Lambda reading 1.75 always!

[stevesingo]

11APR11.e35

Current config

(72 KiB) Downloaded 1322 times

I have an issue with my Lambda sensor set up on M800.

Sensor is Bosch 0 258 017 066/67 LSU 4.9 (originally ford application)

Once the warm up is complete the value goes straight to 1.75.

No error in the diagnostics.
No exhaust leak.
Engine idling with no miss or excessivley lean/rich mixture.
Engine at 2500rpm=no change.
No exhaust leaks.

Config file and Logged data file attached.

Many thanks

Steve

[Polux RSV]

Did you try air calibration ?

Angelo

[stevesingo]

Air calibration is fine. 1072.

I have another sensor to try tomorrow and I'll double, re-check the wiring.

Thanks

Steve

[Polux RSV]

Looking at your data, when you pressed three times the throttle, the lambda value fall also to about 1.3, with some delay. So it seems working. Admitting your setup and calibration are correct, it seems your exhaust have "negative leaks" , fresh air is entering and fooling the sensor.
And where is located the sensor. It could be too close to the output, measuring fresh air entering from the outlet during exhaust pulses.

Angelo

[stevesingo]

I am fairly certain there is no exhaust leaks. The sensor is mounted 650mm from the engine flange in the header. The header is connected to the exhaust system which is 3+metre. No leaks visible and it is all new. I have had previously a Innovate LC1 mounted in the same place and got consistent results.

Double checked the wiring all OK.

Tried different sensor, air cal before fitting, no change.

I did some more logging and found that the La1 Ipn is fluctuating with changes to mixture; 2.5mA at 0 fuel trim, but when adding fuel, I get a slightly richer reading on the La1, but the idle becomes unstable at +35% fuel trim before I get to La 1.00. With the richer La1 reading I get a lower value on La1 Ipn.

When leaning the mixture, I get a rich response on the La1. Taking away 10% fuel trim produces a 0.5 reduction in La1, another 10% gives a further 0.2 reduction with La1 sitting around 1.05. 25% reduction in fuel causes unstable idle and La1 remains 1.1-1.3. Again with the richer La1 reading I get a lower value on La1 Ipn, but it never reaches 0.

I am beginning to believe the the Characteristics of this sensor are different to the default characteristics specified in the M800 software. In that the M800 default characteristics for the LSU 4.9 state at Ipn 2.06mA the La1 should read 1.75. My sensor Ipn is 2.6mA so will always read 1.75 until it drops below 2.06mA. The lowest I saw was 0.65mA.

Does anyone have any suggestions at to why this might be? What is strange is that when I lean the mixture I get a richer reading!

Any suggestion welcomed.

Steve

New Bosch 0 258 017 025 LSU 4.9 ordered as a known quantity.

[IDP]

I think the sensor part number in your original post is a LSU 4.0 not a 4.9

[stevesingo]

I will look in to that tomorrow.

[stevesingo]

It definitely has a 4.9 connector.

Looking in the M800 software, the characteristics are the same in relation to Ipn/lambda for all LSU sensors using Characteristic Table #38.

Any Motec staff care to comment?

Steve

[stevesingo]

I have fitted the new sensor and the result is in the log file attached.

As you can see adding fuel in 5% increments has no difference on Ipn (mA), which if I understand correctly, should fall with the richer mixture and become negative when richer the La=1. At +20% fuel the idle becomes unstable!

Strangely, when I remove fuel, I get a reduction in Ipn albeit, fluctuating, until the point where the engine fails to idle at the desired rpm.

I tried a similar test at 4k rpm and this gave similar results. No change

What should the Ipn be in free air?

Had a good look at the exhaust and can't see any evidence of leaks, which should be easy to spot as it is all new.

Any thoughts.

Thanks

Steve

[Blu302]

Do you have a high back pressure?

Try this page for some basic lean and rich testing: http://wbo2.com/lsu/position.htm
That website has a good amount of info on wideband sensors.
The different sensors (4.0/4.2/4.9) have different wire arrangements and different plugs. There are now 6 cables in some "5-wire" sensors. The 6th is an Rcal wire with a laser trimmed resistor for calibration. Check the site above as it has the connectors and wiring as well.