Omega Owners Forum
Chat Area => General Car Chat => Topic started by: Webby the Bear on 14 August 2013, 18:24:43
-
Hi guys,
I have a code 19 for crank sensor. going to replace it but as i'm almost about to go back to school i'm doing some revision and i'd like your opinions on testing it (I know a lot of you said don't bother as I have the code but I am interested in electric diagnostics as it's my worst area...that's why the call it elec-trickery ::)).
This is what I would do based on the small amount I know and what I believe makes sense. just let me know if you think this is correct, where i'm going wrong and what you would do different :)
1. as I know the wire is a big fault possibility due to it being close to the exhaust and oil filter housing I reckon that's a good place to start......if I back probe the connector end of the CPS and back probe the CPS end, touch multimeter leads (with alligator clip extension wires obviously) to the two back probed areas i'm thinking that I should see no resistance (apart from the 1/2 an ohm or so in my leads)?
2.) undo connector. probe connector (that CPS connector attaches to). this is to test that voltage is able to get to the CPS. with ignition on take voltage reading. or would engine need to be on? would engine even start without CPS in place? this would prob set a code if there wasn't one set previously.
3.) back probe two wires at CPS end. with car running touch pins with multimeter, this should give battery voltage, no?
what do we think?
remember i'm not going to do all this as I've got the code and its a known fault..... I am merely testing myself, trying to think logically just if the EM code wasn't there to help.
:y :y
-
You need a scope tbh.
Crank sensor is just a hall sensor
-
There won't be a sensible voltage reading at the crank sensor. The sensor itself generates a small AC voltage that is amplified inside the ECU so the chances of picking anything up using a multimeter are small.
The only thing you can really do with a multimeter is to check that the two active pins of the sensor have the correct resistance between them (about 400 ohms, if memory serves). Even then, it's not conclusive, but when my last one failed, it had gone open circuit.
You can probe the connections for continuity through the loom right back to the ECU, of course, but because it's normally the sensor, this is probably a waste of time on an Omega IMHO.
-
There won't be a sensible voltage reading at the crank sensor. The sensor itself generates a small AC voltage that is amplified inside the ECU so the chances of picking anything up using a multimeter are small.
The only thing you can really do with a multimeter is to check that the two active pins of the sensor have the correct resistance between them (about 400 ohms, if memory serves). Even then, it's not conclusive, but when my last one failed, it had gone open circuit.
You can probe the connections for continuity through the loom right back to the ECU, of course, but because it's normally the sensor, this is probably a waste of time on an Omega IMHO.
Thanks kev,
I thought that most components received a voltage (most components full battery voltage, some 5 or 6 volts) going in to them. so the voltage that goes to the sensor is actually a lot smaller than this amount thus as you say wont be picked up via MM?
Re the resistance check, that's something i'm used to doing but I remember mark damning the use of a resistance check. if I had the Crank sensor off the car could I also check for resistance at either end of the sensor.... with it being a wire that should allow current flow I shouldn't see any resistance other than the aforementioned 1/2 ohm that my leads created?
a continuity test ive used for bulbs.... the MM simply beeps if there is continuity. does this means that electricy is allowed to flow? if so and, just for completeness, would checking for voltage coming out of the loom be just the continuity test? but as you say my check for batt. voltage coming in wouldn't be valid if it only receives a small amount...thus leaving the continuity test???
hope you understand my garbled post :-[ I know what I mean.... trying to convey it is difficult :-[
-
agreed above and you need a voltage log (some recording equipment) if its not completely dead..
-
I read that a TSB was issued in 2001 about 'voltage spikes from the alternator damaging the camshaft sensor. This is prevented by unpicking the wiring loom and re-routing the cam and crankshaft sensor cables together around the front of the airbox, keeping them away from the alternator. ' I wonder if the same spike damages the crank sensor. Or was the crank sensor already routed away from the alternator (I didn't think it was close anyway)?
See here scroll down to 'recalls'
http://www.honestjohn.co.uk/carbycar/vauxhall/omega-1994/?section=bad (http://www.honestjohn.co.uk/carbycar/vauxhall/omega-1994/?section=bad)
-
I read that a TSB was issued in 2001 about 'voltage spikes from the alternator damaging the camshaft sensor. This is prevented by unpicking the wiring loom and re-routing the cam and crankshaft sensor cables together around the front of the airbox, keeping them away from the alternator. ' I wonder if the same spike damages the crank sensor. Or was the crank sensor already routed away from the alternator (I didn't think it was close anyway)?
