Poor idle and stalls + onboard computer fuel use

This is the Forum for all your Citroen Technical Questions, Problems or Advice.

Moderator: RichardW

Post Reply
pcspinheiro
Posts: 31
Joined: 23 Nov 2012, 09:42
Location: Denmark
My Cars:

Poor idle and stalls + onboard computer fuel use

Post by pcspinheiro »

Hi everybody,

I'm the recent owner of a 2002 Xsara estate exclusive 2.0i 16V (136 bhp) that was cheap but did not come without its hiccups... I've posted these problems in a french citroen forum but I got very little feedback over there.
The car starts fine and runs OK (although I would expect a little more power for 136 bhp...) but once warmed up the engine idles unstably and bellow 700 rpm and easily stalls if one is not careful with the clutch. Also, when shifting, the rpms can sink to bellow 500 although so far it has not stalled because of this... I also became aware of the fact that the on-board computer displays odd fuel consumptions... when coasting in gear it drops ever so slowly but I would expect it to drop quite fast and to zero or close... I can coast 100m in fifth gear and not see the fuel use drop bellow 4l/100.

Could the 2 problems be related and what could it be? I was leaning towards the throttle position sensor but I guess that would not affect the idle speed... so would really like some input on that. Maybe multiple sensors are malfunctioning? I do know the ECU was changed under the previous ownership "it burned", I was told. How to test?

Thanks in advance and best wishes!

Paulo.
User avatar
Mandrake
Posts: 8615
Joined: 10 Apr 2005, 17:23
Location: North Lanarkshire, UK
My Cars:
x 664

Re: Poor idle and stalls + onboard computer fuel use

Post by Mandrake »

Can you put it on a Lexia to check fault codes and live sensor data ? Check the Lexia thread for Lexia owning forum members nearby.

Candidates for slow unstable idle when hot and a tendency to stall include the MAP sensor and lambda sensor. The lambda sensor would only cause issues when warmed up as the sensor is ignored when the engine is cold. A quick and easy test for the lambda sensor is to disconnect it and see if the idle is ok when warmed up, as the engine ECU is able to run in open loop mode without the sensor connected.

The MAP sensor being faulty can cause a slow unsteady idle and a tendency to stall especially when you blip the throttle quickly, try leaving the MAP sensor disconnected to see what I mean - the engine will run, but not very well! It will struggle to idle and have massive flat spots in the performance when depressing the throttle...

The throttle position sensor is unlikely to cause idle problems as the butterfly will be completely closed with all the air going through the idle control valve which is operated by the idle stepper motor.
Simon

1997 Xantia S1 3.0 V6 Auto Exclusive in Silex Grey
2016 Nissan Leaf Tekna 30kWh in White

2011 Peugeot Ion Full Electric in Silver
1977 G Special 1129cc LHD
1978 CX 2400
1997 Xantia S1 2.0i Auto VSX
1998 Xantia S2 3.0 V6 Auto Exclusive
Homer
Posts: 1503
Joined: 26 Feb 2003, 10:52
Location: Yorkshire
My Cars: Current:
Volvo V60 D4 180

Previous:
BX16RS (two of),
BX19TZI,
Xantia 2.0i saloon,
Xantia 2.0 Exclusive CT turbo Break,
Peugeot 807 2.0 HDi 110,
Renault Grand Scenic, 2.0 diesel (150bhp)
C5 X7 2.0 HDi 160 which put me off French cars possibly forever
x 16

Re: Poor idle and stalls + onboard computer fuel use

Post by Homer »

pcspinheiro wrote:I can coast 100m in fifth gear and not see the fuel use drop bellow 4l/100.

