[Tgo]

Just when I thought of scouring the internet for some info, I came upon this thread here itself.

I am in a very peculiar situation. In my 110 PS Duster @ 170k km, the engine is randomly having the following problems:

• Loss of acceleration, reverting to idle speed
• Taking longer to crank
• Check engine light and injection system fault light glowing

The car is at renault workshop, they have been transparent in sharing all the diagnostics being carried out on the vehicle for the last 10 days. They did the following:

• Checked fault codes
• Checked sensors (probably swapped too)
• Checked wiring

The diagnosis was that the ECM is not able to read engine speed / crank angle sometimes and as a result the related fault codes are being registered. Resulting in temporary limp mode. Since wiring and sensors are not suspect, they replaced the flywheel since that is the mechanical component giving signal to the related sensors. That didn’t solve the problem, so they installed the old flywheel back.

Now they are saying that Renault has suggested to replace the ECM (cost 53-55 k). This is what brings me to this thread.

Though the diagnosis seems to be done following the seemingly correct steps, I can just not fathom that the ECM has failed. This car is run in Rajasthan and there is not even the slightest chance of water damage. Being a sealed part having no moving components, I can’t believe there is something at fault inside it which comes up randomly. If it is a bad sector (like in a hard disk), it will remain bad always and exhibit the problem always. If it is a bad code, can it not be re-installed (flashed) onto the ECM chip? Why isn’t Renault suggesting that? Specially when there is no sign of physical damage to the ECM?

Since I have been in touch with a person to source a salvaged ECM or a cheaper OE replacement ECM, why is he saying it will have to be coded as it comes empty from his source?

I basically need help in understanding how the ECM fails so that it exhibits a particular fault randomly and also in understanding what needs to be done in case I go for a ECM swap (New / salvage / OE replacement)

[Sankar]

Originally Posted by Tgo

The diagnosis was that the ECM is not able to read engine speed / crank angle sometimes and as a result the related fault codes are being registered. Resulting in temporary limp mode. Since wiring and sensors are not suspect, they replaced the flywheel since that is the mechanical component giving signal to the related sensors. That didn’t solve the problem, so they installed the old flywheel back.

Why is the wiring and sensor not suspected? Before replacing the ECU:
1) test the sensor (ask them to refer wshop manual). Check for oxidation and dirt or debris in the sensor harness plug.
2) And test the continuity of the circuit from the sensor to the corresponding pin on the ECU's socket (they will need the wshop manual to identify the pin).
3) Check the ECU pin and the ECU socket pin for any oxidation which will prevent the pins making intermittent contact and throw an error.

Replacing flywheel will not solve this problem because there is nothing to go wrong in a flywheel, the sensor is a hall sensor and reads off a notch or a missing tooth from the flywheel.

Originally Posted by Tgo

Since I have been in touch with a person to source a salvaged ECM or a cheaper OE replacement ECM, why is he saying it will have to be coded as it comes empty from his source?

He means the immobiliser and the injectors mainly. Again this can be done by the workshop if they have the tool, they will need to refer to the section which details the procedure for an ECU change. What all needs to be coded or adapted will be there.

[clevermax]

Coming here from the ECU remaps thread.

I too wonder why the wiring and crank sensor is not suspected in the first place. Though not related, I am having an issue with my Palio's fuel pump right now, and I am beginning to suspect that it is an ECU fault. In my case, I stand to believe that the switching transistor (which commands the Fuel Pump relay) within the ECU has blown, but this suspicion started only after debugging and making sure that the wiring, Fuel pump relay, the pump itself are all in perfectly good condition. I had checked the voltage levels of the wire coming to the relay and found out that there is no command signal coming there but a constant 9.8 volts which suggests a blown transistor. There are no wiring faults.

[sarathlal]

Originally Posted by Tgo

Though the diagnosis seems to be done following the seemingly correct steps, I can just not fathom that the ECM has failed. This car is run in Rajasthan and there is not even the slightest chance of water damage.

Could heat be a culprit with the circuits? Loosely basing on how graphics card pins tend to disconnect from the motherboard when heated up. All will appear back to normal when things cool down.

Heat causing the solders to melt, is one probability for circuits to go kaput without visible exterior damage.

Originally Posted by Sankar

Why is the wiring and sensor not suspected?

