NGK AFX wideband and E85

[LSxpwrdZ]

No there is no sense in it. You could however switch your programmable output on the LM1 to lambda and have it output 0.70-1.10 lambda and then enter that on the list portion instead of the transfer function so that 0v = 0.70 and 5v = 1.10. That will increase the resolution of accuracy on your analog output. I also used to have an LM1 and did this to narrow the range down so that small voltage fluctuations wouldn't affect the reading as much.

There is no need for commanded lambda because you have EQ Ratio. The EQ like stated earlier, is either equal to lambda or the inverse so there is no need in wasting a custom PID spot for commanded lambda when it can be integrated into the lambda error custom PID.

If you wanted commanded lambda just for a visual then you could make a custom PID named commanded lambda and it would either be 1/[PID.6010] or [PID.6010] depending on which OS you are running.

[98Camarod]

Can you post the config required for the lambda logging?

[jim2092]

No SENS.119 or PID.6010 in my very recent Beta for my 2005 C6 E40. Research continues.

[jim2092]

No SENS.119 or PID.6010 in my very recent Beta for my 2005 C6 E40. Research continues.

No solution has been found. An inquiry has been made on this forum with HP.

[DSteck]

I use [SENS.119] on E40s all the time.

[Atomic]

Just to add to the confusion. I log commanded AFR and display lambda error directly...

100*((1.4*[PID.2811]+9)/14.57-[SENS.121]/STOICH)/([SENS.121]/STOICH)

2811 is the analog out from my ngk controller.
"STOICH" is whatever you have in the air fuel ratio > stoich afr value.

[DSteck]

Boy you guys like to complicate stuff.

I'll post up the proper E40 lambda error PID soon. First, I want to let this boil a bit. Lol.

[jim2092]

Boy you guys like to complicate stuff.

I'll post up the proper E40 lambda error PID soon. First, I want to let this boil a bit. Lol.

[jim2092]

I use [SENS.119] on E40s all the time.

???? No sens.119 or pid.6010.

[LSxpwrdZ]

Instead of using [SENS.119] like I did in my Gen3 configuration use your [AUX.xxx] PID for the Lambda sensor. Then you'll have to scroll through the PID's to find the correct PID number for Equivalence Ratio. I know it's in all the LS applications I've ever messed with. In the end it is very simple:

Lambda Error % = 100*((Lambda-Commanded)/Commanded)

[DSteck]

Still cooking.

[jim2092]

http://i323.photobucket.com/albums/n...mbdaconfig.jpg


So I should reconfigure the EIO for lambda using the lambda sensor instead of the AFR, using all the appropriate values.

And use the pid.68 (only mention of EQ in the whole list) where EQ is called for?

[LSxpwrdZ]

Yep

Keep it simple, no need to over think everything with conversion etc. Once you get the wideband input as the lambda sensor it takes out all need of anything AFR in the scanner, even the custom PID calculations.

[jim2092]

" You could however switch your programmable output on the LM1 to lambda and have it output 0.70-1.10 lambda and then enter that on the list portion instead of the transfer function so that 0v = 0.70 and 5v = 1.10. That will increase the resolution of accuracy on your analog output. I also used to have an LM1 and did this to narrow the range down so that small voltage fluctuations wouldn't affect the reading as much."

So if I use the above numbers for higher resolution, then the custom wide band pid numbers from your previous post changes from "1/7.35 + 0.68" to what?

[jim2092]

Boy you guys like to complicate stuff.

I'll post up the proper E40 lambda error PID soon. First, I want to let this boil a bit. Lol.

And this information probably couldn't hurt either.

[DSteck]

I honestly haven't had time. I've been building MAP sensors nonstop and just screwed myself yesterday by buying Diablo 3. My free time is gone now. Lol.

[jim2092]

No there is no sense in it. You could however switch your programmable output on the LM1 to lambda and have it output 0.70-1.10 lambda and then enter that on the list portion instead of the transfer function so that 0v = 0.70 and 5v = 1.10. That will increase the resolution of accuracy on your analog output. I also used to have an LM1 and did this to narrow the range down so that small voltage fluctuations wouldn't affect the reading as much.

There is no need for commanded lambda because you have EQ Ratio. The EQ like stated earlier, is either equal to lambda or the inverse so there is no need in wasting a custom PID spot for commanded lambda when it can be integrated into the lambda error custom PID.

If you wanted commanded lambda just for a visual then you could make a custom PID named commanded lambda and it would either be 1/[PID.6010] or [PID.6010] depending on which OS you are running.

Can't seem to complete conversion to lambda using the .70 and 1.10 figures.

Is step one setting up the LM1 Programmer Analog out 1/2 to read 0.0 volts=.7 lambda and 5 volts=1.1 with "use lambda" selected and in my case to have the controller to read correctly-set the multiplier to 14.3 (tho not important)?

[jim2092]

Is step two to set up the LM1 in HP scanner as follows? (using your spreadsheet)

[jim2092]

[vltrb-0]

Is step two to set up the LM1 in HP scanner as follows? (using your spreadsheet)

your equation isn't right there

VOLTS / (5/DEVICE RANGE) + (MINIMUM VALUE OF DEVICE) = LAMBDA

so volts is variable pending sensor

then we get

volts / (5/(1.1-0.7) + (0.7) = lambda

the 5 here represents the voltage range of our EIO

so that leaves you with

volts / (5/0.4) + 0.7

volts/12.5 + 0.7 (you can also see from the scale on the right it isn't correct as at 0v you should get 0.7 and at 5v you should get 1.1)


also can you please create a log and upload it (include the cfg file)

i'd imagine your getting frustrated by now, keep reading and take your time, even leave the car for a few days. Taking your time is far easier then repairing a hand grenaded motor

i'm no expert by a long shot but i've found the only way to learn is to really understand the content your applying, every action has a reaction, use alternate calculations to qualify your findings, if you can't then you still don't understand the content and it's back to the drawing board