There are some software limits on what it will display. I don't know what that is in English money but IIRC the computer will never show worse than 9mpg or better than 99.9mpg
User avatar
Mandrake
Posts: 8615
Joined: 10 Apr 2005, 17:23
Location: North Lanarkshire, UK
My Cars:
x 664

Re: Poor idle and stalls + onboard computer fuel use

Post by Mandrake »

If that's 4 litres per 100km, it works out to 70.6 MPG.
Simon

1997 Xantia S1 3.0 V6 Auto Exclusive in Silex Grey
2016 Nissan Leaf Tekna 30kWh in White

2011 Peugeot Ion Full Electric in Silver
1977 G Special 1129cc LHD
1978 CX 2400
1997 Xantia S1 2.0i Auto VSX
1998 Xantia S2 3.0 V6 Auto Exclusive
pcspinheiro
Posts: 31
Joined: 23 Nov 2012, 09:42
Location: Denmark
My Cars:

Re: Poor idle and stalls + onboard computer fuel use

Post by pcspinheiro »

Thank you for your replies!

Good point about the lambda probe and MAP sensor! I have just received my lexia kit (being in Denmark I can't really reach out to you fellas in the UK) and I will hook it up to the car as soon as I get a XP boot on my Win7 laptop and everything installed. From my extensive reading on the net it could be a number of other things that also included the coolant temp sensor, dirty injectors, etc, etc.

However, regarding the fuel use I think you guys are missing my point (or I'm missing yours...): when one lets go of the gas pedal completely and lets the car coast IN GEAR the fuel use should be zero or close quite quickly, at least that is what I have seen in other cars with on-board computer, since the ECU will reduce the injector open time to essentially zero (this would not be the case if coasting in neutral as some fuel is required for idling). In my case, however, I see the fuel consumption decrease VERY slowly. I suppose the ECU uses, among others, the throttle position sensor to calculate fuel consumption and a faulty reading could cause this display? Maybe the MAF sensor reading could be the culprit too?

Does the Lexia software allow to compare actual live readings with expected values to determine a faulty sensor?

Thanks again!

Paulo.
User avatar
Mandrake
Posts: 8615
Joined: 10 Apr 2005, 17:23
Location: North Lanarkshire, UK
My Cars:
x 664

Re: Poor idle and stalls + onboard computer fuel use

Post by Mandrake »

Most modern petrol engines do indeed have a fuel cutoff that activates on the overrun that shuts off the injectors completely when you're coasting down hill. On my V6 the Lexia will show on the live sensor data whether the overrun fuel cutout is active or not, (indicated by a 1 or 0 flag) it may also do so on your engine ECU.

The ECU will only allow the cutoff on overrun to activate if it thinks the butteryfly is completely closed in the idle position, so if the butterfly sensor is faulty it may not activate, likewise if the accelerator cable is adjusted incorrectly.

Butterfly opening percentage is one of the things you can read in real time on the Lexia so you will quickly see if the butterfly sensor is working properly and the butterfly is closing fully on idle.

Personally I wouldn't worry too much about the trip computer reading, I would focus on getting the idle and performance fixed first, then check the butterfly sensor reading is correct and that the overrun cutout is activating with a test drive, if that's all ok I wouldn't fret about the trip computer figures, as they're only an approximation...
Simon

1997 Xantia S1 3.0 V6 Auto Exclusive in Silex Grey
2016 Nissan Leaf Tekna 30kWh in White

2011 Peugeot Ion Full Electric in Silver
1977 G Special 1129cc LHD
1978 CX 2400
1997 Xantia S1 2.0i Auto VSX
1998 Xantia S2 3.0 V6 Auto Exclusive
pcspinheiro
Posts: 31
Joined: 23 Nov 2012, 09:42
Location: Denmark
My Cars:

Re: Poor idle and stalls + onboard computer fuel use

Post by pcspinheiro »

OK, so I hooked up the Lexia to the car this morning and got a quick look at some data and faults (unfortunately my 2.5 yrs old daughter saw me leaving and insisted on joining me so I had to cut it short...). there were stored faults for oil temp sensor, coils 1 and 2 misfire, and fan units. I'm posting some pics bellow. As you can see the RPM sit in a low 672 with some surges to over 700. The oxygen sensors seem to be cycling OK although at idle and the MAP sensor detects pressure changes on acceleration. The throttle position sensor was fine too, immediately showing "idle" as soon as the foot was off the gas. Strangely enough the coolant temperature sensor kept displaying 40-45C throughout the measurements (the stored faults are, I assume, old and have 95-98C). However, the dash showed me 90-98 so I disconnect the sensor and immediately the radiator fans came on, the dash gauge dropped to min and a blinking STOP light appeared. All went back to normal once the probe reconnected but no change in RPM throughout. What is curious is that I often get the "check engine light" when driving in traffic at the moment the fans kick in if I'm accelerating, at idle this doesn't seem to happen. Could all this be caused by a coil pack on its way out? I guess there is no cheap way to test...

