Question/idea regarding 02 sensor...

[hasenpfeffer]

Just curious if anyone has considered the possibility that O2 sensors may be faulty in a way that still provides (faulty) data to the ECU, making the bike run poorly, but not the same as it being unplugged.

I don't consider this very likely at all, but, it is at least possible. Sensors don't ALWAYS fail completely. I've had a TPS sensor that was only faulty when it got hot, so it was rather intermittent and very difficult to figure out. Perhaps an O2 sensor could be partially faulty in a somewhat similar manner. I've also had a case where a weatherpak pin inside the connector was actually out of spec just enough to cause intermittent contact, but visual inspection did not show any problems. This was also very difficult to figure out. Once the actual weatherpak pin was replaced, no problem since.

Maybe some people just have faulty O2 sensors. Or, also, maybe using an AIT spoofer causes fowling of the O2 sensor, which is causing spoofer problems over time. I haven't heard anyone try to replace their O2 sensor yet, they aren't infallible.

 

[drzcharlie]

I have has the Eruption Mod since 50 miles. When I pulled the o2 sensor it was pristine. Your theory may have validity though.

 

[nev..]

It seems like each Terra has it's own personality with regards to the off idle stalling and other drivability issues. That being said I think mine is probably as bad as any of them. It will stall on just about every takeoff regardless of the ambient temp or the engine warm up. I have given up on the IAT spoofers due to the learning curve of the ECU.

I'm dubious that any ECU learning curve is adjusting your spoofers out. In the spectrum of bikes out there, some have no stalling issues whatsoever, most have some issues, and some have nothing but issues, and from your description it sounds like your bike is way way down the end of the spectrum with nothing but issues, and that's probably an issue that needs to be dealt with, because none of the bandaid solutions which seem to appease the majority of bikes in the middle of the bell curve seem to help you.

 

[glitch_oz]

Just curious if anyone has considered the possibility that O2 sensors may be faulty in a way that still provides (faulty) data to the ECU, making the bike run poorly, but not the same as it being unplugged.

That would make for one hell of a batch of shonky sensors then, and all of them bought and used just by Husky to be used on the TR's?
And all of them displaying slightly different faults? Naw....that doesn't sit right. Haven't checked who made the sensor, but most of them, used in the automotive field, are NTK's
(part of NGK spark plugs, the reason they also use the same 18x1.5mm thread), not really a little Dinky-Toy company nor known for doubtful products.

 

[mag00]

That would make for one hell of a batch of shonky sensors then, and all of them bought and used just by Husky to be used on the TR's?
And all of them displaying slightly different faults? Naw....that doesn't sit right. Haven't checked who made the sensor, but most of them, used in the automotive field, are NTK's
(part of NGK spark plugs, the reason they also use the same 18x1.5mm thread), not really a little Dinky-Toy company nor known for doubtful products.

They are made by NGK. Stranger stuff has happened. Our chains for instance. Custom order to RK. Smaller pin diameter and you cannot use a standard link to put it back together if you bust it apart.

Our ECU doesn't seem to be used on any other models. Custom ordered by Husky/BMW

So a bad batch or bad design of custom ordered parts by the brainchilds at BMW MM Husky Itallia is not a stretch by any means. AFter all, all the o2 spoofers are is a way to fix the o2 signal, and the results do move in the right direction.

Then, adding to the mix would be the factory setting of the throttle plate. Still not enough discussion on that. That is a sensitive setting made at the factory, but not shook down by the techs. Seems that most Moss operators do not go after a slight tweak to make the bike run perfect, but rather rely on what the electro gadgets tell them. A small tweak in the Throttle plate, then moss may be the reason why some of us do not have the issues, especially after we have them tuned.

I see great testing and documentation on spoofers, but have not seen one test on the throttle plate.

I have heard of one or two BMW owners with similar problems have great success by adjusting the throttle body idle set screw.

Add a low quality o2 sensor to the mix and sure, it is plausible.

 

[drzcharlie]

Plausible but IMHO not likely. The ECU does learn. It has been demonstrated over and over again. The O2 sensor works as designed IMO but I'll concede a possibility of some being bad.

Bottom line is that the ECU is the culprit. The ECU is the heart of the FI system. It tells the injectors and throttle body internals how much air and fuel to add. It gets its information from o2 sensor and AIT and applies algorithms to make it happen.

FI and Lambda are seemingly mysterious and daunting. However, Lambda is a simple concept once you wrap your brain around it. Stock lambda is 14.7:1 and shifting that value the ECU's job.

