Hi
Just wondered if anyone else has come across this problem, i have a 02 1.6v stilo with the engine fault light on, i have been told the code's it is saying are P1325 and P0325. Does anyone know what this means, can still drive but the car loses power quite quickly......any advice would be great.

Both codes are saying something is wrong with the knock sensor circuit



Knock sensor is a simple little device attached to the cylinder block but modern engines are lost without it. Engines run best if the ignition for each cylinder is just retarded a fraction from the engine knocking. So the ECU advances the ignition up to this limit intil the knock sensor says "Whoa! That's enough" for each cylinder. Now without feedback from the knock sensor it is lost as to what to do

So you'd need to investigate just where the problem is in the circuit from the ECU through the knock sensor connector and on to the knock sensor itself. First place I'd go would be to clean the knock sensor connector contacts with electrical contact cleaner, then the ECU connector

Have to do some electrical circuit checking after that. K50 is the knock sensor and it sends signals to pin 41 and pin 48 at the ECU M10

Hi
Many thanx for the info, it looks like it may be the ECU but is there any way of making sure that it is the ECU without changing it, and where would you suggest i get one from and does it have to be new. Are there any other things to check before changing the ECU...
Many Thanx Andrea

First place I'd go would be to clean the knock sensor connector contacts with electrical contact cleaner, then the ECU connector

Have you done those jobs?

Hi
Just thought i would let you know that it cost £50 to get someone to check the ECU which turns out to be at fault, another one has cost me £219 so here's hoping it works...fingers crossed....
Thanx for your help much appreciated.
Andrea

Many ECU 'faults' are actually poor contacts on the connectors to the ECU.

Did you try cleaning the connectors?

I am following this thread with great interest, we have a 2002 1.6 16V Stilo here with basically the same problem: a knock sensor circuit fault that after clearing reappears after anything from 100 meters to several thousand kms of driving causing limp home mode with power loss.

Already tried the lot: renewing the knock sensor, cleaned/replaced connectors, bypassed cabling, checked all other sensors, did testdrives with the examiner connected to check for irregularities (none) and even tried a brand new ecu. But so far no luck :bang:

Found a few other threads here and elsewhere on the net describing similar problems but none lists a clear solution (except one that had it solved with a new ecu, which didn't work for us).

So we're running out of options. At this point our thought are even going towards a mechanical problem in the engine, maybe a piston with too much clearing which is making noises that the ecu mistakenly interprets as knocking :shakehead:. Engine is a bit rough...

9000 said:

So we're running out of options. At this point our thought are even going towards a mechanical problem in the engine, maybe a piston with too much clearing which is making noises that the ecu mistakenly interprets as knocking :shakehead:. Engine is a bit rough...

Click to expand...

Not had the cam belt replace recently (as in just before the problem started) ?

Very common cause of issues if not fitted properly.

Or is it burning oil if so it may be dieseling
with pre ingition and setting the knock sensore off.

Both surfaces on the knock sensor and block need
to be metal clean.

John

Thanks for the quick replies!

Argonought said:

Not had the cam belt replace recently (as in just before the problem started) ?

Very common cause of issues if not fitted properly.

Click to expand...

We did check cambelt timing (using the official fiat tools) and that is al perfect.

sixpotman said:

Or is it burning oil if so it may be dieseling
with pre ingition and setting the knock sensore off.

Both surfaces on the knock sensor and block need
to be metal clean.

John

Click to expand...

Yes, we cleaned them .. Sensor was replaced by a new one and torqued down with specified torque of course.


