How to get fueling perfection

This does not apply to Power Commander users. Look away.

Got a custom map? Got a little dip or a fluffy spot here or there? Here's a way to fix it.

Items required:

1 x wideband sensor and controller with programmable narrowband output
1 x 15 ohm 50W resistor
1 x 12V relay

Wideband controller kit:


[You must be registered and logged in to see this image.]


Dummy load (15 ohm 50W resistor)


[You must be registered and logged in to see this image.]



The stock narrowband sensor cannot be used trim to just-on-the-rich-side-of-lean-best-torque where the 8V sings. The narrowband operates in the range of 14.5-15.5 AFR which is great for your scooter to nip down the shops or Joe average Cage Driver who thinks cars and bikes are just for transport. It also helps in passing emissions tests. Those who listen to me blather on about the 8V will know that the 8V is not your average motor.

Furthermore, the 5AM ECU is coded to to receive a narrowband voltage, which is in the range of 1.1V-0.1V where 1.1V = 14.5 AFR and 0.1 = 15.5 AFR or thereabouts. Wideband sensors operate in the 0-5V range. You cannot connect a wideband sensor directly to the ECU, because you will fail epically and get a nice red triangle and SERVICE warning for your trouble.

The ECU can be tricked however. Get yourself a nice wideband controller and sensor. There are several on the market that can be programmed to simulate a narrowband output. In this case, I'm using an Innovate LC-2.

The stock sensor operates on ~9V. The ECU also monitors the current for the heater in the sensor. The wideband controller requires a 12V supply, so you cannot simply connect the wideband controller across the heater supply from the ECU. If you connect the controller directly, it will not function and you'll get the SERVICE warning again.

The O2 sensor typically requires 3-5 amps to function. Normally the ECU monitors the current to confirm proper function of the sensor heater element. This is where the resistor comes into play. It will be used as a dummy load to trick the ECU into thinking the heater is fully functional.

Program the controller to simulate a narrowband output. In this case, I used 1.1V=12.2 AFR and 0.1V=12.8 AFR. This equates to Safe-best-power-at-WOT for most engines,  but for 8V it's just a tad rich lean-best-torque.
The output from the controller must be fed into the ECU S+/S- lambda inputs. I butchered the original narrowband sensor and utilized the connector. The output from the ECU to the heater is connected across the dummy load plus the positive side to the relay coil. The contacts of the relay are then used to supply +12V to power the wideband controller.

Circuit diagram:

[You must be registered and logged in to see this image.]



Lambda must be switched on in the map. The ECU trims the fuel to the perfect AFR. My custom map for my Mistral-Arrow exhaust had a little dip around 4000 RPM. The power delivery after trimming is now perfectly linear. In fact it's so good now I don't think I'm going to bother with the custom Ti pretzel.

Controller tucked away under the seat:

[You must be registered and logged in to see this image.]


I clamped the dummy load to the frame, which acts as a big heatsink.

[You must be registered and logged in to see this image.]



Power delivery is superb. Fuel economy is excellent.

[Them] Wait, what? Aren't your maps already perfect?

[FIG] Yes, but unless your bike is tuned exactly like mine, there may be a few differences.

[Them] OK, so we could use this application with any map?

[FIG] Almost. There would need to be some nod to the pipe you're using, with proper tweaks to the  correction tables, and not just fuel thrown into a rev range, 'cos that's wot Mick did and his bike goes like a shower of shit'

[Them] So some semblance of a well made map would be required?

[FIG] Yes indeed.

[Them] I think I love you..

[FIG] I get that a lot..

Well, the wiring diagram looks simple enough that even I could attempt it!

Pete

Simply magnificent Mark. Excellent results without too much effort is seems.

Agreed Al, but it's a $200 fix that some may think unnecessary. I can't help myself though..

beats the hell out of fixing the fueling on a Yamaha FZ09!!

would there be any benefit to having a dual o2 sensor controller? one sensor for each head-pipe?

i was hoping you'd do a write up on this!

Brad, it's not going to be easy. You'd need to use all or part of a Stelvio or Norge wiring loom (which are wired for dual lambda) and flash the ECU with a modified dual lambda map, then try to figure out a way to eliminate all the SERVICE warnings you'd get from the lack of ABS and other gizmo's those bikes have that the GRiSO doesn't. Plus the different dash. Ay, Caramba! Too hard.

Because I've got the gear I'll do it when I get the chance.

So with the lamda back on does the ECU *remember* the trims like the stock CL system or does it remain fully dynamic?

Pete

It's just like the stock CL system. It will remember, but they are still volatile. When you disconnect the battery etc, it will forget.

Gotcha.

