
in the old days, Picture tube boosters were like this.
you will have to play with the phase polarity to get it to
boost.
The ratio of this transformer should be close enough to give you
a slight raise in voltage.
the primary has a CT so you can play a bit..
You could also use the 1:1 audio xformer they have with the same
config, that would double it. use a shunt R on one side to drain
out over voltage.
so in your case. 0.4 volts would be 0.8 in theory how ever, roughly
thinking about loss from the sensor due to extra load and all it most
likely will end up to be around .6 volts/

Question: I have in my box of magic junk a pre-wound pot core of a
power ferrite material with a 1:1 ratio, almost an inch and a half in
diameter. I think each winding has 860 uH inductance. Would it work?

 

[Jamie]

Question: I have in my box of magic junk a pre-wound pot core of a
power ferrite material with a 1:1 ratio, almost an inch and a half in
diameter. I think each winding has 860 uH inductance. Would it work?

Nope
you need something with much higher L in it.
You could try it but I don't think it'll work.
something in the H/ high mh range..

 

[kell]

Nope
you need something with much higher L in it.
You could try it but I don't think it'll work.
something in the H/ high mh range..

I typed in "audio transformer" at the Radio Shack site and found one
http://www.radioshack.com/search/in...dio transformer&origkw=audio transformer&sr=1
but they don't seem to give the ratio? I'd like to use 1:1. Don't
see how I could use a 600:1 tranny, which I think is the other typical
ratio for an audio transformer, isn't it? I guess I'll have to go
into the store and look at the package to find out.

 

[kell]

I typed in "audio transformer" at the Radio Shack site and found onehttp://www.radioshack.com/search/index.jsp?kwCatId=&kw=audio%20transf...
but they don't seem to give the ratio? I'd like to use 1:1. Don't
see how I could use a 600:1 tranny, which I think is the other typical
ratio for an audio transformer, isn't it? I guess I'll have to go
into the store and look at the package to find out.- Hide quoted text -

- Show quoted text -

Okay, I found a 1:1 transformer at RS, item 273-1374
600-900 ohms
300Hz to 5KHz response

 

[Tony Williams]

Some basic facts... there are only two wires coming from the cps.
With the car turned off, the cps is supposed to measure about 275
ohms (mine measures 230). As I indicated before, the cps
generates a small signal (detectable with a voltmeter set on ac)
when the flywheel is turning and the slots are flying past the
sensor.

It is a Variable Reluctance Sensor.

While I had the connector unplugged, I also tested the pins on
the wiring harness side of the plug. With the ignition key
turned on both pins show 3.37 volts with respect to ground and
zero volts across the pins.

The amplifier will be a differential amplifier with
about a few k of input resistance. There is probably
an added dc bias current at the inputs that allows
detection of open-circuit sensor, (and sometimes
short-circuit).

It is therefore not a case of simply providing an
extra stage of amplification. Any extra amplifier
will probably have to fake the floating dc resistance
of the VRS. This will be a transformer output, padded
up to about the same 240 ohms output resistance.

I need some kind of fix, even a temporary one, just to
get the car running.

In the event of a 'failed-sensor' detection, (real or
false), the electronics might well cut the engine.

Poor connections can often cause a false failed-sensor.

So the first thing to try is to get rid of that plug
and socket, and do a hardwire, (as suggested in the URL).
Use a three-way screw terminal block.

A VRS has an output impedance that looks like an inductor
and resistor in series. In an industrial app I have seen
the output increased with a capacitor across the output
leads.... of the order of 0.1uF or less.

Perhaps you might try your cranking experiment again,
stabbing various cap values in parallel to see what
the effect is.

 

[kell]

It is a Variable Reluctance Sensor.


The amplifier will be a differential amplifier with
about a few k of input resistance. There is probably
an added dc bias current at the inputs that allows
detection of open-circuit sensor, (and sometimes
short-circuit).

I could be wrong about this, but it seems to me the meter should see a
bit of voltage across the pins in that case, which it doesn't. This
vehicle is from 1989 and may be a bit less modern in that respect.

