Using Long Term Adaptation info to Tweak Lambda (02 sensor) fuel map

[bbcmat]

TE510 - 2009:

I have Ibeat v2 – and wanted to take another crack at using the 02 sensor and closed loop fuel injection in my 2009 TE510.


I set up the TPS with the wide open throttle reading 100.2% - this required an Idle TPS reading of 1011mV – then set the idle air thumbscrew at 2.75 turns out. This is a generally “good” baseline for Powered-Up mode. Bike surged/bucked between shifts and during light roll ons. Did DIY TPS tweak and 1 ½ turns on the idle air screw – surging and bucking 95% reduced, idle approx 1850 rpm.

I purchased 10 gallons of fuel and set it aside for testing (Sunoco 94). Rode approx 200 kms of mixed street/sideroad and hooked up to Ibeat. 02 Long Term Adaptation table read:

100% 100% 100%
111.7% 109.8% 100%
114.1% 118% 100% with idle 115.6%

Using a paint pen and the throttle position readout, I marked 20% / 40% / 70% throttle positions on the handlebar control. This would give me reference points for observations and testing points while riding. I don’t think I spend enough time at 70% throttle for the ECU to populate the matrix – and thus 100% values at max throttle max rpm are not valid. Turning the idle airscrew out lowered the idle % value but the surging/bucking returned. Reset adaptation matrix.

I changed FB1 / FB2 / FB3 values from 100/100/100 to 110/100/100.

Idle compensation value dropped to about 100%. Went for mixed street ride 100+kms. 20% throttle much better, 40% seemed weak, 70% nothing special. Biggest change was in the transition from cruise (slow throttle movements) to power (large quick throttle movements). The transition was much smoother and rolling on the throttle did not go from wheeze to “hang on” as it did before. I turned in the idle screw 1/8th, which gave smooth off idle transition in the city but created a bit of heaviness (rich) on the side roads. (??)

Long Term Adaptation Table:

100% 100% 100%
103.9% 98% 100%
103.9% 100% 100% with idle 100%

Reset table.

Changed FB1/FB2/FB3 to 112/107/105

Went for mixed street ride (same loop) – 40% much better, excellent 70%, bottom end 20% needed more and more airscrew as the ride went along – idle now at 2050 rpm. Actual riding temperature about +6degrees Celsius. Throttle transitions getting very smooth – drifting out of slower turns not such a cheek clenching experience.


100% 100% 100%
100% 96.1% 100%
102% 96.1% 100% idle 102%

Reset values.

Changed FB values to 110/105/105 and airscrew to 1/5/8 turns out – then tweaked to 1/3/4 turn out to have idle compensation to 100%.

When for mixed ride – everything about the same except for idle air screw needing adjustment during ride – more as night continued – then back in city stoplight to stoplight.

100% 100% 100%
100% 100% 100%
103.9% 102% 100% idle 102%

Reset adaptation.

Changed FB to 112/105/105 and covered ½ of both radiators to reduce cooling at higher speeds. Set idle air screw at 1 5/8 turn out for 100% idle compensation at 95 degree Celsius water temperature.

Went for mixed ride. Excellent on all accounts. Airscrew inconsistencies stopped with rads covering up rads. Fuel mileage less than buck n surge factory setup but still much better than tuned “powered up” settings (my bike 92/95/98 idle air bypass 2 ½ turns out with DIY TPS tweak).

This is the best the bike has run in street riding (powered up or not) as far as best compromise between smoothness / throttle response and overall power. All my testing was done at night with variable amounts of fog and wetness on the road plus temps in the single digits (Celcius). Good economy during cruise – good power when whipped.

This weekend I plan to change the gearing back to either 13/50 or 14/50, take the rad covers off and slop around on some tight single (02 sensor in). I will be surprised if it’s best vs Powered-up – but worth a try anyways.

The long term adaptation table is an excellent tool to help set FB1/2/3 for improved ridability and performance with the 02 sensor is in place.

