Wideband O2 sensor

[jammess]

Ordered an innovate Motor Sports LC-2 controller with wideband O2 sensor today from Amazon. With this system I'll be able to program virtually any AFR for closed loop operation and I will be able to observe what happens in real time and save it as a chart recording on my laptop. I think the Yamaha ECU AFR at cold startup is fairly rich and during the warmup period is influenced by the CO settings. When the bike reaches operating temperature at idle I think the ECU goes closed loop and AFR is determined by input from the O2 sensor with the resulting AFR being 14.7:1 with no influence from the CO settings. My plan is to start out with a closed loop AFR of 13.5:1 and see what happens. I hope to see an improvement in idle smoothness with less tendency to hunt. I think I'll be able to know exactly how the CO settings are effecting AFR during warmup.

I have an LC-2 with wideband O2 sensor on my '93 R1100RS that I run closed loop at 12.9:1 which results in very smooth engine performance while still getting acceptable fuel economy in low 40s for mpg. The BMW Bosch Motronic ECU has adaption capability meaning open loop operation is modified over several tanks of gas by closed loop settings. I would imagine the FJR ECU operates in similar fashion. The LC-2 has two analog outputs so one can program two closed loop AFRs and with a toggle switch one can select one of two AFR settings to get a direct comparison on a test ride. Hot rod tuners consider 12.8:1 to 12.9:1 to be the AFR for lean best power.

What I should do is stop screwing around with this beast and just ride it.

 

[Fred W]

What I should do is stop screwing around with this beast and just ride it.

Now, where's the fun in that?

So, you will be installing the wideband O2 sensor in a new port and leaving the stock O2 sensor in place? I would be interested in establishing a few things about these bikes, since you'll have the capability to see the AFR. One would be exactly when the FI goes into its "closed loop" state with the O2 sensor. Another would be exactly what effect and at what rpm ranges the CO setting influences.

Part of the idle instability may be attributable to the advanced, and fixed, valve timing that allows relatively high rpm operation on these engines. That much advance and overlap is good for high rpm power, but would be detrimental at idle speed. But having an optimum F/A ratio at idle will help eliminate lean stumble for sure, as well as making roll on smoother, and on-off transitions less exciting.

It would also be interesting to confirm that the fuel is cut during trailing throttle, as has been surmised from riding these bikes. If that is true then you'll only get the off-on throttle so smooth without reprogramming the fuel cut out of the ECU, or learning to never fully chop the throttle.

Seems like we could learn a lot of other cool things about these bikes with a peek inside what the FI map is actually doing under various conditions.

 

[jammess]

"So, you will be installing the wideband O2 sensor in a new port and leaving the stock O2 sensor in place?"

Nope, will replace narrowband with wideband. That's where the LC-2 comes in as it sends an analog signal from the wideband to the ECU and the ECU will simply richen or lean out the mixture just as it does with an input from a narrowband O2 sensor only the end result won't be 14.7:1 unless the LC-2 is programed for this mixture.

A guy name of Roger04 did a ton of research on the effects of using an LC-2 and posted his findings extensively on www.bmwsporttouring.com , he owns an '04 R1150RT as did I which got sold when I found my FJR. Nope, don't miss the top heavy beast.

 

[Mr_Canoehead]

I have an LC-1 - it is a really cool tool because it has a narrow band output that will fool the ECU into running richer. It is a bit of a science project to get it all set up, though. You need to be good with a soldering tool. For those of us old enough to remember Heathkit, that's what it reminded me of.

Roger used the LC-1 along with the GS911 to monitor what the Bosch ECU was doing on the BMW. I'm not sure how you can see what the Mitsubishi ECU is doing on the FJR.

 

[jammess]

Hi Mr. Canoefead, Yeah, Roger did much with his GS911 with his Motronic 2.4. A GS911 won't tell me much about my bike with the older 2.2 Motronic. With the realtime recording you get with the LC-2 you can easily tell when the ECU goes closed loop plus other interesting things. I'm particularly interested in observing the effects of the CO settings. I think the CO settings only effect the AFR during warmup when the ECU must be in open loop. Probably even after warmup anytime a throttle change causes an open loop condition. I used an LC-1 on my R1150RT but have an LC-2 on the R1100RS, much easier with far less wiring with the LC-2.