We lack the ability make changes to Lambda because we cannot access the ECU's codes. If we could we'd have the ability to manipulate the o2 sensor, injector pulses and air injection by telling it to tell the ECU what WE want it to hear and see. That's it, nothing more, nothing less other than variations at different rpms.

The "band aids" are our only means of coping right now. Doing things like changing the voltages at the O2 sensor and telling wide band o2 sensors that WE like one Lambda setting better than the stock one thus fooling the ECU's samplings (no, they can't be learned). Break the ECU Hex Codes and we have something. Until then we are doing the best we can. Breaking the Hex Codes is beyond my capability or expertise so I have to defer to those smarter than I am.

On another thread here one hapless soul attempted to get us (in the states) to register with the NTSB to show that we have a common problem with lean stumble and how dangerous it is. If more of us would register instead of bitching whether it'll do any good or not perhaps NTSB could get action on the stumble which would undoubtedly lead to revamping the ECU programming.

In the mean time we struggle forward. I have proven to myself that fueling upwards of 9% makes this bike a beast and a pleasure to ride. Now, how best to do that? LC2, PCV, EJK...the focus is on these devices now. Or maybe the o2 bung extender, it could be that simple.

Regarding, my testings the o2 bung testing begins shortly after I have my coffee.

 

[mag00]

I don't think "lean" is the problem. Combustion can and will happen in lean conditions.

If max power is what you are after, well, carry on. I always loved the flames from the dragster pipes. If you want a dependable 200+ miles per tank,I do not subscribe to the adding fuel as the fix.

 

[mag00]

Coffee done, onward ho.

Looking forward to the test results.

 

[hasenpfeffer]

Plausible but IMHO not likely. The ECU does learn. It has been demonstrated over and over again. The O2 sensor works as designed IMO but I'll concede a possibility of some being bad.

Bottom line is that the ECU is the culprit. The ECU is the heart of the FI system. It tells the injectors and throttle body internals how much air and fuel to add. It gets its information from o2 sensor and AIT and applies algorithms to make it happen.

The ECU also receives input from the TPS and a pressure sensor. It all works together. Messing around on a TBI injection system of a Jeep told me just how much they all work together. If you mess around with 1 sensor, then you pretty much have to readjust others as well. And the throttle plate stop is an extremely important piece of the puzzle from my experience. Messing slightly with that can change things quite drastically.

Note that the pressure sensor is on the under side of the ECU. Before going straight to complicated fixes, just try simple things first. Maybe the under side of the ECU is clogged up with dirt for some reason? Maybe the air pressure under the seat deviates significantly from the airbox for some reason? Just trying to offer other, simpler, ideas. In my experience trouble shooting lots of things, starting with addressing simple things first goes a long way.

Just trying to bring up the point that similar symptoms can easily be caused by different problems. Just because you experience a stumble, doesn't necessarily mean the same root cause. And adding third party equipment often just covers up the root cause, not necessarily fixing it.

It's also somewhat puzzling to me that lots of bikes run quite well, while others don't. Mine happens to run well, maybe not perfect, but certainly no wear near spiderman's description. Well enough even that I have no intention of changing the stock sensors in any way. This to me, generally means the fundamental design works. If a particular bike deviates from this considerably, then usually that means something isn't working according to the design. This doesn't usually mean change the design. It usually means fix the thing that isn't in spec.

That all being said, the TR is a first year design, including the ECU. So the design isn't infallible either, by any stretch. And, it's been suggested before, maybe some have ECU's themselves that aren't working according to spec for some reason.

 

[drzcharlie]

Alright here we go. Kyle is right kinda. The differences between extended bung and no extended bung are negligible but non-the less present. My observation was that my lean stumble came back immediately with the stock Lambda or 14.7:1 (on my bike YRMV) and it was not present during my runs at 13.2:1. Here are the charts:

13.2:1 Test - with no Extended Bung

14.7:1 Test with no extended bung (My ECU didn't like this test, it hunted a long time)

After the first two runs I installed the extended bung and ran the following tests.

13.2:1 with bung extension installed

14.7:1 with bung extension installed.

The difference is repeatable (or nearly so). So it makes a difference but not much of one. Fueling differences did not effect the overall difference seen with the extended bung however, it is such a negligible difference I doubt that it is helping that much. Fueling differences did have an immediate affect on the lean stumble. 14.7:1 re-introduced the stumble in my engine, putting the AFR back to 13.2:1 eliminated the stumble.

My personal opinion is that the extended bung may be of value if a length is found that more greatly affects the shift in AFR. Otherwise, well, you make the call.

I will leave the bike hooked up with extended bung and to my computer if anyone wants to suggest another test of any kind. I will put things back together in the morning.

