G oog le BadWeB | Login/out | Topics | Search | Custodians | Register | Edit Profile


Buell Motorcycle Forum » XBoard » Buell XBoard Archives » Archive through September 30, 2008 » NB 02 Sensor Enrichmane Testing « Previous Next »

Author Message
Top of pagePrevious messageNext messageBottom of page Link to this message

Mmcn49
Posted on Friday, September 26, 2008 - 10:14 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Brought an O-Scope home to look at the o2 sensor and spent the day testing, (my company is on a 4-day schedule).

The o2 sensor runs at 6-8 MHz shifting rapidly from min to max voltage with a jittery waveform. Plugged in all kinds of settings, (several dozen) but will only give a brief overview here.

Initially did a few tests with my MLV maps, but then loaded in the stock maps. Loaded several o2 voltage settings from stock .41-.49-.56 V, up to .67-.75-82 V.

At that setting EGO correction shot up to 200%. The bike started running lousy around 135%. With the stock CL map set at 14.7%, I was attempting to get a steady, (103-14.2:1) to (107%-13.2:1) CL EGO reading.

Brought the o2 voltage back to stock and connected a 100K pot, (Variable Resistor) between the Violet wire with Grey tracer at pin 4 on the Grey ECM connector, (o2 sensor wire) and the battery’s negative terminal. The pot’s range was measured to be 5 Ohm to 99.5K Ohm through 10 turns.

Steadily brought the resistance down to somewhere between 1500-2500 Ohm. Before installing the pot, EGO correction varied between 86 to 92%. As resistance approached 2500 Ohm, EGO slowly increased to about 99 to 102.

If resistance was dropped just a little more EGO would shoot up to 200. Just before the point where EGO shot up, if you turned the throttle just a little, the ECM went into OL and CL ego would reset to 100%.

Continued making o2 sensor voltage changes and ended up with 47-49-63 V and the pot set at about 2500 Ohms. At this setting CL EGO would vary between approximately 99 to 103, but most importantly the region would shift from CL to OL when the throttle was twisted and stay there.

Clicked on the Other Maps Tab. Changed Open Loop Enrich. Delay to 0. Changed Default Open Loop Corr. To 107% and 113%. Went for three test spins and determined 107% is the best setting.

In conclusion when resistance is lowered there is a very fine line between keeping CL EGO where it forces the ECM to go into Open Loop and where EGO correction becomes unstable and goes through the roof.

The weather is supposed to be nice tomorrow and plan on going for a nice long ride to see how the Buell performs. I have high confidence that all will go well. Perhaps someone reading this can come up with additional circuitry to make the EGO correction percentage more stable.
Top of pagePrevious messageNext messageBottom of page Link to this message

Jos51700
Posted on Saturday, September 27, 2008 - 09:22 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Why can't you just ride it, and let it do what it's supposed to do?

Seriously, with ECMspy now on everyone's harddrive, is there a point to this?
Top of pagePrevious messageNext messageBottom of page Link to this message

Mmcn49
Posted on Saturday, September 27, 2008 - 10:12 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Because it doesn't do what it's suppose to do as good as it should!

ECMSpy identifies & solves many hidden issues that HD/Buell would rather us go to dealerships for.
Top of pagePrevious messageNext messageBottom of page Link to this message

Redbuelljunkie
Posted on Saturday, September 27, 2008 - 11:04 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Do you recall what prompted the HAL 9000 to have an epic failure. DDFI-3 has the same problem- it suffers from a "Hofstadter-Moebius Loop" contradiction. It is programmed to be powerful, efficient, and EPA legal all at the same time. This is impossible.

Adjusting the mapping can dial out some of the contrary information, allowing more specific output- usually performance oriented. Buells must leave the factory slightly "faulty" in order to please the EPA... it's up to the purchaser whether or not to do something about it.