This morning for fun I set all CO settings to -10 and went for a ride up the coast which was drop dead beautiful (

) this AM at 61 degrees with moderate ocean swells and only a touch of receding fog bank I noticed a smoother idle with no discernible effect on power output and no tendency to surge. Even warmup at idle seemed smoother. I can't wait to install the LC-2. Never tried a minus setting before.

Roger had obtained a Bosch service manual for the Motronic 2.4 which apparently was extremely useful in his extensive investigation into the workings of the beast. A Motronic 2.2 replacement from Bosch costs over $2K! Much to appreciate about the FJR for sure.

 

[mech 1 twa]

FJR also has a catalyst in exhaust system. Will this effect overall A/F reading? It should change as Catalyst heats up.

O2 is down stream of catalyst. It would be interesting to read both sides of exhaust system for A/F ratio.

 

[jammess]

The O2 is upstream of the cat so won't have an effect.

 

[bigjohnsd]

An A/F ratio of 12.8-12.9 to 1 is close to what is normally accepted as "Rich best Power" and is definitely safe, Lean Best Power is fraught with engine destructive consequences.

https://www.turbobygarrett.com/turbobygarrett/airfuel_ratio_tuning_rich_vs_lean

"So what is meant by a rich or lean AFR?

A lower AFR number contains less air than the 14.7:1 stoichiometric AFR, therefore it is a richer mixture. Conversely, a higher AFR number contains more air and therefore it is a leaner mixture.

For Example:

15.0:1 = Lean

14.7:1 = Stoichiometric

13.0:1 = Rich

Leaner AFR results in higher temperatures as the mixture is combusted. Generally, normally-aspirated spark-ignition (SI) gasoline engines produce maximum power just slightly rich of stoichiometric. However, in practice it is kept between 12:1 and 13:1 in order to keep exhaust gas temperatures in check and to account for variances in fuel quality. This is a realistic full-load AFR on a normally-aspirated engine but can be dangerously lean with a highly-boosted engine. "

 

[dcarver]

I will be following this thread with great interest. KrZy8 has about a 200 rpm hunt at idle which irritates the hell outta me. Of course, she's also way overdue for a valve check, the new balanced FI's haven't gone in yet, the plugs are way old, blah blah blah.

...the techie side of the house is much more interesting...

 

[jammess]

Got the wideband O2 sensor with LC-2 controller installed this afternoon. Will program it up tomorrow morning and see what happens. Will start with an AFR of 13.5:1 and go from there.

 

[Mr_Canoehead]

@jammess do you still have the pair system installed? That will cause some issues with O2 levels when it operates.

 

[jammess]

@jammess do you still have the pair system installed? That will cause some issues with O2 levels when it operates.

Good point. I removed the system early on.

 

[jammess]

OK, I've had a fun and informative morning. The LC-2 does its thing after you ground the black/blue wire which completes the circuit for the O2 sensor simulated input from the LC-2. Doing this doesn't seem to adversely effect the FJR's other sensors that being the crankshaft position sensor, lean angle sensor, intake air temp sensor, coolant temp sensor, cylinder ID sensor, TP sensor, and intake air pressure sensor. These sensors all share the same common connection which is above ground by about 4 ohms. I had to ground this point to satisfy the LC-2 and allow it to be recognized as a valid input by the ECU. ECU threw no diagnostic fault codes.

These things I can say that I'm pretty darn sure of: (1) The FJR does not go closed loop at idle, ever. (2) The O2 sensor does not effect operation at idle. (3) The only time CO settings have an effect is at idle before closed loop. The ECU stays in open loop at idle. For fun I set in a CO of +20 and saw the AFR go from 13.5:! at a CO setting of "0" at idle to 12.4:1. My nose also confirmed this rich running condition. My bike being a 2010 has factory CO settings all zeros which at idle produces an AFR as previously noted of 13.5:1. The CO settings have no effect on AFR above about 1500 RPM or so.

An over the road test showed that an LC-2 setting of 13.5:1 results in smooth powerful operation so I'll leave it there, for now. I see no reason to go richer on the FJR but not so with a 1993 BMW R1100RS.