MAT

 

[glangston]

If tuning is creating better performance and/or mileage we need an acceleration test and a mileage test. Just a thought.

 

[Coffee]

TE510 - 2009:

The long term adaptation table is an excellent tool to help set FB1/2/3 for improved ridability and performance with the 02 sensor is in place.

In summary, are you adjusting the FB123 parameters for minimum long term adaptation value (valuese) changes?


It sounds like you are getting the bike adjusted for your liking

 

[bbcmat]

Yes - adjusting FB1/2/3 values to have the adaptation value of 100% (but not less than 100%) seems to correspond to the best (smoothest) transition from using active 02 sensor value and the base map - and the best power while on the base map.

Note that the only way I could get rid of off-idle surging/bucking was to do the DIY TPS tweak and run the idle air bypass 1 1/2 to 1 3/4 out. No FB1 value would get rid of it.

MAT

 

[Coffee]

Sounds good!

It is my understanding that other experiments are now being performed on this topic, your work will speed up those experiments

 

[bbcmat]

I put this setup through the paces in the bush. The same smoothness that keeps the suspension from jacking while cranked over on pavement is too "flywheel" on the single track. The exhaust header is ripping hot as well - not a rider problem but I'm glad the leaves are still wet.

Fueling was good but the ignition curve just isn't aggressive enough (IMO) - no snap - just pull.

On the ride home (pavement) the 02 sensor/map/ignition curve was an excellent fit. The supermoto version of this bike must be an absolute treat.

I burned about 20% less fuel than I would have expected with the power-up plug installed.

MAT

 

[Ex HVUK]

I started some investigation into the effects of the long term adaption table values with a 2010 TE510 but before I could reach a conclusion I gained some new information from the factory which seems to negate any further testing.

The new and somewhat key piece of information was that the factory assure me the ECU does not refer to the values in the long term adaption table once in race mode. So irrespective of the values stored in the table, once you remove the lambda, the fueling is entirely dependant on the base map and the CO/FB set up.

I was also informed that the base map is optimised for running in race mode.

Dave

 

[bbcmat]

Excellent factory feedback - my experiences in both lambda threads support this.

The long term adaptation table is an information source - not an active component.

MAT

 

[Coffee]

I started some investigation into the effects of the long term adaption table values with a 2010 TE510 but before I could reach a conclusion I gained some new information from the factory which seems to negate any further testing.

The new and somewhat key piece of information was that the factory assure me the ECU does not refer to the values in the long term adaption table once in race mode. So irrespective of the values stored in the table, once you remove the lambda, the fueling is entirely dependant on the base map and the CO/FB set up.

I was also informed that the base map is optimised for running in race mode.

Dave

Great info!

For the record, on the 2008 TE250 that I tested the FB1 did nothing in open loop but did affect closed loop - FB1,2 did nothing at all.
http://www.cafehusky.com/forums/showthread.php?t=4815

Again, that is for a 2008 TE250, I've not done any closed loop experiments on the 2009 bikes because there was no OEM O2 sensor available so the FB's could be adjusted, but in open loop the CO values make a huge difference:
http://www.cafehusky.com/forums/showpost.php?p=46277&postcount=83

 

[BlkLT1Z28]

This option isnt available for my bike......so my iBeat tells me. Seems the 08 bikes cant use the Long term adaptation?

 

[Coffee]

This option isnt available for my bike......so my iBeat tells me. Seems the 08 bikes cant use the Long term adaptation?

Nope. Not for 2008.

 

[seymore]

This option isnt available for my bike......so my iBeat tells me. Seems the 08 bikes cant use the Long term adaptation?

What bike do you have?
Do you still have the O2 sensor installed?
How well does your bike run?

 

[aethelred]

Can somebody please post the manual pages for the IBEAT software or create a WIKI type page?

Tyhe manual in pdf format would be a good read.

 

[seymore]

Can somebody please post the manual pages for the IBEAT software or create a WIKI type page?