Based on the symptoms, this could be the main culprit but would take patient troubleshooting. Loose connections with the adapters are difficult to discover, since under normal conditions everything might appear proper. Before going the ECM replace route, try to analyze all possibilities related to this.

[SPIKE ARRESTOR]

Originally Posted by Tgo

I basically need help in understanding how the ECM fails so that it exhibits a particular fault randomly and also in understanding what needs to be done in case I go for a ECM swap (New / salvage / OE replacement)

A fault registered in the ECU does not imply ECU is faulty (unless it is some hardware related fault inside the ECU), it implies some algorithm inside the ECU has detected an error, this error can be software or hardware related. They could reflash the ECU and clear all the errors, to see if it crops up again.


Since they want to replace the ECU, ask them if they can replace your ECU with a working ECU from some other car, and see if it throws up same errors. If it throws up the same error again, then it is not the ECU.



Spike

[Tgo]

Originally Posted by Sankar

Why is the wiring and sensor not suspected? Before replacing the ECU:
1) test the sensor (ask them to refer wshop manual). Check for oxidation and dirt or debris in the sensor harness plug.
2) And test the continuity of the circuit from the sensor to the corresponding pin on the ECU's socket (they will need the wshop manual to identify the pin).
3) Check the ECU pin and the ECU socket pin for any oxidation which will prevent the pins making intermittent contact and throw an error.

Replacing flywheel will not solve this problem because there is nothing to go wrong in a flywheel, the sensor is a hall sensor and reads off a notch or a missing tooth from the flywheel.



He means the immobiliser and the injectors mainly. Again this can be done by the workshop if they have the tool, they will need to refer to the section which details the procedure for an ECU change. What all needs to be coded or adapted will be there.

Sensor and wires checked multiple times.
Our original flywheel has been reinstalled as it was wrongly diagnosed as faulty.

So basically it is to be done at Renault or someone who has access to tools used and knowledge for this purpose.

Originally Posted by clevermax

but this suspicion started only after debugging and making sure that the wiring, Fuel pump relay, the pump itself are all in perfectly good condition. I had checked the voltage levels of the wire coming to the relay and found out that there is no command signal coming there but a constant 9.8 volts which suggests a blown transistor. There are no wiring faults.

In my case I would have to check what then? The K9K engine uses two sensors, one getting signal from the water pump pulley and another from the missing teeth on the flywheel. Readings of both are in sync and help the ECM determine engine speed and crank position.

Originally Posted by sarathlal

Could heat be a culprit with the circuits? Loosely basing on how graphics card pins tend to disconnect from the motherboard when heated up. All will appear back to normal when things cool down.

Heat causing the solders to melt, is one probability for circuits to go kaput without visible exterior damage.

A view of the pins on the ECM. If something inside is faulty how can that circuit be identified?

Originally Posted by SPIKE ARRESTOR

A fault registered in the ECU does not imply ECU is faulty (unless it is some hardware related fault inside the ECU), it implies some algorithm inside the ECU has detected an error, this error can be software or hardware related. They could reflash the ECU and clear all the errors, to see if it crops up again.


Since they want to replace the ECU, ask them if they can replace your ECU with a working ECU from some other car, and see if it throws up same errors. If it throws up the same error again, then it is not the ECU.

Please explain difference between reflash and clearing the codes. are they analogous to reinstalling windows and repairing windows?

They did install a good ECM on this car but since the problem occurs randomly they could not conclusively say it was solved or not. Also the test drive with the good ECM was done with the new flywheel so they thought the flywheel was the culprit. LATER, they put the original ECM back and it reoccured. Thats when they started focussing on the wiring and sensors for the second time after finally concluding that it is the ECM at fault.

Current symptoms:
Car is having erratic loss of acceleration accompanied by check engine light and the ignition system fault light, sometimes in gear the car does not accelerate until half clutched and revs raised above 2500 rpm. Takes longer to crank sometimes. Sometimes there is no check engine light after start-up. Sometimes it comes and goes without loss of acceleration. Totally unreliable performance.

[a4anurag]

Originally Posted by Tgo

Current symptoms:
Car is having erratic loss of acceleration accompanied by check engine light and the ignition system fault light, sometimes in gear the car does not accelerate until half clutched and revs raised above 2500 rpm. Takes longer to crank sometimes. Sometimes there is no check engine light after start-up. Sometimes it comes and goes without loss of acceleration. Totally unreliable performance.

