Lambda reading problems

[kielcivic]

There are some on here who will say each and every sensor needs its own calibration ( que Rice Racing ) and they no doubt have their evidence to support that. So maybe that might be an option for you to to have your sensor given a test and calibration chart to use.

I've only done a few cars, but never had any issue that caused me any concern. I did also try a couple of sensors in the same exhaust pipe when I changed to the S8 and could do so and again saw nothing that gave me any concerns about them reading different.

Obviously each and every case may be different. As you have the AEM fitted to the car, wire it's output into one of the analogue inputs on the ecu so you can log it too, at least then you can see what both are doing on the same graph. You could even choose to use that as your sensor for any corrections to be applied.
Might also be worth trying with closed loop off too.

I use the LZA09-E1 on my own car and others and is what I tested. I also tried an original L1H1 for a short period too, and it read pretty much the same, with the same linearization as the LZA ( both supplied graphs were the same anyway )

I have a PLX wideband, LSU4.9 on one of my exhaust pipes as well as the Syvecs, purely as I've always had a separate wideband display in the car anyway, but I log this on the Plex dash as well as the Syvecs from both pipes.
Again, other than minor differences at times which is to be expected, it will always give similar readings.

None of the above a great help....although around idle with only 12" from fresh air, not sure I'd be expecting to get honest readings from any sensor, so maybe one is just more sensitive to this than the others.

On power once sufficient volume of gas is moving, it should probably be ok though.

Although if this fuel issue is also affecting boost, it must really be huge AFR differences ?

Hi steve
Can I run my aem through the syvecs as well as the ntk?
I think your right about the reading with how the exhaust is! How far away from the turbo would you fit a lambda sensor?

[RICE RACING]

NTK's are more temperature tolerant than Bosch units but still you can make them do weird things if the bung temp gets too high.
On turbo rotaries I run them about 1 meter away from the turbine outlet. But it sounds like you have air coming back into the system from what Stevie is saying?
Your EGT is very low on the run you sent me so no reason why you could not run the sensor closer to the turbo outlet (right on it) like this below.
This is using a racing quality NTK sensor with special heat sink I made. NOTE: This uses one of my self made internal 347SS remote sampling chambers using venturi effect inside a solid copper heat sink. I tested this to constant 1150deg C for 2 hours straight on the sample probe end and the hex temp was never over 550 deg C

[stevieturbo]

As Peter has covered above...it kind of depends on what EGT's you are seeing..and overall design.

Clearly the trend with these drag cars now is effectively no exhaust, which clearly makes things difficult for lambda sensors.

So for me, anything I work at has a full exhaust, so ensuring the sensor can run cool but also nowhere near fresh air is not a problem.

Something like Peters adaptor should cover you for both the above scenarios though

But yes you could simply just assign a spare AN analogue input to the AEM wideband and configure lambda 1 ( or lambda 2 ) to that input configured as a 5v input and sort your linearization as per AEM's schedule
Probably need to assign lambda control for all cylinders to say Bank 1 and so only Lambda input 1 would be doing any closed loop work, the other Lambda channel would just be for logging/sanity purposes