Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[demolite]

In an MOT emission test, the meters display CO, HC, and Lambda. Limits are CO =<0.2%, HC =<200ppm and lambda between 0.97 and 1.03. What does the lambda measurement mean?

It's a calculated residual oxygen factor. It tries to determine if engine is running on average a stoichometric mixture. Because of the catalytic converter, this value is only mathematical. Air leaks and such might skew this value.

[demolite]

I finally fixed the ECU grounds. The three small bolts holding the three ground wires to the dispack tray were dirty and only finger tight! Good news is, lambdas work nicely now and voltage goes to 800+ mV when accellerating.

Bad news is I broke something. I removed the scuttle and the two small vacuum pipes from the brake booster line for more access. Cleaned the grounds and then assembled all back.

When warm, the engine runs fine but cold starts became very poor. This happened right after fixing the grounds, so I guess I must have done something by accident.

Symptoms are:

- Very rich fuel mixture on cold start (outdoor temp 20 celcius)for about a minute (can see it on my MID as up to 10L/h and very strong petrol smell around car (this used to be about 4L/h))
- Diagnostic tool shows air consumption as 40kg/h during this time -> can anybody verify?
- Fuel trim for banks 1 and 2 differ about 30% on idle, about 4% on part load. Bank 1 seems to run very near normal trim, bank 2 is the one that needs much more. (These can be the other way around as diagnostic tool is inconsistent) 

Was thinking of a serious vacuum leak but couldn't find one... Any help?

[Andy H]

I finally fixed the ECU grounds. The three small bolts holding the three ground wires to the dispack tray were dirty and only finger tight! Good news is, lambdas work nicely now and voltage goes to 800+ mV when accellerating.

Bad news is I broke something. I removed the scuttle and the two small vacuum pipes from the brake booster line for more access. Cleaned the grounds and then assembled all back.

When warm, the engine runs fine but cold starts became very poor. This happened right after fixing the grounds, so I guess I must have done something by accident.

Symptoms are:

- Very rich fuel mixture on cold start (outdoor temp 20 celcius)for about a minute (can see it on my MID as up to 10L/h and very strong petrol smell around car (this used to be about 4L/h))
- Diagnostic tool shows air consumption as 40kg/h during this time -> can anybody verify?
- Fuel trim for banks 1 and 2 differ about 30% on idle, about 4% on part load. Bank 1 seems to run very near normal trim, bank 2 is the one that needs much more. (These can be the other way around as diagnostic tool is inconsistent) 

Was thinking of a serious vacuum leak but couldn't find one... Any help?

Have you cleared the long term fuel trims?

[demolite]

Have you cleared the long term fuel trims?

Yes I've cleared all learned values after the repair a few times. Trims become like stated earlier over a few days. Start problem is even with all trims zeroed.

[Kevin Wood]

Coolant temperature sensor is round the back on the coolant bridge. I wonder if you dislodged the connector to it?

That would explain poor cold starting.

[demolite]

I've been getting to the bottom of this.

Idle BLM values skyrocketed after I reconnected the grounds. The other value hit 255 while the other was a little over 200.
Engine seems to use the learned idle BLM value for cold starts and 700-1200 rpm idle. When its 255 (+100%) theres a lot extra fuel when the vacuum was leaking less (at higher idle).


It seems that the bad grounds were masking a problem with the fuel vapor solenoid beign stuck open. Thats why no vacuum leak was found, it was sucking air from the canister. I'm testing it now with the fuel vapor hose disconnected from the breather and the port blocked. Looks good, BLM values are staying in the low 140's, and no fuel smell anymore. The automatic gearbox does actually engine brake a lot better now, cant say about fuel consumption yet.

[mandula]

Hi,

I've got fault code P0130(4) for awhile now, always not present and cleared it, and it keeps coming back.
I replaced 1st lambda, but still the same. ECU casing is grounded via extra cable to battery ground terminal.

I was looking livedata of 1st lambda, and noticed voltage won't rise even above 700 mV, just jumps between 50 - 650 mV.
I think that is too low (is it?).

I read that it could also be ECU fault, but now I spotted this topic. Will check groundings soon, but what do you think, could this just be ground fault?