 

[dcarver]

Does a 2010 have ODBC connection point? Or were you picking up various signals discretely?

Did you notice any change in idle rpm hunt from CO settings from 0 to +20?

 

[jammess]

And now the rest of the story. It doesn't work connecting the black/blue wire (floating ground) to ground because doing so screws up the other sensor inputs to the ECU which causes the bike to not idle well at all in open loop with much sputtering and occasionally dyeing. The LC-2 uses battery negative for common and just no way around that. Long and short is if you can't connect the floating ground to ground the ECU sees an open circuit where the O2 sensor is concerned. IF you ground the black/blue wire all is well above idle in closed loop. I see no way around the ground issue. I even tried a ceramic capacitor between the floating ground and ground and again no joy. Maybe Yamaha uses this floating ground to isolate the ECU from stray electrical noise on the sensor inputs, who knows.

I guess I have a spare LC-2 with wideband O2 sensor for my BMW.

For what little its worth I know when the FJR goes closed loop and I know when the CO adjustments come into play.

 

[dcarver]

Bummer.

In my low-tech way I've been playing around with various CO settings and simply listening to the idle.

I went from CO = 100 on 1,2,3,4 to CO = 25 on 1,2,3,4 and the idle is much better.

No clue where it was from factory, which does not matter as this is engine #2.

I hope Fred can join in to provide true insight not my BS.

 

[jammess]

Hi dcarver, No, I didn't notice much of a change in idle RPM at different CO settings. Well, maybe a slight drop at +20. For the heck of it I checked the idle AFR at different CO settings with a warm engine.

0 = 13.5:1

+7 = 13.0:1

+10 = 12.9:1

+15 = 12.7:1

+20 = 12.4:1

The readings fluctuated a small amount around the above mid points by maybe a tenth plus or minus.

Tomorrow I'm going to try and contact Innovate Motor Sports and see what they say about working with an ECU that uses isolated grounds. I can't believe they have never run into this issue before. Before I decided to try the LC-2 I studied the bikes wiring diagram and I wondered if this above battery negative ground would cause a problem but I spent my $153 anyway. Oh well, now I have a spare for the R1100RS.

 

[Fred W]

Been following along here, but I don't quite understand what the problem is with the grounding. Where exactly is this black/blue wire that needs to be grounded? Is it the wire on the wiring harness O2 connector? Or is it a wire coming from the LC-2?

If you do not ground that wire will the LC2 still maintain a target A/F ratio even though the ECU never goes closed loop? If so, why would you care that it doesn't?

Also interesting that you don't believe that the stock bike never goes closed loop at idle. I'm pretty sure I read in the FSM that that was one of the few times that it does. What was your indication that it doesn't? The varying AFR when changing CO settings is exactly what I would expect if the FI was running closed loop and the setting was varying the target voltage. I thought that was what that setting was doing, not changing the fuel map.

 

[jammess]

In the FSM wiring diagrams it shows a black/blue wire attached to one side of the O2 sensor which acts as a common return for not only the O2 sensor but all the other sensors previously listed as well. This is a floating ground or common and only connects to a pin on the ECU. The earlier Innovate LC-1 had two ground wires one for power (12V) and one for signal ground or common. The LC-2 only has one ground wire that being for power as well as signal ground.

The only time the ECU looks at the signal from the O2 sensor is when it is in closed loop. I think this is common with this technology.

After I grounded the black/blue wire I programed the LC-2 for and AFR of 14.7 and increased the idle speed to around 2.5K revs and watched the real time recording AFR rise to that value. Then I reprogrammed the LC-2 to 14.0 and again increased the idle speed and watched the trace go to that value. At the time I had the CO settings at +15. I expected the bike to go closed loop after warm up at idle just as the BMW does. Would like it if you could find that description of operation in the FSM. I admit my data and conclusions could be flawed because of this floating ground issue causing incompatibility problems using the LC-2.

Today I'm going to disconnect the LC-2 from ECU then put my laptop in the trunk and go for a ride and see what the recording looks like. I think this should force continuous open loop operation with no input from an O2 sensor.