Reading AFR off Fuel trim and Narrowband?

[ben_r1]

Hi all, I don't usually delve into engine management so this may be a stupid question, feel free to correct me in any way you see fit. Hopefully some of the more clued up members may be of assistance.

Basically I'm going for a remote fastchip remap on my meg 265 but the generic map is for standard cars. Mines not exactly a snarling time attack machine but does have a Scorpion exhaust, KTR intercooler, Forge recirc and panel filter.

I know as with the Bluefin remote maps for the 250 many users have suffered with lean running, some to the point of the car going into limp mode! I know the issue with remote maps is this is always a small risk so I'd like to keep an eye on the AFR of my very expensive engine.

With the Fastchip you can read ECU including STFT, LTFT and Lambda voltage. In the absence of a wideband can these be used to reliably get an idea of what's going on or is it best to fit a dedicated Wideband sensor and controller?

 

[Chip-mk1]

Its a narrow band sensor, so all it can tell you is if its richer or leaner than its band.

So for example at 14:1 which is engine meltingly lean, it will say "Rich".


You cant tune a turbo car reliably without a wideband.

 

[ben_r1]

Found this online which apparently corresponds to a narrowband reading on a meg 225:

A/F ---- V
16:1 --- .76
15:1 --- .80
14.7:1 - .81
14:1 --- .84
13:1 --- .89
12.5:1 - .91
12:1 --- .93
11:1 --- .98

 

[ben_r1]

Its a narrow band sensor, so all it can tell you is if its richer or leaner than its band.

So for example at 14:1 which is engine meltingly lean, it will say "Rich".


You cant tune a turbo car reliably without a wideband.

Are the Fuel trims of any assistance in determining what's happening?

 

[Dan@SJM]

Narrow band isn't accurate enough at full throttle, it's very accurate between 14.3 and 15.2, and a wideband is fairly accurate between 10 and 18:1. And the fuel trims are relying on the narrow band.

 

[Chip-mk1]

Found this online which apparently corresponds to a narrowband reading on a meg 225:

A/F ---- V
16:1 --- .76
15:1 --- .80
14.7:1 - .81
14:1 --- .84
13:1 --- .89
12.5:1 - .91
12:1 --- .93
11:1 --- .98

If it really is capable of reading from 11-16, its not really a narrowband.

Trust it if you like, but personally I wouldnt.


The shape of a narrowband output voltage is normally more like this:

As you can see, its only really designed to be use like a switch to say above or below, not give a specific value, even a very small change in voltage once you are either side of the narrow point, is a big change in AFR, and hence any inaccuracy will be massively exagerated.

 

[Chip-mk1]

Are the Fuel trims of any assistance in determining what's happening?

Not really no, as the fuel trims are from a set point(s) in the map (henk may know what they are on your car, I dont) where it samples AFR and then it tries to guess how the rest of the map should be effected, its notoriously unreliable the moment you modify the engine as you are changing the VE of the engine its making assumptions about (cams are the worst thing for it)

I am of the opinion that anyone mapping a turbo car without a wideband, is taking risks that I personally wouldnt take, cant really say anymore than that, if you talk to Henk and he is happy to do it then thats cool as unlike me he probably has experience of literally hundreds of this exact car at the spec you have, so he can base his figures on those results, but its never going to be truely correct for your car if its based on assumptions and inaccurate readings, thats not to say that it wont work, just that it will be unlikely to be the best it could be.

 

[ben_r1]

Is there any type of t-piece for exhaust sensor so it can run the wideband for data logging and the narrowband OEM or do they need to be in direct flow of the EG?

Don't really fancy drilling and welding my new exhaust system.

 

[Chip-mk1]

Needs to be in direct flow and as close to the front as possible really so you get a fast enough response to know that the value you are seeing is whats happened very recently not 2 seconds ago, as if its not updated very fast, it will be logged against the wrong point in the rev range.

when I use a tailpipe sniffer I have to take this time delay into account when working out which cell is really the one that needs adjusting, but to do that takes a while making a few changes and seeing where they land to work out what the delay on the car in question is.

 

[Chi]

I think if it were my car Id be getting a custom map..

 

[Chip-mk1]

I think if it were my car Id be getting a custom map..

Agreed, would make more sense really to just take it to MWM or RSTuning or SJM and end up with a car back that is how it should be.

 

[ben_r1]

But the beauty of the RSTuner is that I have 2 megane and a clio and all three can be mapped for nearly half the price of getting every one custom mapped, and it has useful features such as data logging and fault code reading.

RS Tuning is a 6 hour round trip for me as well although in an ideal world if have it booked in there yes. So in conclusion remote maps are bad in any shape or form?

 

[Chi]

Definitely not but for a 265 id prefer peace of mind..

 

[Chip-mk1]

But the beauty of the RSTuner is that I have 2 megane and a clio and all three can be mapped for nearly half the price of getting every one custom mapped, and it has useful features such as data logging and fault code reading.

RS Tuning is a 6 hour round trip for me as well although in an ideal world if have it booked in there yes. So in conclusion remote maps are bad in any shape or form?

A remote map done by someone conservative should be as safe as a generic map, but it wont ever be able to be as correct as if someone has the car in front of them or at the very least has accurate data logs from a wideband, in that context the remote map should work well (although they still cant truely optimise the ignition of course!)

 

[ben_r1]

Why don't manufactures fit wideband to aid more control over the engine from the factory?

If the pre cat lambda controls AFR ( I assume ) what's the purpose of the post cat lambda?

 

[Chip-mk1]

Why don't manufactures fit wideband to aid more control over the engine from the factory?

Cost, although lots do, VAG in particular.

If the pre cat lambda controls AFR ( I assume ) what's the purpose of the post cat lambda?

To see if the cat is working


On a 172 you can safely remove the second one and put a wideband in there during mapping.

 

[ben_r1]

Could you not replace the pre cat lambda with a wideband and modify the ECU to accept the new signal in place of its original, or is this outside the scope of an OEM ECU?

 

[Chip-mk1]

Could you not replace the pre cat lambda with a wideband and modify the ECU to accept the new signal in place of its original, or is this outside the scope of an OEM ECU?

The ECU is no doubt capable of it, but actually altering the software would be a hugely involved task, as its not like renault are going to hand over the source code and these things are notoriously difficult to reverse engineer, and what would be the point really?
Widebands also go out of sync often too, so it would take huge amounts of time on R&D to accurately spot when the sensor was giving an untrue reading and should be ignored etc.

It really isnt something you would want to retrofit, the cost would probably run into the millions to do it properly, where as you could just fit an aftermarket ECU that is designed to allow such things to be added?

 

[ben_r1]

Okay got it. Thanks for your time everyone really appreciate it!

 

[Faboka]

Sorry to high jack but what wideband kits do people recommend? I looking for one to play with mapping my 182 at some point but I want one that I can log against RPM. Was thinking the innovate lc-2 hooked up to the iphone kit.

 

[Chip-mk1]

The innovate is ok, uses the same cheap bosch sensor that most of them do, you'll struggle to get a lot better with spending silly money.

 

[Dan@SJM]

I use an LM2, and record my data onto an SD Card

 

[Faboka]

Thanks.

Dan do you have the lm2 plug into the obd port to capture rpm?

 

[Dan@SJM]

Yes,

I log:
TPS %
RPM
MAP
ECT
AIT

 

[Faboka]

Cheers Dan