Citroen C2 rough idle with new O2 sensor

[OneRandomBX]

Hello everyone
I swapped the O2 sensor (the one before the catalytic converter) on my 2005 C2 1.4 petrol the other day.
It started fine but after driving it I noticed an intermittent issue.

The engine runs rough on idle and the revs go down when I give it some throttle, sometimes it even dies.
I can rev it up by giving it gas very slowly and when it goes above around 2k RPM it runs completely fine.
Shutting it off and starting it again usually makes it run normally but it starts acting up again after a while.

It's drivable but of course hard to drive in low revs and partial throttle.
I've disconnected the sensor and now it runs completely fine.

Do you have some ideas about what the issue could be?
Thanks!

[PaulC5]

Possibly the sensor you have put on is an aftermarket part and not working correctly. I think O2 sensors do not work on the fueling until they have warmed up, hence it running ok when first started.

[ozvtr]

I swapped the O2 sensor (the one before the catalytic converter) on my 2005 C2 1.4 petrol the other day.

Why?

Does the engine light come on when you first turn the key to ignition, but don't start the car? Have you owned the car since new?

Probably the O2 sensor, but you should be getting a check engine light.

[OneRandomBX]

I changed it because I was getting a check engine light (would appear for a few days and then disappear for a few days)
And I had error codes for that O2 sensor (P0134 O2 Sensor Circuit No Activity Detected Bank 1 Sensor 1)
I put in a relatively cheap sensor because I was told here that even the cheapest Chinese ones work fine.
I haven't gotten a check engine light yet after the swap, but that's maybe just because i haven't driven it enough.

[ozvtr]

Do you have an OBDII scanner?
Can you check the long and short term fuel trims? Should be below 5% (may peak over 5% but average below 5%)
Can you monitor the upstream O2 sensor (sensor 1 bank 1) voltage? Should oscillate between about 0.2V and 0.8V about every second.
Downstream O2 sensor (sensor 2 bank 1) should flat line at about 0.5V.
Fuel trims should tell you if upstream O2 sensor is out of calibration.

I have not had trouble with Chinese O2 sensors in my TU3JP engines. However you might just be unlucky.

[OneRandomBX]

No unfortunately I don't have a scanner.
I do have a picture of the readings from before i swapped the sensor though:

O2B1S1 (V) : 0.450 (this value was constant, I assumed it was the placeholder value when the sensor isn't responding)
SHRTFB1S1 (%) : 0.8
O2B1S1 (V) : 0.705 (this fluctuated quite a bit)
SHRTFTB1S2 (%) : 99.2

I just had a quick drive (with the sensor disconnected), when I came back I turned it off, plugged it in and started it up. Idled for a while and drove back and forth a bit and it ran normally.
Tried fiddling with the connector itself under the bonnet. It seemed to run rough for a moment a few times when moving the connector, could it be a bad connection? I don't see any corrosion on the contacts

Update: fired it up to run into town and it's acting up again. Disconnected again so I can drive safely

[ozvtr]

You would have to assume the problem is with the sensor, given the symptoms. But...
The engine ECU can tell if the sensor is removed and "fall back" into a predetermined map to run the engine. Not perfect, but acceptable.
If the engine ECU gets some form of signal from the sensor, it will try and make sense of it and do the best it can.
The problem I have is that both sensors (the old and the new) have intermittent problems. This tends to point to a problem somewhere else, like the wiring.
If you plan to do more engine work on your cars but don't want to spend too much on a scanner, buy a cheap OBDII scanner that will graph O2 sensors and fuel trims. You should be able to find one for less than 50 euro.
Even a scanner for 20 euro will do live data for those parameters. This will give you a much better picture of what's going on.

The fuel trims are by bank, not sensor. The TU3 engine only has one "bank" so it will only have one short term and one long term fuel trim. The fuel consumption is compared to a nominal fuel consumption then given a positive or negative deviation in percentage. Positive if the engine ECU is adding fuel. Negative if the ECU is taking fuel away. Most ECUs cannot add or subtract more than about 25%.

The 0.45V on sensor 1 is a bias voltage. If it reads that all the time the O2 sensor is not connected, or dead.

Sensor 2 should not fluctuate much. It reads the amount of oxygen in the gases downstream of the CAT. If the voltage is high there is less oxygen. If the voltage is low there is more oxygen. The CAT should combine oxygen with unburnt fuel, leaving less oxygen. So, typically the voltage is high. If the voltage goes up and down in sympathy with the upstream O2 sensor, the CAT is definitely not working.

[OneRandomBX]

I've been thinking about this for a bit and I have a few questions.

I'm assuming that the previous O2 sensor was faulty and that my car just ran open loop all the time, and ran fine.
Then when I put in a working O2 sensor it tried running closed loop but is running rough because there is a problem somewhere else.

Could the issues be caused by a vacuum leak or faulty MAP sensor?
Does my thinking even make sense?

I found an old thread here where similar symptoms on a C2 were described, and it turned out to be a vacuum leak.
Would my car even run right open loop if it had a vacuum leak?

Anyway, I'm ordering an OBD2 scanner and multimeter today, so I'll be able to collect more data once they come in, in about ten days.

