RENAULT CLIO 2 - Erratic performance and stalling - Open Loop Mode investigations

Open Loop Mode during idling

I had opportunity to make some more observations with the help of the Torque app. In brief the observations were

• When idling or running at low speed or when the accelerator was lifted, the ECU always switched to Open Loop mode.
• There was one instance when Torque reported running in Open Loop mode because of low temperature.
• There were several instances when system failure was reported at the time when erratic performance was being experienced.
• Erratic performance was always at low speeds/engine RPM

So what I would like to know now is exactly when the engine should be running in Open Loop mode and for that I consulted Lambda Power and this article.

Open Loop Mode not expected at Idling

Summarising it states, for a narrow band sensor, that

• Open Loop will be used during accelerating and decelerating/engine braking
• Closed Loop is used during idling and part load conditions

That makes sense because taking O2 readings is a process that incurs a certain amount of time lag. Therefore, Closed Loop would not be suited to rapid changes in engine loads but is suited for idling. Unfortunately, at idling the opposite Open Loop is being observed!

Support for an O2 (Lambda) sensor fault

A little bit more research on the Lambda Power site reveals these comments. The O2 sensor can be a cause of poor idling and engine racing that is often attributed to the idle control valve.

My next questions are "How does the ECU decide when to swap modes?", "How can I verify the operation of the O2 sensor?"

How does the ECU decide when to swap modes?

A quick search for answers to the first question was unsuccessful but an obvious answer would be the accelerator/throttle position. More information is needed before pursuing this line of investigation of fault cause.

How can I verify the operation of the O2 sensor?

For this I have once more consulted the helpful Lambda Power site and these articles.

Some work on this has already been carried out and described briefly in the previous post. Namely, observing the O2 values whilst artificially introducing excess air to weaken the mixture and then doing the opposite by adding propane. These tests seemed to indicate the sensor was functioning, but well enough?

Why is the fault intermittent?

The final question is "Why is the fault intermittent?" That doesn't suggest a worn out or faulty device. Perhaps an inspection of the wiring to the sensor would be a valuable first step! My first rule is always "Look for the simple things first".

RENAULT CLIO 2 - Erratic performance and stalling - Coil Pack Replacement

It's not the coil pack!

After reflecting on the discussions in the previous post I decided to change the coil pack. Unfortunately, it has made no difference.

RENAULT CLIO 2 - Erratic idling and stalling

Renault Clio electrics

It's been a busy time car wise. This time the problems have been with my son's 2001 Clio 2. A good little car but it's Achilles's heel is, once again, electrics.

Intermittent idling problem

The fault and related observations are as follows

• Engine starts OK but hunts excessively and struggles to idle.
• Then will be unresponsive to the accelerator.
• Will stall easily when the engine does eventually respond to the accelerator.
• Runs fine for days and then will go faulty for a relatively short period.
• Fault can develop at anytime. Not just on restarting.
• Doesn't seem to be related to heat/cold/damp.
• Clicking heard from RHS front wing area. (The petrol vapour capture vessel for emission control).
• Coil/electrics indicator light on. 

Must be possible to deduce the cause

Having consulted the internet this problem does seem to be relatively common. However, there is, as usual, no consensus of opinion and, as usual, lots of people have spent lots of money and time replacing all manner of things with no clear conclusion or feedback. So my thinking is "it must be possible to isolate the problem by thorough investigation". Especially since speculative parts replacement is so potentially expensive.

The first thing to say is I've not fixed the problem yet. Unfortunately, every time I have access to the car it runs faultlessly, except for once. Fortunately, that one time, I gained some information using my OBD2 reader and Android Torque app.

This is what I've learnt

OBD data when engine was running badly

• ECU running in Open Loop even though the car was at normal running temperature.
• No Short Term Fuel Trim data being reported.
• O2 sensor above 0.9V
• Reported throttle position anywhere between 5 - 13%
• Idling RPM about 1700 rpm but very erratic.
• Manifold pressure appears to respond to changes in reported throttle position.

Then, whilst collecting the data, the engine started to run normally and this is what I observed.

OBD data when engine was running normally

• Short Term Fuel Trim data available and converges in the range  -0.6% to 13%
• O2 sensor voltage falls and oscillates between 0.1V (lean) and 0.8V(rich) approx. 
• Throttle position settles at 7%

What conclusions can be drawn from these observations?

Why Open Loop Mode?