Here is some livedata from semi-warm motor running idle:

[demolite]

Based on the data, I'd say there could be some some voltage drop on the B1S1 circuit, but the O2 seems to be switching normally albeit maybe a little on the slow side. You are right, the 650mV is too low a maximum for a fully working circuit. What happens when you accelerate hard? 650 mV again or more? Accellerating hard takes the system to richer open loop settings and this way you can test if the O2 can generate more voltage, and if it gets to the ECU.

I take it from the OBD2 code you're getting that you have a facelift model with the 2.2 engine? On these engines, the O2's are 4 wire (instead of older 3 wire), and dont use the exhaust and engine block as a ground return path. They physically can't suffer from the same type of ground voltage problems like the older 2.0, 2.5 and 3.0 engines if the wiring is still connected as the manufacturer intended (a 4 wire sensor is used and connected to original 4 wires to O2). Bad ground tends to shift the readings down, and I saw a value of 0 mV (as it went negative) often from my O2.

There could be problems on the O2 wires if someone has connected them wrongly (signal ground and heater ground switched or connected together) or grounded a O2 signal ground wire on the first O2. Also check for high resistance on the signal wire, connectors and signal ground wire.

Be sure to check the O2 heater circuit for problems as well, as a cool O2 will switch slowly and can generate lower voltages. If you need any help making any measurements, I'll be glad to help.


Just a thought..
There could also be problems with stray current through the ECU casing (due to the extra grounding), causing voltage drop inside the ECU. Ecu on the 2.2 is normally grounded (according to wiring diagram) at the  chassis point below the battery. Grounding the case to the battery terminal could theoretically make current run from

the chassis point -> the ecu ground wire -> the inside of the ecu -> the case -> the battery

causing voltage drops in its internal ground plane. Try disconnecting the extra ground.   

[mandula]

Thanks for your very informatic reply!

I have Z22XE engine with original wirings from ECU to lambda, my new lambda required to cut old lambda connector to be attached to new one. Colors matched on wires so I think no problem there.

I removed extra grounding from ECU casing, stil the same with engine idling.

However, when I revved up to 2000 rpm and above, I noticed first that 1st lambda voltage dropped almost to zero and stay very low, at the same time before closed lambda loop changed to open loop and 1st lambda did not make any sense. Voltage commonly stay low but some times got up to 500 to 600 mV and then dropped down again. 2nd lambda started to show high readings, 500 to 600 mV at revs.

When I got back to idle, lambda loop stay open and 1st lambda at zero or little higher.

I think that is what causes the fault code.

I could not yet get data to chart, because excel got confused for thousands on RPM data are separated with space, I'll try to get it sorted out so you can get better idea whats going on.

[Andy H]

Thanks for your very informatic reply!

I have Z22XE engine with original wirings from ECU to lambda, my new lambda required to cut old lambda connector to be attached to new one. Colors matched on wires so I think no problem there.

I removed extra grounding from ECU casing, stil the same with engine idling.

However, when I revved up to 2000 rpm and above, I noticed first that 1st lambda voltage dropped almost to zero and stay very low, at the same time before closed lambda loop changed to open loop and 1st lambda did not make any sense. Voltage commonly stay low but some times got up to 500 to 600 mV and then dropped down again. 2nd lambda started to show high readings, 500 to 600 mV at revs.

When I got back to idle, lambda loop stay open and 1st lambda at zero or little higher.

I think that is what causes the fault code.

I could not yet get data to chart, because excel got confused for thousands on RPM data are separated with space, I'll try to get it sorted out so you can get better idea whats going on.

Worth checking that they used the same colours........

[demolite]

Try to get a reading with the engine worked hard, as in open loop and running rich. Also if you can, measure the O2 signal wires directly at the connector (preferably at the sensor connector). This way you can see if there is voltage drop to the ECU on either the signal or the ground line. The sensor should be connected while doing this; you can use sharp long needles to backprobe the connector. When letting off the gas, low voltage is typical, but on high steady revs, it should be switching in the 100 - 900mV range. How's the car running at high revs? Do you have an exhaust leak?

I counted the O2 cycles on your diagram and got 16 cycles (450 mV to highest value then to lowest value and back to 450mV) in 100 seconds. That is very slow, a new sensor should cycle about once in 1-2 sec, although this is dependant on the ECU too. On my 2.5 both sensors cycle about 20 times a minute at idle, 40+ times a minute at 2000 rpm. What are your short term and long term fuel trim values? Better yet, can you include them in the next graph?

new lambda required to cut old lambda connector to be attached to new one

Attached how? Some sensors have an outside oxygen reference path inside the signal or ground wire (oxygen molecules move between copper strands). Although this is not true for all O2 sensors, I suggest they shouldn't be soldered for this reason. I have never seen one fail if soldered, but there is a chance.