I hope this helps everyone, I need to go put my bike back together. Sorry for the delays in getting this done.

 

[drzcharlie]

The ECU also receives input from the TPS and a pressure sensor. It all works together. Messing around on a TBI injection system of a Jeep told me just how much they all work together.

I agree, I should have noted that in depth rather than state "throttle body internals" What you said is what I meant to imply. Thanks for pointing this out.

 

[isldtime]

Well, it looks like the bung needs to be 2 feet long not inches.
At least Kyle's thinking was in the right direction and bravo to Charlie for proving the theory.

 

[vodka]

Just curious if anyone has considered the possibility that O2 sensors may be faulty in a way that still provides (faulty) data to the ECU, making the bike run poorly, but not the same as it being unplugged.

 

[vodka]

It would seem that the Husky Terra O2 sensor ,part 7674452 is probably the same as used on BMWs eg 1200R,Xcountry 650 etc..BMW part 1178 7674452. Can't be absolute but both prob sourced from NGK as mentioned before so faulty sensor specific for Terra is unlikely.​

 

[kyle01]

I wonder if removing it from the stream further (I'm thinking 90* bend) would have much of an effect? Honestly, this bike serves it's purpose in my lineup and I enjoy it, but as far as offering much in the way of "performance"- well, that's not really it's forté. It's the turd of my lineup. (a lovable turd that I ride a whole lot more than some of it's stablemates)

My bottom line: I want it to run decently without destroying itself. That's it.

 

[hasenpfeffer]

A possibly far fetched theory about the O2 stand-off...I'm wondering if the stand-off is causing a change in temperature of the sensor, possibly affecting its capability. Extending the bung may be cooling it off a bit.

Also, I never new this before, O2 sensors use the cabling to introduce fresh air into the sensor. So it is very important to be sure the cabling area into the sensor is relatively clean and free from any grease and such.

 

[PaulC]

Thanks Charlie

 

[drzcharlie]

My pleasure. Kyles idea was a great one and it was worth a shot and after all, it did work, kinda.

I think I will stick with the LC2 for the foreseeable future (although, the PCV interests me). The LC2 has two programmable analog outputs and I can run two different fueling arrangements on. What I plan to do is:

1. Play with the 13.2:1 setting by backing it off a couple of steps at a time to get the best balance of power, driveability and fuel economy. I will try 13.4:1 next and so on.
2. Use the second analog output so that I can switch to it for long highway runs. I will likely target settings somewhere in the 13.8:1 - 14.2:1 range.

I consulted with Roger on switching arrangements for this and it will take a single pull, single throw switch and sorting which ECU wire will need to be spliced so that I can switch between the two fueling arrangements. I hate to have to manually switch the settings but as I said above, until the we have a broken hex code, it is what it is, or, at least that's way I am going to pursue it.

 

[drzcharlie]

I don't think "lean" is the problem. Combustion can and will happen in lean conditions.

If max power is what you are after, well, carry on. I always loved the flames from the dragster pipes. If you want a dependable 200+ miles per tank,I do not subscribe to the adding fuel as the fix.

What else are you going to try? PCV? EJK? What are your thoughts on this and how are you going to approach the happy medium?

Here is where I am at with the LC2:

1. LC2 set at 13.2:1 AFR (ridden 500+ miles) - 55.01-56.0 Miles per gallon
2. At 55.01 mpg that equates to 1.818 gallons per 100 miles or 6.681 liters per 160.9 K
3. That converts to 205 miles per tank of fuel (3.74 Gal. US of 95RON)

We both know real world riding will affect that number downwards. So, I would guess (and it's just a guess) that realistically I will probably get between 180-190 +/- miles per tank under hard riding mixed on/off road use. This of course will depend on the terrain, my wrist angle and environmental conditions. That's pretty close to your "dependable 200 miles per tank."

I think I can live with that part.

As for loading the piston with carbon? We share those concerns. Well, after nearly 550 miles since changing to 13.2:1 AFR I see no evidence of soot on plugs, o2 sensor or exhaust outlets. In fact the plugs look great as did the o2 sensor. My conclusion? It's not worth worrying about.

I'd be interested in what your thoughts are on this. How do you intend to bring your bike to achieve fueling nirvana? You've been pondering this problem longer than me and you have had some good insights. Please share your thoughts with us on this.

 

[drzcharlie]

My next project will be to build a switch circuit and then refine my fuel settings for (lack of better terms) "power" and for "economy". I will hunt parts this morning and get the wiring sorted while I am up on the lift. Unless my wife has other plans for me...like "honey do's."