See here scroll down to 'recalls'
http://www.honestjohn.co.uk/carbycar/vauxhall/omega-1994/?section=bad (http://www.honestjohn.co.uk/carbycar/vauxhall/omega-1994/?section=bad)
Cam and crank sensor wires don't go anywhere near the alternator or airbox on an Omega, so I think that one's a red herring.
-
to test that a component is getting voltage am i right in thinking you back probe the place your looking for voltage. connect one of your MM leads to that. and the other to a known good earth. and that should give you a reading?
-
to test that a component is getting voltage am i right in thinking you back probe the place your looking for voltage. connect one of your MM leads to that. and the other to a known good earth. and that should give you a reading?
Basically, yep.. keep moving further down (or up) the line until you find voltage between there and a known earth and voila, you've narrowed down a break in a wire or faulty plug.
However, as Kevin says, the crank sensor is one of the sensors on the car that is not powered - it doesn't have a 12V feed (or any kind of +ve feed) from anywhere.. It is, in effect, a tiny tiny generator producing it's own alternating current: http://en.wikipedia.org/wiki/Hall_effect_sensor
Another unpowered sensor is (or could be, depending on type) the O2 (lambda) sensors. They receive 12V on the Omega but only for the heater circuit.. the sensor itself generates voltage due to the chemical reaction occurring inside the sensor element. On a four wire sensor you'd have +12V and ground (for the heater resistor) and Signal +ve with Signal gnd. On a three wire the ground connection is shared between both functions and on a two wire sensor there is no heater circuit at all, just a +ve and ground connection coming from the sensor.
Mark will be along in a moment to tell me that it's not called 'ground' on a car since there is no connection to the earth .. but since most everywhere you see automotive electrics & electronics discussed will refer to it as 'ground' .. so will I ;) (More correctly it's a 0V rail, but..)
-
By far the best way to diagnose electrical issues of this nature is to forget what voltages we might or might not expect from the ECU because, without knowing the internals of the ECU, we simply don't know what to expect.
Look on the wiring diagram to determine which ECU pin is connected where and then measure the resistance along the path from sensor to ECU connector to determine if the path through the engine loom is sound, then check for shorts between the connections.
-
The problem is that many of the sensors are not resistors and a hence a resistance check is of no use as they exhibit non linear resistance and/or maybe active.
The only sensor that springs to mind that can be validly checked using a resistance is a temperature sensor.
MAF - No
Pressure sensor - No
Crank sensor - No
Cam sensor - No
O2 sensor - Yes/No (heater - yes, sensor element No althoguh volts may give some indication with an analogue meter)
Air temp sensor - Yes
Coolant temp sensor - Yes
Knock sensor - No
Throttle Position Sensor - Yes (but much harder to do on DBW cars)
-
funnily this talks remind me an adviced/recommended mechanic who tried to measure crank sensor while engine is off ;D
that was long time ago.. :-\
-
And to confirm a car has a ground but not an earth:
A ground is considered to be a 0V reference (e.g. ground plains etc).
An earth is a reference which is at the same (or very close to) the earths surface potential.
Important to remember as the ground on a car could be many thousands of volts above the potential of the earth, this will become more obvious as electric vehicles are used becuase under a fault condition the vehicles chassis ground could beocme live at mains volts.
-
As an example regarding the resistance, try measuring a diode.
If the resistance measuring voltage of the meter is low the you will get a high resistance measurement, if its above 0.6-0.7 then you will get a low one. This will also change depending on which was round you measure it.
So its a BAD way of testing most sensors. :y
-
And to confirm a car has a ground but not an earth:
A ground is considered to be a 0V reference (e.g. ground plains etc).
An earth is a reference which is at the same (or very close to) the earths surface potential.
Important to remember as the ground on a car could be many thousands of volts above the potential of the earth, this will become more obvious as electric vehicles are used becuase under a fault condition the vehicles chassis ground could beocme live at mains volts.
Are you sure? I would expect the car to only ever to become live with respect to the battery negative, never the earth (planet), even in fault conditions. For example if you connect a multimeter to any battery positive and the other lead to the earth you stand on, there would be no reading (or next to no reading). Unless somone has bolted the battery negative lead to the earth (planet)!
I suppose its possible for the negative lead to become detached from the battery and touch the ground (planet) while at the same time the positive lead becomes disconnected and touches the car body, or more likely just the positive lead to become detached and touch the earth (planet). Then the car body might become live with respect to the planet .
-
And to confirm a car has a ground but not an earth:
A ground is considered to be a 0V reference (e.g. ground plains etc).