First of all, should the ECU have been in Open Loop? As I understand it this is a mode of operation when the engine is cold and the O2 sensor is not warm enough to measure O2 accurately. Hence the ECU uses a pre-programmed set of parameters for air flow and fuel flow (as measured by the fuel trim data). Is this a "red herring"?

Reasons to run badly - Sensor Data

Regardless of Open Loop mode why would it run badly? Open Loop mode I understand simply requires air flow which can be calculated from the engine RPM and then adjusted using the manifold pressure. The ECU then knows how much fuel is required. So that would suggest checking the

• Crankshaft sensor (Replaced relatively recently) (£32)
• MAP sensor (£112)

Reasons to run badly - Fuel and Ignition

So if the sensors are OK that would mean the fuel and ignition components are faulty. That is the

• Throttle Body (£250)
• Fuel Injection system
• Ignition pack (Replaced relatively recently) (£27)

Crankshaft Pulse Sensor

The crankshaft pulse sensor was replaced relatively recently and Torque appears to show RPM data without suggestion of error. When the previous sensor failed it caused starting problems and this is not currently a problem. Judged to be a low probability of being the cause. 

MAP Sensor

Torque again shows data being reported and with no suggestion of wild or missing data. Also, as a simple sanity check, when the engine is not turning the reported manifold pressure is consistent with atmospheric pressure.

Throttle Body

Some of the internet posts I've read report similar symptoms and suggest the throttle control valve could be faulty and they recommend removal of the intake manifold (not a 2 minute job!) to either clean the valve or replace it (an expensive part to speculate on!).

Through Torque I can observe changes in manifold pressure following ECU requested changes in throttle position. Therefore, if it was the throttle control valve causing the problems, would I observe this? I don't think so.

Fuel Injection System

I've got no way of testing the fuel injection system but Torque does show a reported O2 figure from the exhaust sensor that indicates a very rich mixture when the engine is running badly. From this I infer that there is sufficient fuel and so I'm not inclined to suspect this area as a cause.

An interesting monitoring period was when the engine settled down to normal idling. It came out of Open Loop operation and into Closed Loop. This was evidenced by the reporting of the STFT (Short Term Fuel Trim) data and the O2 figures responding accordingly.

Ignition Pack

And finally to the ignition pack. Again this has been replaced relatively recently and it would be a surprise to find it the cause. The dash ignition fault tell tale does at times light but unfortunately, when I had Torque connected there weren't any error codes available to give further inormation. That said, previously when a code was available, it was of little help.

Some background on Fuel Trim and O2 control.

In Closed Loop operation the ECU attempts to keep the air:fuel ratio at its optimum of 14.7:1 (In practice, probably a little richer). It adjusts fuel flow by adjusting the time the fuel injector is open and this is called the Short Term Fuel Trim. (It's called short term because it's a correction factor to the Long Term Fuel Trim which is adjusted over a much longer time period).

The ECU measures the effect of the fuel trim adjustment by means of the O2 (sometimes called lambda) sensor. A high voltage, say 0.9V means the mixture is rich i.e. too much fuel and so the trim is reduced. A  low voltage means too weak i.e. too little fuel and so the trim value is increased. The optimum voltage is around 0.45V for cars such as the Clio ( Note: There are narrow band and wideband sensors. The latter are most likely to be found in more modern high economy cars but the control principal is the same).

Next Steps

It's difficult to make progress identifying the fault when the car is running OK! One strategy could be to try to create the fault by 'playing' with the various sensor components.

(I did also wonder if the O2 sensor was OK since this is a common source of poor engine performance. To test it I followed a procedure of letting excess air into the intake manifold by removing a pipe followed by injecting propane into the air intake (from a plumbing torch). These steps have the effect of making the air:fuel mixture weak and rich respectively. Using Torque the O2 readings can be observed changing accordingly. To all intent and purposes the reading should respond immediately, which they did).

BMW 320D - Air Bag Occupancy Seat Pad Sensor Replacement

Air bag warning light intermittent operation

Some time ago the air bag warning light began to operate intermittently. After some research this was found to be a common fault on the 3 series and caused by the seat sensor failing. This is what the sensor looks like. (The top of the picture is the lifted seat fabric. The view is from the back of the seat looking down).

Emulating the sensor pad

Confirmation that it was a sensor fault was obtained by using a simple device documented on other sites and forums. This device mimics the sensor by means of a zener diode and resistor. In fact the mat itself contains a diode but the resistor in the mat is in the form of a resistive grid that changes resistance when stretched i.e. when someone sits in the seat. The diode provides a means of creating a reference voltage from which  to measure the resistance changes. (Well that's how I think it works).