Most of us can't accept riding something that doesn't operate as it should- especially if it can be fixed.
Top of pagePrevious messageNext messageBottom of page Link to this message

Speedfreaks101
Posted on Sunday, September 28, 2008 - 06:31 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Thats pretty damn funny Redbuell.
Top of pagePrevious messageNext messageBottom of page Link to this message

Jos51700
Posted on Monday, September 29, 2008 - 08:50 am:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Well, from the above posts, I gonna assume (and possibly regret the assumption) that you're playing with an '08 bike?

I just dont see why you're playing with altering sensor values and all that, when you can change the maps, and still allow the bike to function like it should. Really, if you want to play with sensor values to alter the way runs, plug in a power commander or TFI, and it up right (We all know the reputation they have).

Powertrain engineers have set the bike up for the constraints they have to work within, and believe me, they hate those constraints as much as you or I. That's why race teams exist.

However, altering sensor values bypasses alot of the work they have done that you can still use to your advantage. Like I said, what you're doing has already been done, and capitalized on (Power Commander).

I'm all for modifying for more efficiency and power, but your methodology is lost on me. (Unless you've an '08 bike, for which ECMspy is no comprende)
Top of pagePrevious messageNext messageBottom of page Link to this message

Mmcn49
Posted on Monday, September 29, 2008 - 03:35 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

I just dont see why you're playing with altering sensor values and all that, when you can change the maps, and still allow the bike to function like it should.

You can datalog w/ECMSpy and make map changes with MLV. You can spend hundreds on a professional dyno tune, but no matter what map changes are made the NB o2 sensor will apply corrections to the changes.

Looking at the EGO correction window in ECMSpy you can see that the bike runs great even with stock maps when ego corr is 100-107%. It also runs great if the Closed Loop map area can be tricked into going into Open Loop.

Normally when the ECM goes into cl, the ego corr drops to somewhere between 86-92% and the bike runs rough.

Nightrider sells devices that when connected between a Harley two wire o2 sensor and the wiring harness will change the AF ratio from 14.7:1 to 14:2:1 or 13.2:1.

It does so by altering the o2 sensors signal and tricking the ECM into thinking its running a lot leaner than 14.7:1.

You may disagree but the only purpose for an NB o2 sensor is for keeping cl, (where we do much of our riding) at 14.7:1. Yes it can adjust for atmospheric changes, but in the real world these changes are miniscule. It can also adjust for altitude changes, but I went to 5000’ with it off and the bike ran just fine.

Nightrider also has a schematic on their web site, (free to anyone) for making an o2 sensor enrichment device.

application/octet-streamNB o2 Enrichment Device
\Floscandata2\Users\Ed\NB o2 sensor.doc (27.1 k)


John as a Buell Tech and HD Master Tech would you please answer the following for me:

Does the two wire HD o2 sensor have the same threads and tip depth as the Buell's single wire sensor?

Buell sells three types of ECM's, stock, race and a race ECM that is only sold to racers and cannot be purchased by average riders.

Has the Closed Loop area in the Professional Race ECM been eliminated?

I've been fairly sucessful in making the bike run pretty good with MLV, but when ego corr. kicks in, it doesn't run as good as it could. The more I correct and enrich cl cells, the lower ego corr. gets.

My initial testing on the o2 sensor proved that ego corr can be raised and that the ECM can be forced to run in open loop across the entire fuel map.

My goal is to be able to change cl from 14.7:1 to 14.2:1. 14.2 will have only a tiny effect on emissions, will affect mileage slightly, but will make the bike run much smoother. The o2 sensor will still do its thing, it just won't be able to go to 14.7


The one remaining problem to solve} is stability. If it will fit, an HD 02 sensor may be the answer.
Top of pagePrevious messageNext messageBottom of page Link to this message

Gemini
Posted on Monday, September 29, 2008 - 05:25 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

the narrow band o2 sensor keeps the ratio at 14: ] because that is the switch point from high voltage to low voltage. you can work around this in two ways if you choose to. use ecmspy and adjust the o2 sensor voltage switch point voltage or go the pricey route and get a wideband o2 sensor and change the voltage to a/f ratio table. so that .50 volts is 14.2. this way you can keep the bike in closed loop(which i prefer to do).
Top of pagePrevious messageNext messageBottom of page Link to this message

Id073897
Posted on Monday, September 29, 2008 - 11:44 pm:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Normally when the ECM goes into cl, the ego corr drops to somewhere between 86-92% and the bike runs rough.