It is therefore not a case of simply providing an
extra stage of amplification. Any extra amplifier
will probably have to fake the floating dc resistance
of the VRS. This will be a transformer output, padded
up to about the same 240 ohms output resistance.

Well, yesterday I pushed it to get it across the street to a legal
spot. Damn jeep.

In the event of a 'failed-sensor' detection, (real or
false), the electronics might well cut the engine.

Poor connections can often cause a false failed-sensor.

So the first thing to try is to get rid of that plug
and socket, and do a hardwire, (as suggested in the URL).
Use a three-way screw terminal block.

I tried pinching the connectors to make them tighter, but I didn't
want to cut the connector off the wiring harness.

A VRS has an output impedance that looks like an inductor
and resistor in series. In an industrial app I have seen
the output increased with a capacitor across the output
leads.... of the order of 0.1uF or less.

That's a good tip.

Perhaps you might try your cranking experiment again,
stabbing various cap values in parallel to see what
the effect is.

I had another idea of sticking a dime-store magnet on top of the cps
to make the magnetic field stronger. I figure it probably failed
because heat and vibration weakened the magnet. There's certainly
plenty of both at the top of the bell housing, and that vehicle has
200k on it.
The new sensor I bought has the same resistance as the old one, so I
doubt the old one has shorted turns.

Thanks for the info Tony.

 

[Phil Hobbs]

I could be wrong about this, but it seems to me the meter should see a
bit of voltage across the pins in that case, which it doesn't. This
vehicle is from 1989 and may be a bit less modern in that respect.
Well, yesterday I pushed it to get it across the street to a legal
spot. Damn jeep.


I tried pinching the connectors to make them tighter, but I didn't
want to cut the connector off the wiring harness.

That's a good tip.



I had another idea of sticking a dime-store magnet on top of the cps
to make the magnetic field stronger. I figure it probably failed
because heat and vibration weakened the magnet. There's certainly
plenty of both at the top of the bell housing, and that vehicle has
200k on it.
The new sensor I bought has the same resistance as the old one, so I
doubt the old one has shorted turns.

Thanks for the info Tony.

If you use a $5 Nd-Fe-Bo magnet, you can probably remagnetize the thing,
but be careful you do it in the right direction, or you'll screw up the
timing something awful.

Cheers,

Phil Hobbs

 

[Jim Thompson]

If you use a $5 Nd-Fe-Bo magnet, you can probably remagnetize the thing,
but be careful you do it in the right direction, or you'll screw up the
timing something awful.

Cheers,

Phil Hobbs

Uuuuh? Is there really a magnet there, or just a steel "tooth" going
by a VR pickup that has current flowing thru it.

I know for a fact that ignition (distributor) VR pickups have no
magnets.

...Jim Thompson

 

[Phil Hobbs]

Uuuuh? Is there really a magnet there, or just a steel "tooth" going
by a VR pickup that has current flowing thru it.

I know for a fact that ignition (distributor) VR pickups have no
magnets.

...Jim Thompson

Hmm, could be. The ones I've seen (not automotive, admittedly) used
permanent magnets. That way they don't require power.

Cheers,

Phil Hobbs

 

[Jim Thompson]

Hmm, could be. The ones I've seen (not automotive, admittedly) used
permanent magnets. That way they don't require power.

Cheers,

Phil Hobbs

I misspoke, there are no _rotating_ magnets, but there is one
underneath the rotor assembly.

But flywheel pickups don't use magnets, at least those I know of...
there's no convenient "loop".

...Jim Thompson

 

[Phil Hobbs]

I misspoke, there are no _rotating_ magnets, but there is one
underneath the rotor assembly.

But flywheel pickups don't use magnets, at least those I know of...
there's no convenient "loop".

...Jim Thompson

Any magnetic sensor is going to have exactly the same problem...if you
use a bar magnet with a field of 1 kG, and lots of turns, it doesn't
take much to give you a nice big signal. Remagnetizing the magnet with
the wrong polarity is likely to make the ignition fire half a tooth
early or late.