An earth is a reference which is at the same (or very close to) the earths surface potential.
Important to remember as the ground on a car could be many thousands of volts above the potential of the earth, this will become more obvious as electric vehicles are used because under a fault condition the vehicles chassis ground could become live at mains volts.
Are you sure? I would expect the car to only ever to become live with respect to the battery negative, never the earth (planet), even in fault conditions. For example if you connect a multimeter to any battery positive and the other lead to the earth you stand on, there would be no reading (or next to no reading). Unless somone has bolted the battery negative lead to the earth (planet)!
If refering to the later satement note that this is under a fault condition (as it states).
-
ok, while i really appreciate the input you guys have made... i'm going to confess i'm very lost. :(
to try and get my mind back on track as i really really want to understand it.....i'm going to go back to very basics.....
so we have a component, unless its unique like the crank sensor, it should get a voltage. but unless we know what voltage it should get its pointless looking for it? i suppose as a general rule of thumb we could have a list of certain components that get battery voltage.... cooling fan i believe being an example? so in this instant we could backprobe the loom coming in, touch it to groud and get a reading.... if we dont see 12v or there abouts then we can search for resistance in the wires thus looking for a short (as kev mentioned).
and pretty much from what mark is saying that resistance in a component is pointless because of all the things he mentioned (dont really understand it but thats cos of my lack of understanding; not your explanation :y)
i'm still getting to grips with this but i will get there! :y
-
I think the issue is that you are taking a diagnostic approach that's fine for simple circuits (lighting and motor circuits, perhaps - indicators, wipers, etc) and trying to apply it to circuits associated with the ECU, some of which are far from simple.
Some of the circuits defy normal diagnosis with a multimeter completely (the crank sensor included) beyond verifying that the wires in the loom are making the connection they should be. Some components send digital data to/from the ECU, for example. This scenario is why ECUs report live data, and it's probably as important, on a modern car, to master diagnosis by what live data and stored fault codes tells you as it is using a multimeter IMHO.
Many cars of post-Omega vintage are now doing away with even the simplest circuits and incorporating CAN controlled light clusters, etc. The only way to diagnose these will be using a code reader. That said, their adoption should make problems much less common in the first place. Should. ;)
-
I think the issue is that you are taking a diagnostic approach that's fine for simple circuits (lighting and motor circuits, perhaps - indicators, wipers, etc) and trying to apply it to circuits associated with the ECU, some of which are far from simple.
Some of the circuits defy normal diagnosis with a multimeter completely (the crank sensor included) beyond verifying that the wires in the loom are making the connection they should be. Some components send digital data to/from the ECU, for example. This scenario is why ECUs report live data, and it's probably as important, on a modern car, to master diagnosis by what live data and stored fault codes tells you as it is using a multimeter IMHO.
Many cars of post-Omega vintage are now doing away with even the simplest circuits and incorporating CAN controlled light clusters, etc. The only way to diagnose these will be using a code reader. That said, their adoption should make problems much less common in the first place. Should. ;)
My last company recently bought a new expensive 650tonne injection moulding machine incorporating on board diagnostics, and remote diagnostics (direct to the machine manufacturers). This m/c, in production mode, just kept throwing many and various fault codes, to the point it was totally unacceptable. Problem eventually traced to faulty CAN bus!!
-
I think the issue is that you are taking a diagnostic approach that's fine for simple circuits (lighting and motor circuits, perhaps - indicators, wipers, etc) and trying to apply it to circuits associated with the ECU, some of which are far from simple.
Some of the circuits defy normal diagnosis with a multimeter completely (the crank sensor included) beyond verifying that the wires in the loom are making the connection they should be. Some components send digital data to/from the ECU, for example. This scenario is why ECUs report live data, and it's probably as important, on a modern car, to master diagnosis by what live data and stored fault codes tells you as it is using a multimeter IMHO.
Many cars of post-Omega vintage are now doing away with even the simplest circuits and incorporating CAN controlled light clusters, etc. The only way to diagnose these will be using a code reader. That said, their adoption should make problems much less common in the first place. Should. ;)
I think you've hit the nail on the head there mate.
in that case perhaps its worth making myself a guide as to what a multimeter can look for and diagnose..... off the top of my head battery, alternator (dynamic test), starter motor etc.
anything else that's of the simple variety that I can add to the list.
and in fact thinking about it... if I take the approach of finding out what components can be tested via MM then it'll be easier to search for individual specific MM tests. that make sense? :)