I did not get to drive it to check the injector cutoff but I will as soon as possible. Any other things to look at? Ideas by now?
Attachments
RPM.jpg
fault2.jpg
fault1.jpg
Last edited by pcspinheiro on 25 Nov 2012, 21:09, edited 1 time in total.
pcspinheiro
Posts: 31
Joined: 23 Nov 2012, 09:42
Location: Denmark
My Cars:

Re: Poor idle and stalls + onboard computer fuel use

Post by pcspinheiro »

missing pic
Attachments
upstream O2.jpg
User avatar
Mandrake
Posts: 8615
Joined: 10 Apr 2005, 17:23
Location: North Lanarkshire, UK
My Cars:
x 664

Re: Poor idle and stalls + onboard computer fuel use

Post by Mandrake »

pcspinheiro wrote:OK, so I hooked up the Lexia to the car this morning and got a quick look at some data and faults (unfortunately my 2.5 yrs old daughter saw me leaving and insisted on joining me so I had to cut it short...). there were stored faults for oil temp sensor, coils 1 and 2 misfire, and fan units. I'm posting some pics bellow.
Oil temperature sensor ? Are you sure you don't mean coolant temperature sensor ? At least on my V6 oil temperature is not monitored, and unless its a turbo I can't see the utility in monitoring the oil temperature. Edit: just looked at your fault codes again and see it is indeed oil temperature sensor. Can you monitor the live reading for the oil temperature and see if its approximately the same as the water temperature ?

Try clearing all the fault codes now that you've recorded them and see if any of them return after a test drive. Some of the codes could be years old faults that have been fixed but the codes were never cleared. (A major problem with fault codes is you have no idea how old they are and whether they are relevant unless they reoccur after being cleared...)
As you can see the RPM sit in a low 672 with some surges to over 700. The oxygen sensors seem to be cycling OK although at idle and the MAP sensor detects pressure changes on acceleration. The throttle position sensor was fine too, immediately showing "idle" as soon as the foot was off the gas. Strangely enough the coolant temperature sensor kept displaying 40-45C throughout the measurements (the stored faults are, I assume, old and have 95-98C). However, the dash showed me 90-98 so I disconnect the sensor and immediately the radiator fans came on, the dash gauge dropped to min and a blinking STOP light appeared.
If your coolant temperature sensor is reading a constant 40-45 degrees regardless of the true coolant temperature you've found a major part of your problem. This is a critical piece of information the ECU needs to know to tell whether the engine is warmed up or not. When cold the ECU richens the mixture and also raises the idle rpm. It also ignores the oxygen sensor data (on purpose) when the engine is cold, and only uses the oxygen sensor data once warmed up.

If its using the oxygen sensor data on a cold engine (because it thinks its warmed up when its not) the mixture will be incorrect because the oxygen sensor won't give accurate data when its cold. (when cold it should be running rich anyway)

Its also important to realise that the coolant temperature sensor that the ECU reads to control the operation of the engine is NOT the same temperature sensor that is used to drive the dashboard temperature reading, and is not in the same part of the cooling system. That's why your dashboard sensor is reading ok but the ECU is not getting correct temperature data. It's the sensor feeding the dashboard temperature gauge that you disconnected, not the ECU one.
All went back to normal once the probe reconnected but no change in RPM throughout. What is curious is that I often get the "check engine light" when driving in traffic at the moment the fans kick in if I'm accelerating, at idle this doesn't seem to happen. Could all this be caused by a coil pack on its way out? I guess there is no cheap way to test...