The parts to make the seat sensor emulator cost pence, can be fitted in minutes and can be obtained from Maplin. (I haven't got them the details to hand but will try and post them later).

Although the emulator could be left in place this is probably a dangerous solution long term. The function of the sensor is to allow the air bag system to determine which air bags to fire in the case of an accident. Therefore, the emulator will probably make the system think the passenger seat is unoccupied and therefore not fire the relevant air bags.

Sensor pad replacement

I couldn't find any documentation on the web on replacement of the sensor and so I had to proceed by trial and error. In fact it took me longer than a day to do what in retrospect was a couple of hours work so I thought this post was needed.

Key Facts

• The seat is a large complicated assembly but the seat cushion containing the sensor is easily removable. There are 2 screws at the front and a plate at the back to remove.
• Removal of the cushion is possible with the seat in-situ but replacing it would be difficult/impossible. So it will be necessary to take the seat out.
• Removal of the complete seat is easy.
• The electrical loom connector for the seat is a little puzzling to disconnect. It doesn't just pull apart. Instead look for a slider mechanism. This moves perpendicular to the body of the connector and separates it into two halves.
• The fabric on the seat cushion can be prised off the rim of the metal seat pad shell. (It's not glued etc). This will nearly reveal the sensor but first it is necessary to cut a number of attachments rings. These rings hold stiffening wires in the fabric to the seat cushion. These can be replaced on reassembly with wire ties.

Tools

• Torx wrench bits for screws
• 16mm hex socket
• side cutters/pliers
• quantity of small cable ties (20-ish)

BMW 320D - General update and review

Over the past months and years I've posted quite frequently on my 150000-ish mileage 2005 BMW 320D Touring. Some of the faults I've fixed and some are still waiting attention so I thought a quick summary report would be useful.

• First of all it's a great car, the family workhorse and feels fit for another 150000 miles.
• A couple of years ago it suffered from a strange intermittent, hesitation/running problem that was eventually identified as the swirl flaps not operating properly.
• The problem was the electrical operated vacuum actuator valve for the swirl flaps was failing. Disconnecting/blocking the vacuum pipe to the valve in order to disable the swirl flaps solved the problem! (It's still disconnected because replacing the valve requires the intake manifold to be removed. ).
• The exhaust does smoke at times under hard acceleration but nothing to cause concern.
• The disconnected swirl vanes default to the open position and so will, I anticipate, affect the fuel/air mixture and probably result in smoke creation.
• Kits are available to remove the swirl vanes as these sometimes fall into the engine causing destruction! I think these block the swirl vane intake path but the car performs perfectly well in the open position.
• The turbo developed a whine at low speed some time ago. It hasn't got worse and performance seems unaffected after well over a year. I've no idea why this should be.
• In winter it was noticed that the engine was not warming/achieving normal running temperature very quickly. The radiator air flow flaps were found not to be working.
• The radiator air flow flaps are actuated in the same way as the swirl vanes and it is strongly suspected that the failure is for the same reason. i.e. the electrical actuator valve. Again this valve is under the intake manifold and difficult/impossible to replace without removing the intake manifold. 
• The crankcase oil breather was changed after reading that, if blocked, this was potentially harmful to the turbo oil seals and could cause smoking (and in the worst case engine destruction if the seals failed!!).
• The air bag warning light started to light intermittently. The warning light was being caused by the passenger seat occupancy sensor.
• A work around of a resistor and zener diode was used as a solution for the MOT!!
• The occupancy sensor has now been replaced.  

BMW320D - Smoke on hard acceleration significantly reduced.

Back in August I upgraded the crankcase filter to a vortex type on my 320D to try and reduce exhaust smoke. In the post I said I would report back after testing. Well, several thousand miles later, I can report it was an important job done. By no means difficult and it has significantly reduced the smoking.

BMW 320D - Excessive smoke on hard acceleration. The cause.

Recent posts have documented my concern regarding heavy smoke from the exhaust of my 2005 BMW320D under hard acceleration. After research it seemed it could have been resulting from a blocked crankcase filter but I was waiting for a long journey  before pronouncing on the results of replacing it. Well, I'm pleased to say it worked! It still smokes at times but only under very heavy acceleration. I was fortunate to be followed by a friend on my trip who confirmed the result. However, the turbo still whines!