Cheers,

Phil Hobbs

 

[Jim Thompson]

Any magnetic sensor is going to have exactly the same problem...if you
use a bar magnet with a field of 1 kG, and lots of turns, it doesn't
take much to give you a nice big signal. Remagnetizing the magnet with
the wrong polarity is likely to make the ignition fire half a tooth
early or late.

Cheers,

Phil Hobbs

Might be worth measuring and see if there's current flow in the pickup
coil.

...Jim Thompson

 

[kell]

It is a Variable Reluctance Sensor.


The amplifier will be a differential amplifier with
about a few k of input resistance. There is probably
an added dc bias current at the inputs that allows
detection of open-circuit sensor, (and sometimes
short-circuit).

It is therefore not a case of simply providing an
extra stage of amplification. Any extra amplifier
will probably have to fake the floating dc resistance
of the VRS. This will be a transformer output, padded
up to about the same 240 ohms output resistance.


In the event of a 'failed-sensor' detection, (real or
false), the electronics might well cut the engine.

Poor connections can often cause a false failed-sensor.

So the first thing to try is to get rid of that plug
and socket, and do a hardwire, (as suggested in the URL).
Use a three-way screw terminal block.

A VRS has an output impedance that looks like an inductor
and resistor in series. In an industrial app I have seen
the output increased with a capacitor across the output
leads.... of the order of 0.1uF or less.

Perhaps you might try your cranking experiment again,
stabbing various cap values in parallel to see what
the effect is.

Okay Tony,

The capacitor trick worked. I tried 0.1 uF first; car started ran
bad. Bunged in a 0.01 uF and ALL'S WELL.

Jim & Phil:
The crankshaft position sensor definitely has a permanent mag. It
attracts a piece of steel.
A passive device like that ought to last the life of the car, by
rights. However, there's a lot of heat and vibration at the sensor
mount on top of the transmission bellhousing. Weakens the magnet as
the engine hours pile up. This thing is a routine maintenance item,
but if it fails out in the middle of nowhere, PITA because your car
won't run without it.
This has given me an idea to sell a little device containing a
capacitor that can be hitched onto the wiring harness with those quick-
snap connectors that pierce the insulation of the wires. Cheap
insurance at a few bucks. I think the off-road crowd would buy them.
You're on top of a mountain somewhere and your jeep won't start
because your cps has 130k on it. Snap the magic doodad on your wires
and away you go.

 

[Marra]

My 1989 Jeep Cherokee started having an intermittent problem where it
dies sometimes when I'm driving and is getting harder to start..
After reading this page:http://www.lunghd.com/Tech_Articles/Engine/Basic_Sensors_Diagnostics.htm
I tested the crank position sensor (cps) with a voltmeter. It showed
about .4 volts ac or a little less when cranking. That's below the
acceptable voltage range of .5 to .8 volts stated in the article.
The cps is on the bell housing, very difficult to get to. It's
recommended that you lower the transmission a little. You need a
floor jack, which I don't have.
I got the idea of building a circuit to amplify the signal from the
cps.
Some basic facts... there are only two wires coming from the cps.
With the car turned off, the cps is supposed to measure about 275 ohms
(mine measures 230). As I indicated before, the cps generates a small
signal (detectable with a voltmeter set on ac) when the flywheel is
turning and the slots are flying past the sensor.
While I had the connector unplugged, I also tested the pins on the
wiring harness side of the plug. With the ignition key turned on both
pins show 3.37 volts with respect to ground and zero volts across the
pins.
I need some kind of fix, even a temporary one, just to get the car
running. I park on the street and have to move it periodically.
I thought of using a non-inverting op amp circuit, powering it
independently (just a couple of nine volt batteries for now) because
I'm not sure how it would work using the vehicle's electrical system
to supply the power. Can you use a single supply if the op amp is
amplifying a small ac voltage several volts above ground? There
wouldn't really be a fixed ground for the op amp to refer to.

Buy a new sensor !