I did not get to drive it to check the injector cutoff but I will as soon as possible. Any other things to look at? Ideas by now?
I wouldn't worry about coil packs just yet, get the ECU coolant temperature sensor replaced first, reset the fault codes and see what fault codes return and what other issues remain after that sensor has been replaced. My guess is that will be the major part of your problem.

By the way how did you get the graphs of oxygen sensor data, rpm and so on ? I've used a Lexia for a couple of months and hadn't discovered the ability to graph sensor data! :-D
Simon

1997 Xantia S1 3.0 V6 Auto Exclusive in Silex Grey
2016 Nissan Leaf Tekna 30kWh in White

2011 Peugeot Ion Full Electric in Silver
1977 G Special 1129cc LHD
1978 CX 2400
1997 Xantia S1 2.0i Auto VSX
1998 Xantia S2 3.0 V6 Auto Exclusive
pcspinheiro
Posts: 31
Joined: 23 Nov 2012, 09:42
Location: Denmark
My Cars:

Re: Poor idle and stalls + onboard computer fuel use

Post by pcspinheiro »

Thank you so much for your reply!

I did not know there were independent coolant temp sensors for ECU and dash, where is the ECU one located then? The one I disconnected was close to the thermostat housing on the top right side of the engine. However, in my case, wouldn't the ECU think the engine is always cold and keep the revs up (it does for the first few minutes after starting from cold)?

I honestly don't remember by heart how I got to graphing the data but it was when you go to live sensor data then there is a bouton that looks like a graph with a sine wave, if you click that you can choose up to 6 parameters to record. Press record or start, don't remember, then stop and it will display. Don't try to use the save to disk button, I did and it just made lexia crash and forced me to manually shut down the computer. I did a print screen then pasted elsewhere.

I will look at everything more carefully (without my daughter nagging in the passenger seat and then see if the faults have returned after driving and check the coolant sensor.

Thanks again!
User avatar
IrfanIbrahim
Posts: 85
Joined: 29 Jul 2012, 17:59
Location: Malaysia
My Cars:

Re: Poor idle and stalls + onboard computer fuel use

Post by IrfanIbrahim »

I have quiet the same problem with my xantia but the difference is my RPM is 1.2k which is higher than normal idle RPM..
If your coolant temperature sensor is reading a constant 40-45 degrees regardless of the true coolant temperature you've found a major part of your problem. This is a critical piece of information the ECU needs to know to tell whether the engine is warmed up or not. When cold the ECU richens the mixture and also raises the idle rpm. It also ignores the oxygen sensor data (on purpose) when the engine is cold, and only uses the oxygen sensor data once warmed up.

If its using the oxygen sensor data on a cold engine (because it thinks its warmed up when its not) the mixture will be incorrect because the oxygen sensor won't give accurate data when its cold. (when cold it should be running rich anyway)

Its also important to realise that the coolant temperature sensor that the ECU reads to control the operation of the engine is NOT the same temperature sensor that is used to drive the dashboard temperature reading, and is not in the same part of the cooling system. That's why your dashboard sensor is reading ok but the ECU is not getting correct temperature data. It's the sensor feeding the dashboard temperature gauge that you disconnected, not the ECU one.
Now this part interested me :)
I've told to change my coolant temperature sensor as the old one could be faulty, it appears the original part are out of stock so i was told by the merchant to use a peugeot's sensor. The fit was the same but the length of the sensor are shorter than original. Maybe it could be the wrong coolant temp sensor on my Xantia that cause this idle problem and engine stalls... :roll:
1994 Xantia 2.0I

Keep calm & drive a xantia!
pcspinheiro
Posts: 31
Joined: 23 Nov 2012, 09:42
Location: Denmark
My Cars:

Re: Poor idle and stalls + onboard computer fuel use

Post by pcspinheiro »

Does anyone have an exploded view of the thermostat housing region (I assume it should be around there...) so that I can locate the correct coolant temp sensor without going on a dismantling spree?

thanks!
User avatar
Mandrake
Posts: 8615
Joined: 10 Apr 2005, 17:23
Location: North Lanarkshire, UK
My Cars:
x 664

Re: Poor idle and stalls + onboard computer fuel use

Post by Mandrake »

What is the code for your engine ? I have the data here but don't know exactly which engine you have. It might be one of the following:

XU7JP, XU7JB, XU7JP4, TU5 JP, TU3JP, XU10J4RS.

