[HarriS]

Just rode with datalogger. Looks interesting or better yet strange. Works very well but the O2 voltage is either close to zero on overrun with closed throttle or around 0.9V when the throttle is open any amount at any rpm or speed. Either the 20102 TuneDyno is running awfully rich or the logger does not show the voltage correctly. Any experiences? Seat of the pants the bike runs just fine and feels strong without dead spots or any hiccups.

 

[dr_gallup]

Are you using a wide band sensor? The stock sensor just shows rich or lean compared to stoichiometric, impossible to tell where you really are.

 

[Devious2xs]

At WOT, the stock O2 sensor has issues.

At cruise, it should give you an idea, but you still need a wideband.

 

[HarriS]

Are you using a wide band sensor? The stock sensor just shows rich or lean compared to stoichiometric, impossible to tell where you really are.

I use the stock sensor. The sensor should show 0-1V depending on mixture. I have used narrowband A/F gauges so I have a clue about what it should read. It just seems that the datalogger reads about 0,9 volts all the time. That means about 12.5-12.7 A/F if the sensor reads correctly. I would like to have maybe 13.5-14 on light throttle highway cruise to cut down the fuel consumption.

 

[Devious2xs]

HarriS,
For you to have a leaner cruise - In TuneEdit, the tune you are using needs to have the AF set at 14.5 or leaner at 10% and less throttle for the stock O2 sensor to work.

Some tunes have had this changed from stock, so they are richer at cruise.
Check it.

 

[HarriS]

I just checked every tune I have and they all have 14.50 at idle only. By the sensor to "work" you mean closed loop, right? The sensor should work and give varying voltage no matter what the A/F target is or is it on closed loop or not. At idle the log shows the voltage jumping up and down like it should. The sensor is not an active component and just generates voltage according to the oxygen content of the exhaust. There must be something not linear in the way the ecu measures the voltage or how the datalogger reads it.

 

[Devious2xs]

Yes.

The ECU is only in closed loop when the AF table has a specific throttle/rpm cell marked 14.5 or leaner (higher).

 

[HarriS]

The main question is, does this mean I have 12.5 A/F throughout the rev range and at all throttle positions? Has anybody else used the datalogger with stock sensor?

 

[Devious2xs]

I used the data logger a few times when I first got the software. But I never got a good file that I felt was accurate at anything other than steady throttle position.

So I mounted up a wideband sensor and LM-1 logger.

 

[dfib]

I used the data logger a few times when I first got the software. But I never got a good file that I felt was accurate at anything other than steady throttle position.

So I mounted up a wideband sensor and LM-1 logger.

Where can I get a wide band sensor, oh great guru of the tune?
Available at autozone or similar, or special order?

 

[Devious2xs]

Are you looking for just the sensor, or a complete kit with logger, plus gauge, etc.?

 

[Guest]

I dunno if you can use a wibeband from a car or not, but a lot of kits usually run around $500-$800 from what I've seen.

 

[Devious2xs]

If you are just looking for a gauge, sensor and wires, you can find one for a couple of hundred bucks. More complete systems that log will cost more. You can wire a wideband UEGO sensor so that it will work with the stock ECU in narrow band, and still provide wideband for the gauge or logger.

Just one source --of many:

http://www.jegs.com/webapp/wcs/stor...ltsPerPage=25&pageIndex=1&searchTerm=wideband

 

[dfib]

Well... Would a wide band provide more accurate data with the tuneboy logger, or an I better off getting a different set up for logging? Does Wayne have tuneboy software that works with the pocket pc? That would be easier to lug around than the lap top. If tune boy isn't the way to go, what options would you suggest?

 

[dfib]

This sensor?

http://www.jegs.com/webapp/wcs/stores/servlet/ProductDisplayPopup?storeId=10001&catalogId=10002&productId=761747&itemId=132150&langId=-1&showValue=1

 

[Devious2xs]

For me an aftermarket LM-1 system from www.innovatemotorsports.com works well for many applications. But I have a gauge from J&S electronics that displays AFR and knock - I just need to get it mounted.

http://wbo2.com/3a1/default.htm is also a very nice system.

I am hoping to put together an article about wideband tuning withing the next few weeks. Right now I am finishing my S3 engine work and the related article.

Too many projects, too little time.




As far as I know, WM at TB has no plan to do a pocket PC version. Several have tried to get him to do it, just not enough time.

 

[dfib]

:shocker: That is entirely too confusing with out beer in the system.

 

[dfib]

Ok, I had some free time so hooked up the lap top and went into tune edit to try logging. Mine is grayed out, as is the option to load the logging patch? :wtf:

Not to be deterred I used the logging program that I found in the tuneboy directory, that worked in a fashion. Looking back at the log I noticed it had a top speed of 121, I know I got on her a bit, but nowhere near 121, 90mph tops, maybe.

Initial idle shows an o2 V of 1.274, once warm idle values ranged from .875 - .27 ? :wtf: Is this what you (DVS) were talking about when you said it was inconsistent?

Thoughts?

 

[Devious2xs]

Yes.

Generally speaking,the way narrow band works in closed loop is it checks voltage, then checks the table. If it needs more fuel it adds it, if it needs less, it reduces the injector pulse width. Then it does it again. Kind of on, off, on, off,... with a slow sampling rate.

Close loop with wideband (5-wire UEGO sensor) is considerably smoother. With a 7-wire unit, the speed is even faster and smoother.

The good doctor of injectors (Dr. Gallup) may be able to explain this better than I can.

 

[dr_gallup]

I don't know any specifics of the Keihin system but the fundamentals are all the same. In general, MC systems have been rather primitive compared to automotive systems because the emissions requirements have been pretty lax. This is changing. All automotive & I assume current generation MC ECU's have a portion of memory that is referred to as "adaptive". With narrow band sensors the ECU only recognizes a rich or lean condition, it does not try to determine how rich or how lean. It constantly makes small corrections back and forth between rich and lean and at steady state conditions you should see the O2 sensor output dither up & down if everything is working correctly.

The ECU "learns" which areas of the closed loop map are rich and which are lean and stores correction values. That way the system does not have to spend nearly as much time relearning each time you start up. There are both long term and short term adaptive memory cells. The number of adaptive cells will be smaller than the static maps with each cell covering a perhaps a 3 by 3 grid. The system will also have limits on how quickly these values can change and how much authority they are allowed. Most systems clamp the adaptive corrections at about 30%. The programmers figure that if you need bigger corrections then something is seriously wrong. Of course, that something may be the supercharger you bolted on.

All of the above only affects closed loop operation. During open loop, the ECU is using map values and pre-programed corrections for temperature, atmospheric pressure, battery voltage, etc. The O2 sensor is ignored during open loop operation.