Based on the slow switching and low voltage there is a possibility that the sensor is bad. Even new ones can fail, some already at the factory.

I don't want to believe the fault is with the ECU, they are usually pretty robust.

[mandula]

Here's some graphs, hope it makes sense..


RPM - lambda 1 & 2 - fuel trim - O2 loop (open/closed)
Cold start, waited for idle to settle and then revved couple of times.


RPM - fuel trim
Detailed view from above graph.

I have equipments to measure 1st lambda from connector, hope I have time to do that maybe tomorrow.


About lambda connection: new lambda is Delphi make, it had this kind of connectors on all four wires [link].

I cut old lambdas wires near its connector and attached this old lambda connector to new lambda, color to color. There was two whites, one brown and one black (if I remember right).


I don't notice anything abnormal when idling or revving the engine, no shaking or misfire-kind feelings. Idle consumption stays at 1,1 - 1,2 l/h (AC off), average cons. about 8,5-9,5 l/100km. Usual cold start idle consumption starts from 5 l/h and may take couple of minutes to drop to normal level. Gasoline does not smell out of exhaust gasses when hot engine, only at cold starts but not very noticeable.

Exhaust flex pipe replaced one year ago, also whole pipe from cats back. I haven't check condition of exhaust manifold, will do that tomorrow.

There is one thing that has troubled me for awhile: You can see from above graph that there is that period of time from cold start that idle is high and is settling down. Sometimes that idle is even higher than shown in graph (even up to 3000 rpm), and if I rev up the engine or start driving, revs can get "stuck" up, lets say if I rev up to 2500 rpm and release acc pedal, revs stay 2500 until drops after 10-20 seconds.

[demolite]

Something weird going on at 270 seconds on the first graph, just as you give it that last touch of gas before letting off. Looks to me like O2 sensor was disconnected at that point. Ecu notes fault and goes open loop. O2 signal comes back at about 290 seconds, ecu stays in open loop because it takes time to start "trusting" the O2 again.

Some theory on my deduction:

Ecu's use a bias voltage on the circuit to determine if O2 sensors are connected, working and are at operating temperature (they are producing voltage). A bias voltage means that an outside voltage is inserted into a signal to bring it to some level in case the signal is very weak. This bias voltage for narrowband zirconia O2 sensors is pretty much always around 450 mV, and the ecu feeds it into the signal path through a resistor. Why this voltage? A narrowband O2 sensor will produce this voltage only under a VERY tiny window of exhaust oxygen concentration. So tiny, that producing such a concentration for any prolonged amount of time is impossible with an engine. (Catalytic converter can produce this voltage for some seconds, see B1S2) For the engine O2, if the sensor voltage is in the area of 420-480 mV for over a few hundred milliseconds the ECU knows this voltage can't be from the O2 sensor but from its own biasing circuit; the sensor is disconnected, not at operating temperature or broken. I'm willing to bet that this situation is what's causing the fault code. The biasing voltage can be seen at cold sensor 0-20 s and possible disconnected/poor contact sensor at 270-290 s.

The disconnection could be caused by engine vibration inside the sensor itself, in the connectors, wires or inside the ecu. Bad contact (high resistance) is a likely culprit. Would also explain shifted low values.

By the way, there was a bulletin from Opel/Vauxhall on ecu's that had wrong kind of protection grease put on the ecu connectors at the factory. This caused poor connections in the ecu connector on some pins after many years from manufacture. Sorry I don't remember which year it was but doesn't hurt to check.

[mandula]

Thanks again for nice reply!

I doubt that lambda itself is faulty, because fault code did not disappear or change since I replaced it.

Can you tell witch pins on ECU should I measure between lamda connector and ECU?

Today I plan to measure wirings, clean groundings and ECU connectors and check lambda wirings.

[Kevin Wood]

My recollection is that the 4 pot Omegas use Titania Lambda sensors not the more common Zirconia devices as used on the V6s. This means they have different characteristics so you can't directly compare the output voltages. If it 'ain't broke...