Tyhe manual in pdf format would be a good read.

The manual is in PDF format and is copyrighted.

 

[glangston]

Can somebody please post the manual pages for the IBEAT software or create a WIKI type page?

Tyhe manual in pdf format would be a good read.

The Workshop Manual in PDF (available for $10 on CD from the dealer) has a lot of instruction on using the iBeat.. Section S

 

[Coffee]

The Workshop Manual in PDF (available for $10 on CD from the dealer) has a lot of instruction on using the iBeat.. Section S

Yes they have lots of instructions but refer people to the ibeat manual, which refers back to the workshop manual.

That is one of the key reasons I went to vegas the 2nd time this year, to figure out how the ibeat settings correlate to the AFR, and by how much. In retrospect it is all pretty simple, but 6 months ago we knew far less than we do now.

This is probably the most complete summary of what is going on, and it could certainly be better written, post #1.
http://www.cafehusky.com/forums/showthread.php?t=4815

 

[bbcmat]

I re-visited the 02 sensor setup again after 100+ hours and 5000+kms on my 510. I gave up last year because the lack of snap and excessive header heat. I also noticed an occasional “losing it’s mind” after engine braking down a long hill or a series of light accelerations. Since then I have been “powered up”

This year I observed mud debris about a foot down inside my muffler after getting into some slop without the spark arrester in. If there is this much reversion, then it is possible under engine braking and light throttle movements there is a mixing of outside air and exhaust gasses at the 02 sensor site. This might explain the variability of the system.

I put the catalytic converter insert back in the mid-pipe and 02 sensor back in. I was surprised at the mellower tone in the exhaust (I’ve never run the insert – I was powered up from new) and the even idle rpm. Completed a stock TPS setup and tried the last FB1/2/3 values still the ECU.

Bike was weak off idle and at 70%+ throttle. Read the 02 long term table and tweaked FB values – then adjusted idle air bleed for best off-idle smoothness. Did this cycle a few times and ended up at:

FB1 = 114%
FB2 = 113%
FB3 = 114%
Tps = 1016mV / 100.2% WFO
Idle air bleed = 2 ¾ turns
Fuel = Shell 91 octane 0% alcohol summer blend
Spark Arrester = out

Idle ended up at 2000 rpm idle with good power and smoothness till 70% throttle and 7000+rpm – a little wheezy from then on. Bike ran like a heavy 350cc with moderate compression. Fuel economy dropped to 22km/L running 15/45 gearing and included some ridicules high speed runs to test for detonation. The bike would log an occasional “high value 02 sensor” fail in the ECU – the cure was to turn out the idle air bleed and lose some throttle response. At this point I should have tried putting the spark arrestor in and putting the IAB back to 2 3/4 – but the bike was already too weak in the top end to keep my interest.

The major downside to this setup was the high heat in the catalytic converter area (right beside the airbox rubber boot) and the fact that the bike didn’t pull hard enough to justify the increase in fuel economy.

Lowering the FB values make the transition from cruise (02 sensor feedback setting mixture) to power (FB map setting mixture) more abrupt – not great when cranked over on knobbies / pavement / night dew. I liked the subdued exhaust note but not the heat in the system.

I yanked it all out and went back to my “economy” powered up tune. As I noticed before – the bike ran so much smoother after using the 02 sensor that I can’t help but believe there are values being updated somewhere in the ECU that are changed when the 02 sensor is run and then removed – and these values are applied to the powered-up fuel mapping. The only other thing I can think of is that the consistently lean fuel mixture burns any crud off the exhaust valves and the bike runs better after.

If they ever start emission testing motorcycles in Ontario I’ll put the “cat” and 02 sensor back in for testing – maybe once a year or so to remind myself how good it runs without it.


MAT

 

[seymore]

Mat,

I think you need to get a PCV with auto-tune and the LCD-200 for data logging and on the fly adjustments.

A guy like you deserves to have 400+ points of adjustment.