Lambda Average Error

[Kurt_XK]

Hoping someone can help me out.  I am receiving a "CL Lambda Status:     Lockout: Lambda Average Error"   I am trying to dial-in my base map and this error condition is making it challenging.  It seems to occur when TP goes to "0" and the engine is coming down to idle.  I have spent a fair amount of time in the idle ranges trying to get the engine to run properly. It would make sense that I have had some real crazy values in there that may be skewing the "average" something wicked.   

Once I start driving, the O2 sensor wakes up, does its thing and starts providing values but it takes forever and by then I am usually out of road...  

I am running a single wideband sensor. 

Thanks

Monday late run.llg

[Adamw]

You have lambda error 2 showing and it seems to appear quite regularly whenever RPM drops below about 1500.  I would be suspicious of maybe the 12V supply to pin 4 of the lambda sensor is dipping in voltage.  The ECU is reporting decent battery voltage but I dont know where the lambda probe gets its 12V from.  I would try investigating that in the first instance, maybe even run a temporary wire down to the sensor from the battery or similar as a test.

[Kurt_XK]

Adam,  

I verified my wiring according and connections. Everything is correct. As far as the 12V supply it shares the same circuit as my Engine Position Module, I would find it odd that would be drawing enough current to effect the performance of the wideband. I am willing to give it a try, It just pains me because this is some of my cleanest wiring and I hate to hack it up. The 12V power supply is fed directly from the battery terminal, runs through a relay, a fuse and to the load.   In searching for help with this I found this: 

On 7/4/2011 at 11:29 PM, Guest Simon Willis said:

I have been using Mixture Map tuning for light loads. It is working brilliantly and really seems to have improved my car. I think there are a few things that could be added to the Mixture Map to make it even more powerful:

-Once a cell has been corrected, it would be good if the calculation of average AFR discarded all previous samples. If you update the mixture map after making a correction then the average AFR is affected by the pre-correction samples. I usually make a whole load of corrections and then restart the datalogging to observe the affect.

To me this really describes what a "Lambda Average Error" is.  I realize I am focusing on this and not the lambda error 2 which is present, but I appear to be getting data even when this is present.  Just one other piece of evidence. Here is a screen capture of the mixture map.... Zones in red seem to correspond to the zones where I have the average error. Zones in green, I have data. I am learning how to use this program more and more everyday, but this just seems to smell like a software or setting issue.  Thoughts? Should I just shut up and run the jumper?  Thanks.  

 

 

[Vaughan]

If you look at your Lambda 1 value it spends a lot of time at 0, neither CLL or the mixture map can do anything with a lambda value of 0 so you need to sort that out.

Does the Lambda come in through analog or CAN? can you share a copy of your basemap?

[Kurt_XK]

I've worked on this base map a bit more, but rev 5 is untested. So this was what I was working off of. 

Jaguar XK rev4.pclr

[Adamw]

3 hours ago, Kurt_XK said:

To me this really describes what a "Lambda Average Error" is.  I realize I am focusing on this and not the lambda error 2 which is present,

No that doesnt describe it at all.

Lambda average is the average of all of the lambda sensors that are connected to the ecu (you can have up to 8).  The way the fail safes work is if a sensor controller reports an error then it's lambda is zeroed out, the lambda average calculation ignores any sensors that are reporting zero lambda and removes them from the lambda average calculation.  In your case you only have one sensor and it is reading zero lambda so it is being removed from the lambda average calculation.  The lambda average error is because it cant generate a lambda average when there are no sensors left that are working.

 

3 hours ago, Kurt_XK said:

I am willing to give it a try, It just pains me because this is some of my cleanest wiring and I hate to hack it up. The 12V power supply is fed directly from the battery terminal, runs through a relay, a fuse and to the load. 

You dont need to hack anything up, just pull the existing #4 pin out of the loom plug at the sensor and run a new temporary wire direct to the battery.  It could just be a bad crimp or similar.  This is the easiest thing to eliminate, really the only other possibilities are a faulty ecu or faulty sensor, both of which will cost money to eliminate.

[Vaughan]

Looks like you are using the internal wideband controller and it is getting Lambda errors which are mostly error 2 (as @Adamw pointed out) might need to be sent back to have the servicing guy look at.

[Kurt_XK]

I realized I didn't answer your CAN or Analogue question. I have a Fury, I have it wired into the designated internal lambda. I am not sure if that makes it CAN or Analogue.  

Vaughn and Adam,  thanks for the Troubleshooting help. I appreciate the expanation of the error code. That makes sense now. I will start with the 4 pin as Adam suggested. 

[Kurt_XK]

I wired the 4 pin directly to the battery and observed similar results. I also adjusted my runtime parameters to include the lambda errors and noted "error 52."  I thought perhaps there maybe a correlation with battery voltage, and the average error,  but after reviewing all the times the lambda1 dropped out, I could find no correlation between it and battery voltage.

UPDATE: I wired everything back to how it was and did some more test drives. I did note the occurrence of an "error 4."  These all seem to be pointing back to a internal issue. What about issues with the lambda sensor itself?  Is there any way to rule out the sensor as an issue? 

 Driveway run.llg