[ozvtr]

Yes, vacuum leaks tend to affect idle. They don't tend to affect the overall performance of the engine...just idle. If the leak was big enough to affect upper range performance, the engine would not idle.
You can hear vacuum leaks. You can also spray brake cleaner around the inlet plenum and see if the idle picks up.
However vacuum leaks don't tend to "reset" when you turn the ignition off then start the engine again.
Doesn't explain the P0134 error code. Could you have two faults? Anything is possible.
Disconnecting the MAP senor also causes the engine ECU to fall back into a fail safe mode. Again engine will run but not optimally.

[OneRandomBX]

Hello, I got my scanner today and I've got a little update:

Connected the O2 sensor, cleared the codes (had a sensor 1 heater and sensor 1 no data code from when I was driving with it disconnected) and started the engine to collect some data.
It started running rough, I shut it off and started it again and it ran fine the whole time I was there, maybe 10 minutes. - Good data set
A few hours later I came down and started the engine again, it ran rough again so I looked at the live data while it was running rough. - Bad data set

I'll make a little table to compare some points.

	BAD	GOOD
LOAD_PCT (%)	42	25	???
ECT (F)	100	86
SHRTFT1 (%)	-25	0.8
LONGFT1 (%)	27.32	29.7
MAP (kPa)	53	29	???
SPARKADV	-1	10
TP (%)	17.6	14	throttle pedal wasn't pressed at any time
O2B1S1 (V)	0.445	0.690	new O2 sensor is working, voltage actually changes
O2B1S2 (V)	0.385	0.450

Of course when it was running well it was in open loop, and in closed loop when running rough.

Can any meaningful conclusions be drawn from this data? At least I took my time to make it look nice...

And another thing happened yesterday while driving in traffic, engine hot, several times while stopped and idling at red lights, I got an "OIL PRESSURE WARNING", just flashing once and disappearing.
Oil level is still good, no weird sounds from the engine and no smoke.
The oil pressure sensor has been dripping with oil for as long as I've had the car, so I'm hoping that it's a faulty sensor :')

[ozvtr]

The long term fuel trim is maxed out at 29%. That's not good.
So The engine ECU is increasing the injector pulse width to get a stoichiometric ratio indication from the O2 sensor. But why?
Could be an air leak but could also be a fuel delivery problem. Can you check the fuel pressure? Have you inspected the injectors? Just because the ECU says it's increased fuel delivery doesn't mean it's actually delivering more fuel! Low pressure, a bad injector or a blockage will cause the ECU to try to deliver more fuel.

The frustrating thing is the fact that the engine runs fine in open loop. That tends to point to the O2 sensor.

The O2 sensor should oscillate between 0.2 and 0.8 Volts (about).
Wait, how is the short term fuel trim -25% but the long term is +27%? That's a big swing.

At what point did the short term fuel trim go to 0.8%? What were you doing?
What are the average of the short and long term. How much do they oscillate by? Look at the average with the accelerator peddle in a fixed position and the car maintaining a constant speed. Then look at the idle results.
The long term should remain relatively stable, but the short term will change dramatically depending on what the car is doing.
If the results show good performance at speed but poor performance at idle, I'd suspect a vacuum leak. But those long term fuel trims are not indicating that. I would guess the performance is poor right across the range?


Catalytic converter is still in good shape. Voltage hovering around 0.45 is good.

MAP values will be low at idle and get higher at faster revs. You can check the MAP sensor by going to the weather bureau and get the barometric pressure of the day. Check the value without the engine running.

Throttle position never goes back to 0%. 10% is about normal.

Engine temp should oscillate between 85 and 95. Thermo fans typically come on just over 100 degrees. So in stop start traffic the temps will be higher.

Spark advance will change depending on speed and load on the engine. The faster the engine the more advance BUT the greater the load the less advance.

[OneRandomBX]

All the data from my previous comment was taken when stopped and at idle.

I first started the car, it went to closed loop and started running rough, I shut the engine off and fired it up again and then it stayed in open loop.
The short term fuel trim was 0.8 when the car was "stuck" in open loop.
Later in the day I fired it up again and it went to closed loop, that's when the short term trim was at -25%

Fired it up again today and it wouldn't even go out of open loop.
The O2 sensor 1 was steady at around 0.450 at idle, it went up when I added throttle. Should it fluctuate when it's in open loop or does that happen just in closed loop?
The O2 sensor 2 was steady at around 0.700.

I shut it off and tried starting it a few times and it wouldn't fire up, that doesn't usually happen...
It fired up when I gave it a bit of throttle and then ran fine as before.

So far all I really know is that the car runs fine in open loop and terribly in closed loop.
I'm just getting more and more confused...

[PaulC5]

Maybe it is just the replacement O2 sensor is not working well enough. What was it like with the old one ?

With the engine off the MAP sensor should show atmospheric pressure which is near enough to 100 kPa. With the engine on at idle it will be lower and a vacuum but if accelerating it can go above 100 kPa if you have a turbo.

[OneRandomBX]

The old one didn't give any data and the engine just ran in open loop all the time, ran with no issues.

With the engine off MAP reads 100.
Measured when it was running well, cold idle at around 1050rpm the value was 30. (open loop)
When it was running rough the value was 53. Also at idle with cold engine, but was running at around 800 with a misfire. (closed loop)

[ozvtr]

I'm not convinced the O2 sensor is swinging as much as it should.

Does it swing between approximately 0.2 and 0.8 volts about once per second?