NGK AFX wideband and E85

[LSxpwrdZ]

Don't even put "Commanded AFR" on your table display! EQ Ratio is what you want. EQ Ratio by definition is the inverse of Lambda so to make a commanded Lambda PID you just use 1/EQ Ratio to get commanded lambda. The only time you should think AFR is when you are entering the stoich value of the fuel you are using into your tune file.

[seevi]

I understand using lambda error to tune the VE/MAF cells but isn't your commanded lambda created by dividing by the stoich AFR value of the fuel you are using?

[seevi]

Whoever started the afr trend needs to be shot!

Do yourself a favor and forget AFR. Learn Lambda!

The fellow who started this thread was asking for help tuning E85 fuel. Smacking his knuckles with the forum yard stick may not be the appropriate response here. Tuning is based on data input from your wideband O2 sensor which determines the mass ratio of air to fuel in the fuel you are burning (E85, E15, E10, Diesel#2, kerosine, Alcohol etc.) If you have the ability to convert your wideband controller to report Lambda, it will report a different Lambda number for each of those fuels based on its Stoichiometric combustion value. Using your tuning software to convert exhaust system data to a Lambda value to determine a correction value is a proven procedure.

Using the EQ in tunings: The equivalence ratio of a system is defined as the ratio of the fuel-to-oxidizer ratio to the stoichiometric fuel-to-oxidizer ratio. Once again, AFR raises its ugly head. You can't work without it!

[shitstang]

I still haven't gotten an answer to my question either, if I wanted a pissing match about how I should tune in lambda I'd go read the other 12 threads that are the same pissing match. All I wanted to know was why the ngk box was reading 9.8 instead of the 14.6 (maybe because of transfer function). The other question was how to get my ve afr error to calculate for e85. I'm not tuning for lambda I've tried the pids none of them work so I said fuck it, plus it's easier for my wife to read afr because the laptop isn't always hooked up. My question is how can I get the box to read correctly and ve afr correction, not grief about how I should tune in lambda. If I wanted grief I'd go argue with the wife.

[seevi]

I still haven't gotten an answer to my question either, if I wanted a pissing match about how I should tune in lambda I'd go read the other 12 threads that are the same pissing match. All I wanted to know was why the ngk box was reading 9.8 instead of the 14.6 (maybe because of transfer function). The other question was how to get my ve afr error to calculate for e85. I'm not tuning for lambda I've tried the pids none of them work so I said fuck it, plus it's easier for my wife to read afr because the laptop isn't always hooked up. My question is how can I get the box to read correctly and ve afr correction, not grief about how I should tune in lambda. If I wanted grief I'd go argue with the wife.

Send me a PM with your email and I'll try and help away from this distraction.

[LSxpwrdZ]

The fellow who started this thread was asking for help tuning E85 fuel. Smacking his knuckles with the forum yard stick may not be the appropriate response here. Tuning is based on data input from your wideband O2 sensor which determines the mass ratio of air to fuel in the fuel you are burning (E85, E15, E10, Diesel#2, kerosine, Alcohol etc.) If you have the ability to convert your wideband controller to report Lambda, it will report a different Lambda number for each of those fuels based on its Stoichiometric combustion value. Using your tuning software to convert exhaust system data to a Lambda value to determine a correction value is a proven procedure.

Using the EQ in tunings: The equivalence ratio of a system is defined as the ratio of the fuel-to-oxidizer ratio to the stoichiometric fuel-to-oxidizer ratio. Once again, AFR raises its ugly head. You can't work without it!

Simply put you are WRONG. If my wideband reports lambda 1.0 @ 2.5v output for gasoline, guess what it'll also read lambda 1.0 @ 2.5v for E85. Stoich is Stoich PERIOD!

To answer the original posters question was nestled in my earlier responses. To use afr error you have to use a completely different transfer function for the wideband input because now you would need to report to HPTuners 9.8 afr when the box is actually reading 14.57 on its display. This is why me and a few others that do this for a living gave the advice to get AFR out of your mindset and use lambda because like we have tried to drive, lambda 1.0 (stoichiometric for all fuels) is the same for all fuels.

[DSteck]

I no longer love seevi.