Generally this is wrong. If this would be true, no one would ride with an AFV above 100.

Regards,
Gunter
Top of pagePrevious messageNext messageBottom of page Link to this message

Jos51700
Posted on Tuesday, September 30, 2008 - 09:21 am:   Edit Post Delete Post View Post/Check IP Print Post    Move Post (Custodian/Admin Only)

Most O2 sensors have the same fitting threads and tip depth. The extra wire on the HD is simply a tip heater, for faster closed-loop response.

Looking at the newest Buell accessory book, it says XBRR and XB models use the same ECM. The XBRR had many wires and plugs that weren't used. Among them, were Map sensor wiring (For a true kickass EFI setup) and O2 inputs. However, the tune-able ECM was totally different.

I see one VERY MAJOR flaw in your thinking here. It starts at
"You may disagree but the only purpose for an NB o2 sensor is for keeping cl, (where we do much of our riding) at 14.7:1. Yes it can adjust for atmospheric changes, but in the real world these changes are miniscule. It can also adjust for altitude changes, but I went to 5000’ with it off and the bike ran just fine. "

The O2 sensor does not do anything but measure oxygen. The rest is in the ECM. (altitude correction is simply an air-density exercise, the bike has no clue what altitude it's at, unlike the HD system with MAP sensors).
The bike references the O2 sensor to maintain the air-fuel ratio, but that ratio in itself is governed by Volumetric Efficiency tables set into the ECM. Those VE tables tell the bike, at X RPM, at Y temp, there should be Z amount of air coming in and that requires Q amount of fuel.
I honestly can't remember if ECMspy allows altering the VE tables, but if it does, there's your answer.

Now, ANY time the bike is in closed-loop, it's running about 14.7-ish (It actually does change based on some settings). At closed loop riding pace, this is GOOD. You don't want more fuel, it's a waste. When you whack that throttle, the bike leaves CL, and goes OL, but at this point, it's choosing the OL maps based on the AFV it developed while riding in CL.

So, you ride up your mountain, and decide to race the last mile to the top.
If you hadn't been in closed loop for the first 2 miles up the mountain, the full-throttle performance will be crappy. Playing with O2 sensor feedback is simply bypassing the most accurate aspect of your FI!

Picture the guys that trailer (Yeah, I know, they're obviously not Badwebber's) their FI Buells up Pike's Peak. The AFV didn't change on the way up, and so it's way off when they get up there. Since the bike isn't running right, they do what's natural, right? Gun the piss out of it, to "clean it out". But it never gets better, because at high TP settings, the bike doesn't go CL, to adjust the AFV!

Closed-Loop is a Good Thing. It only occurs when you're cruising. So, you don't need max-power fueling. In fact, have you ever heard the phrase "lean makes it mean"? Leanness gives more throttle response. Set your bike rich and try the roll-on performance. It sucks!

Now, The Buell EFI does need a little work, especially if your bike is modded. They surge at cruise, and I would like to see them a tiny tiny touch richer, but don't fight the system, work with it! The computer makes more calculations in a second than you can in an hour, so put that to work for you! Besides, if you're a cheap heeb like me, you'll realize, you paid for that fancy technology!

And yes, those minuscule corrections are VERY important. And they DO add up. If you want constant fueling, regardless of altitude, temperature, humidity, etc., then bolt on a carburetor. That's EXACTLY what they DO!!!!!!!! (And we all know how much better an FI bike performs over a carb!)
« Previous Next »

Add Your Message Here
Post:
Bold text Italics Underline Create a hyperlink Insert a clipart image

Username: Posting Information:
This is a private posting area. Only registered users and custodians may post messages here.
Password:
Options: Post as "Anonymous" (Valid reason required. Abusers will be exposed. If unsure, ask.)
Enable HTML code in message
Automatically activate URLs in message
Action:

Topics | Last Day | Tree View | Search | User List | Help/Instructions | Rules | Program Credits Administration