If you can let me know which one you have I can post a diagram. It's usually listed on the plaque near the oil filter or front of the cylinder block. Alternatively if you're not sure, PM (private message) me your VIN number and RPO number and I can look up exactly which engine you have.

PS thanks for the tip on graphing within the Lexia software, I'll have a look at that next time I have it connected, that will be excellent for monitoring data on a test drive when I can't look at the screen while driving... :)
Simon

1997 Xantia S1 3.0 V6 Auto Exclusive in Silex Grey
2016 Nissan Leaf Tekna 30kWh in White

2011 Peugeot Ion Full Electric in Silver
1977 G Special 1129cc LHD
1978 CX 2400
1997 Xantia S1 2.0i Auto VSX
1998 Xantia S2 3.0 V6 Auto Exclusive
pcspinheiro
Posts: 31
Joined: 23 Nov 2012, 09:42
Location: Denmark
My Cars:

Re: Poor idle and stalls + onboard computer fuel use

Post by pcspinheiro »

No problem, I found it by just tinkering around a bit. I'm a research scientist and could not go by without checking what that button would do :mrgreen: .Luckily it was not the ECU or engine self-destruct button :-D

I'm sending you a PM with that info as soon as the kiddos are asleep, have to go out and check the stickers as I don't know what engine it has. By the way, I see that people have access to a citroen site to obtain those diagrams. I was just looking at it and it seems that you have to register as a repair shop. Can you just fill in bogus info (by which I mean to instantly create your own repair shop) and gain access? Or will I be banned too quickly to be worth the trouble?

thanks again!

Paulo.
User avatar
Mandrake
Posts: 8615
Joined: 10 Apr 2005, 17:23
Location: North Lanarkshire, UK
My Cars:
x 664

Re: Poor idle and stalls + onboard computer fuel use

Post by Mandrake »

Ok, based on your VIN number yours is a Xsara series II with an EW10J4 engine, I was originally looking up the Series I Xsara's.

It looks like your engine does in fact only have a single coolant temperature sensor unlike earlier Xsara engines which have 2 or 3 sensors. (One model has three sensors next to each other - one for gauge and warning light, a second one to operate the cooling fans, and a third for the ECU! :lol: )

I've attached three zipped PDF files with all the relevant information I can find. According to these files, the coolant sensor signal is sent to the ECU which measures the temperature and uses this information to control the injection, timing etc. The temperature reading is also then sent over the multiplexed network to the BSI where it controls the cooling fan speeds, instead of having a separate temperature sensor.

If disconnecting the sensor caused the fans to come on, this path of communication from the sensor, through the ECU to the BSI must be working. It's unclear how the temperature gauge gets its reading, without studying all the files in detail I assume it gets its information from the ECU or BSI.

Therefore I have no explanation why the Lexia is showing a constant 40-45 degrees for coolant temperature, yet the gauge on the dashboard changes, since there is only one sensor that goes to the ECU first, and the ECU then sends that data digitally to the BSI... :?
Attachments
Data - Cooling System.zip
(40.75 KiB) Downloaded 99 times
Principles of operation - cooling system.zip
(94.27 KiB) Downloaded 102 times
Generality - Cooling System.zip
(50.54 KiB) Downloaded 87 times
Simon

1997 Xantia S1 3.0 V6 Auto Exclusive in Silex Grey
2016 Nissan Leaf Tekna 30kWh in White

2011 Peugeot Ion Full Electric in Silver
1977 G Special 1129cc LHD
1978 CX 2400
1997 Xantia S1 2.0i Auto VSX
1998 Xantia S2 3.0 V6 Auto Exclusive
Post Reply