[jim2092]

Don't even put "Commanded AFR" on your table display! EQ Ratio is what you want. EQ Ratio by definition is the inverse of Lambda so to make a commanded Lambda PID you just use 1/EQ Ratio to get commanded lambda. The only time you should think AFR is when you are entering the stoich value of the fuel you are using into your tune file.

In the interest of learning this concept from someone who does this for a living and has no reason to mislead those of us who do it as a hobby, below I am listing the formulas I am using in HP. I wonder if you could specifically show/explain what needs to be changed using EQ ratio so that fuel type does not matter. I am using an LM-1 set to "14.7" and "air-fuel ratio" 0-5 volts.

(EIO) LM1 custom - volts/0.5 + 9.6 sensor - Air fuel ratio
LM-1 Lambda - (AUX.20121)/14.3 *E-10
Commanded Lambda - (PID.6001)/14.3
Lambda error - ([AUX.20121]/14.3)/[USER.9004]

[seevi]

Simply put you are WRONG. If my wideband reports lambda 1.0 @ 2.5v output for gasoline, guess what it'll also read lambda 1.0 @ 2.5v for E85. Stoich is Stoich PERIOD!

To answer the original posters question was nestled in my earlier responses. To use afr error you have to use a completely different transfer function for the wideband input because now you would need to report to HPTuners 9.8 afr when the box is actually reading 14.57 on its display. This is why me and a few others that do this for a living gave the advice to get AFR out of your mindset and use lambda because like we have tried to drive, lambda 1.0 (stoichiometric for all fuels) is the same for all fuels.

Note the title of the posting; NGK AFX WIDEBAND AND E85. The AFX reports AFR, not Lambda. According to NGK tech support, there in not enuf interest to add that option to their controller. So, I have two questions for you, 1) what kind of wideband are you using, and 2) what university awarded you a degree in engineering?

I don't tune for E85 every day, but I done a number of flex fuel trucks, a small fleet of Silverados and an Escalade or 2. To quote Joe Miller, "explain it to me like I'm a four-year-old," I use Banish's method to tweak MAF and VE tables using a correction factor until the error is below 3% using a custom, user defined wideband and PID's set up to work in E85 to achieve a 9.8 to 10:1 AFR (8.5:1 at WOT). Simple, Works for me every time. So, if what you use works for you, knock your socks off!

[seevi]

I no longer love seevi.

[LSxpwrdZ]

You are mis understanding what I'm saying... The display on the NGK does not matter! Obviously it is on a gas scale with a stoich of 14.57. The voltage outputs however are 9.0afr at 0v and 16.0afr @ 5v with a stoich ratio of 14.57 @ 3.979v. With that information we can back convert the values of the voltage scale to lambda which is 0.6177 lambda @ 0v and 1.0981 lambda @ 5v. Now set that input up as a custom pid in the scanner and you are now logging lambda into the scanner, which is the data that matters not what's displayed on the wideband screen.

Using EQ ratio to create a commanded lambda value is also easy and effective. The inverse of EQ is lambda (unless you are working on some early gen4 stuff where EQ is lambda). EQ of 1.0 is lambda 1.0. In PE on a NA gas engine commanding 1.13 EQ is actually 0.885 lambda. Very simple!

With everything setup that way all you need to worry about now is inputing the correct stoich value for your fuel type in the tune and then tune away! No swapping back and forth for different fuel types.

[DSteck]

Yea, the NGK display is useless to me. I only use the analog output. That's what matters. The NGK has a range of 0.62 to 1.1 lambda. Put tape over the gauge and don't even consider what it says.

[seevi]

You are mis understanding what I'm saying... The display on the NGK does not matter! Obviously it is on a gas scale with a stoich of 14.57. The voltage outputs however are 9.0afr at 0v and 16.0afr @ 5v with a stoich ratio of 14.57 @ 3.979v. With that information we can back convert the values of the voltage scale to lambda which is 0.6177 lambda @ 0v and 1.0981 lambda @ 5v. Now set that input up as a custom pid in the scanner and you are now logging lambda into the scanner, which is the data that matters not what's displayed on the wideband screen.

