Tuning Anomalies using TuneECU and wideband

[Claviger]

After sorting through some issues with certain tunes, i have compiled my own Hybrid tune.

First things first, it is a composite tune using various bits from various tunes to achieve what I am after, that being a very smooth ride at low throttle openings, no herky-jerky crap, but a butt load of power from 50% throttle and up. That being done with the O2 sensor disabled and unplugged. I don't like ECUs ****ing up my hard work doing maps, and I live in a place that the weather is the same pretty much everyday.... Mission already accomplished Cruising around doing the same speeds as cars AFRs stay in the high 14s, sometimes 15.1:1 - 15.3:1, perfectly fine for low load cruising. Even accelerating like a "normal" person, about the same rate as the family mini-van, it stays in the mid-high 14s, perfect. Went hill climbing a bit to test out the higher load areas in the same RPMs (also drag back brake to increase load), and it dips down into the mid 13s, again, perfect behavior.

Testing out the reactions to snapping open the throttle from 2k-3k areas, and it does a split second (.08 of a second to be exact) lean (normal behavior on snap WOT), then drops down into the 12:1 range. This is where things get kind of weird...

3rd Gear WOT pulls - when it drops to 12:1 it dips further into the 11:8:1 area and stays there until redline.

4/5 Gear WOT pulls - It doesn't drop to 12:1, it drops to like 13.2:1, and floats in the mid-high 13:1 area until at least 5k (cant run out 4th/5th here on Oahu).

What I am not understanding is why 3rd gear is dipping so low and staying there, yet 4th and 5th are spot on. It seems like the ECU has a "hidden" enrichment factor for the lower gears that isn't documented anywhere. Bike still rips 1-3 gears just fine, actually got a CBR to run from 35 mph on the freeway while logging today.....lols just fing lols.

Suggestions?
I have attached a link to where you can get my tune to this post. DO NOT INSTALL THIS TUNE ON A PRE-2013 ROCKET, it is based on tune 20773 which will only work on CAN enabled rockets using the new ECU configuration.

http://www.r3owners.net/threads/tune-for-a-stock-r3-developed-on-a-roadster.20552/

 

[Claviger]

I think I have discovered the culprit!

I live at +40ft above sea level, 1013 +/- 10 hpa pressure, so pretty low compared to vast majority of people. I have a sneaking suspicion that the L tables are to blame and my F tables are pretty **** close as is.

My theory, which I am going to test shortly, is that an accelerating engine in 1,2,3 gear will draw more vacuum as it rapidly climbs the RPM scale wanting increasingly larger amounts of air. As such, I theorize that it will put the fueling in the 100% column on the F table, and further to the right on the L table, more in the 957 column than the 930. If that is the case, the extra fuel will be from the very rich spike in the 957 area.

If I am right, then 5th gear would sit closer to the 930 area where the L table is significantly more lean than the 957 area. My data logs reflect this, however, the ECU only updates the MAP signal for logging every 1 second, so the rapid transients are lost

I am going to add 5% to the 901 and 930 areas, and take 6% from the 957 and 980 areas. If I am correct this will slightly lean the lower gears and make 5th gear more rich.

Answers will come tonight, but, if this is the case, it makes me wonder what elevation Triumph originally tuned the R3s for, because clearly it wasn't sea level

 

[ROCKETHEAD]

Keep it coming...

 

[Claviger]

Confirmed on a short ride just now. Removing 6% from the 957 column increased the AFR from ~11.5 to ~12.5 at WOT in 2nd gear. Awesome. One accidental but fantastic "feature" of the tune I have built, is that when gently accelerating and shifting the AFR drops from the cruise AFR in the 14s down to the mid 12s for about 1 second. The reason this is awesome, it smooths out shifting significantly. No bull****, I could not feel the 1-2 shift a couple of times on the way to work today, smooth like I have never felt on a motorcycle or even in most cars. I certainly didn't suddenly become gods gift to clutch work, so the extra fuel is the only explanation.

Two things need adjusting, and it will take time for me to find the right adjustments.

1. Currently when going from cruise in any gear @ ~14.8:1 AFR to snap 100% Throttle, it dips to ~9.9 for a single update on the screen of the LM-2 so 1/12th of a second, then slowly climbs to mid 12s if held open.
2. 1000-1400 RPM at very low throttle openings (1-4%) are in the mid 15 AFRs, so slightly too lean. I know most dont ride there often, but in traffic I end up there quite a bit so I need to add some fuel to those 4 cells to get it back in the mid 14s.