By now dropped the sump and it turned out that bearings are all ok, but that there is quite a lot of piston play. That's also rather good audible when the engine is cold so could be we're on to something here...

i think you'll by going up a dead end with looking for physical engine problems. P0325 and particularly P1325 are knock sensor electrical circuit problems. Circuit is open circuit (most likely) or short circuit

I'd disconnect the knock sensor and check for continuity between each connector pin there and the ecu connector pins. If that's ok then i'd check for continuity through the knock sensor (it's a very high resistance so a bit difficult to measure at 4MOhms but there should be a circuit there). If that's ok then I'd check for short circuit at the ecu with it all connected. If that's ok you then have to start wondering whether the ecu connector pins are actually making electrical contact with the ECU and ECU circuit board itself

Because it's an electrical sensitive circuit it needs to be shielded from electrical interference with an earthed braided sheath taken to earth to prevent electrical noise

We did do all those checks, and then rechecked them again... We even connected the sensor with a separate shielded cable directly to the ecu to rule out all wiring issues. We disassembled the ECU connector to check the contacts, all clean, cleaned again and in perfectly good health. We measured for shorts and continuity in all directions (while wiggling all cables in all directions as well ). Checked the ground connections of engine and ecu. None of this revealed even the faintest hint of a wiring problem.

I fully agree that the fault codes indicate some electrical circuit problem, but it just doesn't seem to be there. So the next question is how the ecu determines there is a wiring problem. Does it actually contain a measuring circuit at the knock sensor input that is capable of measuring leakage resistances to ground and/or sensor resistance (which by the way is around 560kohm not 4 MOhm, both old and new sensor). Such is the case with SRS and ABS circuits, but I don't believe an ecu has such additional circuitry for something trivial as a knock sensor. Having a background in electronics I can tell that it would be quite a challenge to implement such circuitry given the high (and unpredictable) impedance of a piezo transducer and the high noise levels on the signals. But I could be wrong...

Anyway, I merely start to believe the ecu simply analyses the knocksignal taking in account the engine rpm and load, and if it is out of some specified range for some specified time, it assumes there must be a wiring problem and pops out this code. So the better question would probably be: what did the guy that wrote the software had in mind at the time . Oh well, we'll probably never know.

Hi
The 560K ohms you're measuring is the shunt resistor which is good because it means the resistor is present and correct. The knock sensor peizo is almost open circuit figures being around 4-5Mohms


Have a look at this from Pico Tech which gives some insight into how the ecu determines a knock sensor fault at start up even without the engine running by using sine wave and carrier voltage (although they don't sound too sure themselves)
http://www.picoauto.com/tutorials/knock-sensor.html

Thanks for the link, that's excellent reading!

Seems the Fiat ecu is indeed doing some diagnostic checks before engine start, as switching the ignition on with a disconnected or shorted sensor immediately triggers a fault. Could be the purpose of this built-in resistor is to facilitate this.

However in our case the CE light comes on during driving. So now the question is: does the ecu continue to diagnose the electrical merits of the circuitry in this way while the engine is running? Well, that's something a scope should be able to tell me :chin:

Btw one of the reasons we believe mechanical noise of the engine could be disturbing the knock signals is the fact that we sometimes see the ecu detecting knock and pulling ignition even with a (cold) engine just idling without any load.

We'll keep on digging

Seems to be solved! As often the case with complex problems there were multiple causes:

By using a scope to monitor the diagnosis signals the ecu puts on the knock sensor we were able to determine a poor connection in the ecu connector
But when retesting it with the old ecu it turned out that was faulty as well. So solution was the new ecu and another connector.

Just a nice theory of what probably happened: Old ecu failed somewhere triggering a knock sensor fault. A few desperate previous owners took the car to even more desperate mechanics in order to try to solve the problem (that has hunted the car for a very long time). During that process the ECU connector was disconnected and reconnected countless times, which was more than the (rather crappy) construction could handle.
So when finally the culprit was changed for a good ecu, the fault was still there but now for another reason :bang:

All thanks for your input!

Or another possibility is the other way round: the poor connection in the knock sensor wiring caused spikes on the ecu input killing it.

I guess we'll never know for sure which was first...

Glad you got it sorted and mystery solved. The EOBD was right all the time, "knock sensor circuit is open circuit"
Intermittent faults are fun though aren't they