Using EQ ratio to create a commanded lambda value is also easy and effective. The inverse of EQ is lambda (unless you are working on some early gen4 stuff where EQ is lambda). EQ of 1.0 is lambda 1.0. In PE on a NA gas engine commanding 1.13 EQ is actually 0.885 lambda. Very simple!

With everything setup that way all you need to worry about now is inputing the correct stoich value for your fuel type in the tune and then tune away! No swapping back and forth for different fuel types.

I'm not sure who's misunderstands who here . If you read what I said above about setting up the custom wideband, we're saying about the same thing. I used Lambda to recalibrate the MAF and VE transfer tables. There's no doubt it's easier to understand and less complicated. You've gotten to 8 by adding 4 + 4 and there are some of us who get there by adding 5 + 3. The result is the same.

Again, I would like to remind the contributors here that this was a request for help, that probably added much more confusion to an already confused person, that turned into a pissing match over AFR v Lambda and procedure. These contests are easy to get sucked into and have no intelligent reward.

[jim2092]

You are mis understanding what I'm saying... The display on the NGK does not matter! Obviously it is on a gas scale with a stoich of 14.57. The voltage outputs however are 9.0afr at 0v and 16.0afr @ 5v with a stoich ratio of 14.57 @ 3.979v. With that information we can back convert the values of the voltage scale to lambda which is 0.6177 lambda @ 0v and 1.0981 lambda @ 5v. Now set that input up as a custom pid in the scanner and you are now logging lambda into the scanner, which is the data that matters not what's displayed on the wideband screen.

Using EQ ratio to create a commanded lambda value is also easy and effective. The inverse of EQ is lambda (unless you are working on some early gen4 stuff where EQ is lambda). EQ of 1.0 is lambda 1.0. In PE on a NA gas engine commanding 1.13 EQ is actually 0.885 lambda. Very simple!

With everything setup that way all you need to worry about now is inputing the correct stoich value for your fuel type in the tune and then tune away! No swapping back and forth for different fuel types.

In short, I understand you to say that if you set up the scanner to work in volts and convert directly to lambda instead of AFR you can tune any gas combination with no regard for its stoichimetric value.

If so, please use the data from post #28 for an E40 ECM (early gen 4) and give us the 4th grade explanation version of how those pids need to be set up.

[LSxpwrdZ]

First off what is the wideband analog output setup to show?

0v=X.XXafr 5v=XX.XXafr and I see it's set for 14.7afr for stoich.

Without that information I can't see what the actual tranfer function should be. Here is a spreadsheet I made to help with different fuels or to just configure the input to report direct lambda without creating a custom PID to divide the stoich afr to get wideband lambda.

http://www.hptuners.com/forum/showpo...6&postcount=51

[jim2092]

wideband

stoich - 14.7
0v afr - 10.0
5v afr - 20.0

HP Tuners

stoich - 14.3

[LSxpwrdZ]

Okay to setup your input to read lambda use this equation in your custom wideband PID. Be sure to set the PID to lambda instead of AFR.

1/7.35 + 0.68

That will make the wideband read lambda.

Now you will need to log EQ and verify it's actually EQ or Lambda. The way you can tell this is get into PE and see if Logged EQ goes above 1.0 or below 1.0. Most of the LS3/7 E38 stuff I've done EQ ratio = Lambda so it is just a simple Lambda error PID.

Lambda Error PID is this:

100*(([SENS.119]-[PID.6010])/[PID.6010])

If EQ doesn't equal Lambda (like the gen3 stuff and later gen4) then you'll need to take the inverse of EQ... 1/EQ=Lambda. That is very simple to add to the current lambda error PID. So it would look something like this:

100*(([SENS.119]-(1/[PID.6010]))/(1/[PID.6010]))

[DSteck]

Actually, E40s use EQ Ratio = 1/lambda. E38/E67 are EQ Ratio = lambda.

[jim2092]

Great looking info in posts 37-38. I will begin the absorption process and get back with any questions.

[jim2092]

Since the LM1 is programmable, is their any value to resetting stoich in my case to 14.3 in the LM programmer and then entering the resulting lambda analog output and divisor/adder values that result from your speadsheet in the LM programmer and user defined parameters respectively?

No need for a commanded lambda in this setup?