Otherwise I am seriously happy with the tune I am using thus far. Once I get the AFRs where I want them at WOT and where the bike seems to pull the hardest, I will do a final validation on the dyno to make sure it is making what it is supposed too. All indications so far, even a little rich at ~12.5:1 at WOT, are that its making at least what it did with the R3R_Slip canned tune installed, I suspect similar or a tad more but a much smoother Torque curve.

 

[Claviger]

This tune I am very happy with. I have made the changes I mentioned above and have attached the file for anyone who might want to look it over. Ripped a 4th gear pull to **** near redline on the way home today (tee-hee I couldn't resist) and it floated from 12.4-12.7 after the initial enrichment period which went to 10.2 for a split second then climbed to mid 12s within a 1/2 second. I am confident it makes more power than the 20773 standard tune does, it pulled like woah well over 100 mph.

As an aside, I found that the L tables which I had thought would be the reason I was dipping into the 12s while shifting and providing that super smooth shift were identical to the stock 20773 tables. The only thing different in that area is the F tables, and not by much, so there you have it.

Definitive proof that F tables are used in low-load, low-throttle, low-rpm situations, even below the F-L switch settings. Conversely, L tables are used in High-Throttle, High-RPM, High-Load situations. They are never exclusive, and are always contributing to fueling calculations in a significant way. The fueling is never truly alpha-N or Speed density, but a hybrid of both from idle - redline and 1%-100% throttle.

Another finding I have made is to confirm what Power-Trip was saying all along, the AFR table is unused during fueling over about 10% throttle (some people say 6%, not sure exactly where the change is). It is also completely unused if you uncheck the O2 sensor box in TuneECU. I made a pretty little AFR table for myself to use as a quick reference sheet in the middle of this tuning process and it made exactly zero difference in fueling. So AFR table = useless except for idle/cruise targeting 14.7 AFR, other than that, it is wasted space.

Tune description:
Stock intake/Stock exhaust Roadster.
Raised Speed limit to 185.99MPH.
Fixed Speedo error (verified with GPS datalogging).
Lowered fan turn on temp to 99c.
Turned off crappy narrow band o2 sensor.
Made a pretty (but useless) AFR Table
Street tuned by data logging various conditions, loads, speeds etc ranging from cruising at 20 mph 5 feet from the ocean to climbing steep grades at 80 mph at 1000ft elevation and some full on drag launches and freeway high speed runs.
Smoothed the ignition to remove very odd maping spikes.
Smoothed the F tables to more closely match the needs of the bike.
Reformed the 3-20 percent area of F tables to match the change in the F-L table switch.
Changed the F-L switching point to values that make sense.
Secondaries fully open (The plates are removed on my Rocket).
Adjusted the L table to work better at my very near sea level elevation.
Added a healthy dose of fuel to the F tables to put AFRs where they should be instead of crazy lean like the stock tune(still less fuel than many canned F tables i have seen).

 

[warp9.9]

I like your write up and research and wish I had time to respond with some questions right now but work is calling me and I have to go in. I will say is so far your writing points to the Lamba table as being a table that adjust the fueling in the F tables depending on the voltage reading from the map sensor. ( much like the automobiles did when running boost) Myself I have the L tables set at zero do to boost pressure exceeding the 1 bar pressure rating of the map sensor. This made controlling the fueling easier for me. I do have the new tune boy software which has the ability to set and run different map sensors and a innovate wide band system. I just have not tried it yet because what I have has been working. Still I will give more thought to your testing and write up so I can ask more questions and learn.

 

[Claviger]

Did a longer ride today on revision 5. The 1050-1400 area is still a touch lean, but everything else seems to be working perfectly. Once I tackle the slightly lean spot I will consider this a finished tune.

Got some normal speed freeway time and it looks like mileage is up from the other tunes I haved used expect I'll get around 30-33 mpg, which is a marked improvement from the 24-25 average I have been getting since it was new. These are about 70/30 splits of city/freeway.

Very low load cruise is 15.3:1 AFR bit crack the throttle to gently accelerate and it dips to mid 14s. Open it more for a moderate acceleration and it sits in the high 13s then when at 100% it drops and stay between 12.8 and 13.2.

Very happy with it. I have posted what I am going to put on the dyno and validate its power output. Cruise is exactly how I want it, was getting close to 40 mpg on the way home today, so freeway cruise seems very good. I'll also transpose it to an older base tune for pre-13 bikes once I get it dynod.

 

[BlacknCream]

Do you still have your cat?

 

[Claviger]

The cats on my bike are inside the muffler, so yes, and the small rectangular resonator next to the ABS controller. Cats make very little difference these days in overall power.

 

[Neville Lush]

"The cats on my bike are inside the muffler, so yes, and the small rectangular resonator next to the ABS controller. Cats make very little difference these days in overall power. " Mmm, I don't know about that. Are you sampling the AF gas behind he cat or in front??