Lambda EMAP compensation

[Slim47]

Hello

I have pre-turbo individual LSU 4.9 installed that are connected as +5V sensors. Does anyone have knowledge of the Target AFR adjustment table for EMAP? Perhaps someone has already set up LSU 4.9 and has experience in this.

I see something similar in the first approximation:

[Vaughan]

Are you using Link CAN Lambdas or a different Lambda controller?

[Slim47]

13 hours ago, Vaughan said:

Are you using Link CAN Lambdas or a different Lambda controller?

I am using different lambdas (AEM UEGO).

[Adamw]

You cant do press compensation on the already computed lambda value, the compensation needs to be applied to the measured pump current before that is converted to lambda in the controller as neither the compensation or Ip/lambda relationship are linear or proportional.  You will need a wideband controller that can do the compensation internally.  

[Peace]

20 minutes ago, Adamw said:

You cant do press compensation on the already computed lambda value, the compensation needs to be applied to the measured pump current before that is converted to lambda in the controller as neither the compensation or Ip/lambda relationship are linear or proportional.  You will need a wideband controller that can do the compensation internally.  

Adam
Can you advice controller which can do this job?
There were AEMone, but i believe it is discontinued now.

[Slim47]

1 hour ago, Adamw said:

You cant do press compensation on the already computed lambda value, the compensation needs to be applied to the measured pump current before that is converted to lambda in the controller as neither the compensation or Ip/lambda relationship are linear or proportional.  You will need a wideband controller that can do the compensation internally.  

Why can't I apply an adjustment to an already calculated value? Bosch, as I indicated above, shows the dependence of the adjustment on the Back pressure.

Maybe I don't understand something (

From the Bosch data sheet, we see that at 120 kPa of overpressure in the exhaust manifold there is a 40% error. That is, if the target lambda is 0.78, then this results in 0.088 ((1.00-0.78) * 0.4) errors, that is, we must target 0.69 (0.78-0.088) to get the true lambda at 0.78

[Adamw]

The Y axis of the pressure compensation table is not lambda, it is the % change in measured current that was required to drive the ion pump, often designated "pump current" or "Ip".  Ip is related to lambda, but not directly.  You need both the sensor temperature and the Ip passed through a lookup table to derive Lambda.  The relationship of temperature and Ip to lambda is not linear or even close to linear - a 5% change in Ip is not a 5% change in lambda.    

FYI that datasheet you have is also for an ADV sensor, not the 4.9.  The 4.9 is actually much less sensitive to pressure, it has a max Ip difference of about 15% at 2.5Bar.  

[0x33]

The real question is, when is Link going to release a multi channel (6 or 8) lambda controller. Daisy chaining 7 CAN Lambdas to do the job looks messy as, would be nice to have everything contained in one unit. 

[Rozsko]

But evennif you don't do compensation, the point of pre turbo individual widebands is to balance out the cylinders. No?

[Slim47]

15 hours ago, Rozsko said:

But evennif you don't do compensation, the point of pre turbo individual widebands is to balance out the cylinders. No?

In addition to the fact that the cylinders need to be balanced, you need to know the true value so as not to burn the engine

[Slim47]

23 hours ago, Adamw said:

The Y axis of the pressure compensation table is not lambda, it is the % change in measured current that was required to drive the ion pump, often designated "pump current" or "Ip".  Ip is related to lambda, but not directly.  You need both the sensor temperature and the Ip passed through a lookup table to derive Lambda.  The relationship of temperature and Ip to lambda is not linear or even close to linear - a 5% change in Ip is not a 5% change in lambda.    

FYI that datasheet you have is also for an ADV sensor, not the 4.9.  The 4.9 is actually much less sensitive to pressure, it has a max Ip difference of about 15% at 2.5Bar.  

Complete answer as always. I am grateful to you. You make a huge contribution to our education )

[mapper]

Usally I just use one lambda after the turbo for overall adjustment to target lambda. The individual before read richer, however it doesn't matter, you just need to balance them. 

[Slim47]

4 hours ago, mapper said:

Usally I just use one lambda after the turbo for overall adjustment to target lambda. The individual before read richer, however it doesn't matter, you just need to balance them. 

Yes, this approach works.  But it reminds me of a crutch when you add another lambda being unable to correctly set up the 6 available ))