Soooooo Beetle, are going to run with this perminantly or are you turning off the lambda again when it's all sorted out ?

gotta love creative people...

Ron, if I could figure out how to backup the trims and restore them as needed I might consider turning lambda off. I'm contemplating how to add a battery or capacitive based backup to to save the trims in case of a blown fuse or battery disconnect.

There's a few things I am missing here but it's so interesting I want to understand!

Why should you turn the lambda off? Isn't it working great like this?
What would happen if the battery should go off?

The main issue I imagine would be that every time you loose the trims it's back to square one. Also, because the lamda intervention remains dynamic one assumes that if you end up doing something spectacularly out of the ordinary the ECU will trim to and 'Learn' that condition and then will have to 'Un-learn' it all over again when conditions return to normal. If the machine was runing spectacularly well it would be a shame to loose that due to a temporary inconvenience.

I may be reading that all wrong though. Mark's field, not mine.

Pete

Pete Roper wrote:

The main issue I imagine would be that every time you loose the trims it's back to square one.

Absolutely. That's why some sort of backup would be great.

Quote :

Also, because the lamda intervention remains dynamic one assumes that if you end up doing something spectacularly out of the ordinary the ECU will trim to and 'Learn' that condition and then will have to 'Un-learn' it all over again when conditions return to normal.

Correctamundo. If you have some sort of mechanical issue that affects the fueling, the good 'ol ECU will do its best to trims the problem away. If lambda is switched off, the trims will remain intact so that when your get it fixed or serviced, the ECU doesn't have to retrim all over again. Also, what happens if you have an issue with your wideband gear? It may trim all the goodness away again.

Quote :

If the machine was runing spectacularly well it would be a shame to loose that due to a temporary inconvenience.

Yes indeed. What happens when you have to replace the battery, or a fuse blows. Trims gone. So you switch lambda back on and connect all the gear up again. Then ride 1000km to get it back to where it was. How bothersome!

Are the trims the "self learning parameters"?

In stock map conditions, are they lost when battery is disconnected?

In my experience disconnecting my battery for a few hours makes the bike run smoother, at least for a few days...

voulga77 wrote:

Are the trims the "self learning parameters"?

Yes!

Quote :

In stock map conditions, are they lost when battery is disconnected?

Yes!

Quote :

In my experience disconnecting my battery for a few hours makes the bike run smoother, at least for a few days...

That's because with the untrimmed stock map, the bike will run more rich. Then the ECU trims the buggery out of the fuel and you go back to lean, hunty, GRiSO.

Mark, my output from the ecu to the lambda heating gives 12 volt, or it it under load it drops to 9v? I use it to power the kkl.

Paul, it's 12V to ground, but only 9V across the ECU pins. I think it uses a shunt to measure current.

hey Mark, i have a few questions regarding this.
firstly, there are two versions of the lc2. one 3ft version and one with 8 ft cables.. i'm guessing the 3 ft is more than sufficient.

second, i'm still kinda confused on how this works exactly. i understand that it takes a wideband signal and converts it to narrowband.. but if the ecu can only read and USE a narrowband signal.. how does this improve over the stock o2 sensor which is installed in the pretzel?
am i wrong in understanding that this takes a wideband signal and converts it to narrowband 1.1-.1v?

i definitely plan on using this setup, i just have to know how it works exactly, and why it's better than the stock o2

I think the trim that the ecu learns isn't more then an multipier number. Nothing fance as another correction curve. Don't forget they only made a lambda to pass euro 3

Mark, now you can compare the lambda values that you get with the lambda feedback against without, what does it tell?

Brad,

As I mention in the OP, the narrowband sensor only operates in the 14.5 - 15.5 AFR range. Any AFR outside this range will lock the output at 1.1V or 0.1V respectively. The simulated narrowband signal from the wideband sensor can be set to operate at any AFR range. The output voltage will swing between 1.1V and 0.1V simulating AFR of 14.5 -15.5 to fool the ECU, but the actual AFR will be whatever the controller is programmed for.

Clear as mud?


Oh, and 3 foot cable is more than long enough.




Paul, I haven't checked, but I will eventually. I haven't been near the GRiSO for almost a month. Too busy with boring work/family stuff.

Mark,
Yes thank you. Now i get it lol. Instead of getting the dme to read a wider range of afr's we are lying to it. That's brilliant!
Can this be made to achieve a ratio of 14.7 during cruising and light throttle applications, as well as 12.8 during WOT? I am already sold on this mod but that would be the icing on the cake.
Those afr numbers are theoretical of course. Idk which values this engine prefers for best/max power, and best fuel economy.
Thanks
Brad