May not be directly related but how old is the battery on your car? Was it checked during the tests performed by Renault?

[Tgo]

Originally Posted by a4anurag

May not be directly related but how old is the battery on your car? Was it checked during the tests performed by Renault?

Battery is 1.5 year old. Will get it checked to clear that out.

[adilkhan5]

Check or replace O2 sensors

[Tgo]

I was thinking whether an ECU re-map would solve the problem or not?

I may be wrong here but, here's what I think a re-map does:

It re-writes the data tables that the ECM program uses.

So, if there is a problem where the engine output parameters (live feedback) do not change in-line with a change in input parameters, the ECM would register a fault. So there might be 2/3 scenarios here:

A) The data table (ECM software) has one or more corrupt values.

Possible Solution: Re-write the stock data tables?

B) The sector (ECM hardware) containing the value has corrupted.

Possible Solution: Re-write the stock data tables? new value will be written on a good (accessible) sector.

c) The hardware (Sensors, Wiring, Injectors, Pumps etc.) which is responsible to bring out the change gets the input but is unable to deliver the desired output which is read as a fault.

Possible Solution: Check each and every hardware component related to the fault code (in this case, Crank Angle and Engine RPM) for its designed functioning. Which, brings us back to where we started and the problem would not go away even with a new ECM.

Originally Posted by adilkhan5

Check or replace O2 sensors

As far as I know the K9K diesel does not have an O2 sensor.

[scorpian]

Install a new ecm from workshop and take it on long trial drive to see if the problem you face surfaces again, in case it does then ecm can be ruled out and then primary focus would be on each sensor throwing its respective code. Replace them and again drive to see if problems appear. If they still do then i would look at the wiring harness, earthing points etc.

Incase the new ecm does not throw a code, then probably the old ecm has a fault and you could try re installing the software.I think any person who does remaps may also be able to check your ecm to see if there is any issue in the map or its missing some values etc.

[Tgo]

I have been informed that the replacement ECM has arrived at the service center. They have asked to bring the car for installation and subsequent test drives to either solve the problem or do further diagnosis if otherwise.

Originally Posted by scorpian

Incase the new ecm does not throw a code, then probably the old ecm has a fault and you could try re installing the software.I think any person who does remaps may also be able to check your ecm to see if there is any issue in the map or its missing some values etc.

That's exactly what is going to happen now. In-fact all other steps you said have been carried out at the service center regarding wiring, sensors, earthing etc.

The only problem is that the ECM they have arranged as a replacement is on a condition which does not allow me to return it to them (or keep on hold) if the problem is solved. This doesn't allow me to take my original ECM out and get it checked for ascertaining the defect in it or getting it re-mapped, since I will already have been invoiced for the replacement, the moment it makes the car run fine.

Now, what can I do with a spare ECM? ...another thread maybe.

Will update tomorrow.

[scorpian]

Originally Posted by Tgo

I have been informed that the replacement ECM has arrived at the service center. They have asked to bring the car for installation and subsequent test drives to either solve the problem or do further diagnosis if otherwise.
That's exactly what is going to happen now..

Do you access to any person who does remap nearby to you? explain your situation and ask for a trial map to be loaded to rule out the ecm.
I would do the above if i can come to some sort of understanding with a tuner.

If not then you have no option but to opt for the new ecm. The old one well you could use it for other times when you feel like after doing a remap on it if thats possible. Or get it repaired and sell it off.

[Tgo]

Originally Posted by scorpian

The old one well you could use it for other times when you feel like after doing a remap on it if thats possible. Or get it repaired and sell it off.

Have kept the old one for now since I have not settled in with the thought of having shelling out half a lac for it.

Originally Posted by Tgo

Will update tomorrow.

The car is back on the road with the new ECM. Happy with the outcome for now. Related post.

[scorpian]

Originally Posted by Tgo

Have kept the old one for now since I have not settled in with the thought of having shelling out half a lac for it.



The car is back on the road with the new ECM. Happy with the outcome for now. Related post.

Get it repaired and either keep it or sell it. Plenty of people who can repair them.