Any last ideas before I dive into the world of custom tunes?

[yugo42]

If I remember correctly, I think ECMDroid has entries for max and min AFR that you can just set to 100. That way the system can't adjust based on the closed loop even though it is still reading the o2.

 

[lowkey]

Here is race vs stock EEPROM's enrichment tables(BUEGB vs BUEIB), I have no way to verify if the data was changed or not as I got them from the internet and not an ECM. The ECMID checks out though... I opened a stock XB9 EEPROM and all values for enrichment are the same except WOT is 105% instead of 110% on the XB12 EEPROM's.

 

[lowkey]

If I remember correctly, I think ECMDroid has entries for max and min AFR that you can just set to 100. That way the system can't adjust based on the closed loop even though it is still reading the o2.

You are correct but you want the O2 signal deactivated from the ECM (making adjustments off the signal) for logging. The voltage for NB is 0-1 and WB is 0-5 also the voltage is reversed for AFR rich/lean. It's two separate things that need to be done, lock AFV at 100% to not skew the fuel tables and disable the O2 sensor so the bike runs in open loop off (the entire map is now open loop) of the fuel map cell data only.

 

[cossack84]

Hook up the sensor and pull it out of the header, what AFR and voltage reading do you get? If it reads max lean AFR and 0 volts you will need to track down the data if it reads max AFR but has a voltage percentage I think you could take that as the flat line point in the linear line, do the same in a propane rich cup with the sensor in it to find the voltage point on the other end of the AFR rich spectrum. I would contact for the chart to double check the data works exactly right though to be sure.

My setup is the gauge mounted to the handlebar as I posted then feed the bulk of the harness into the airbox and zip tie it to the filter cover then feed the wiring harness along the engine harness under the seat. The O2 sensor comes back out the airbox and gets zip tied to the clutch cable bracket on the right side near the front engine mount and into either the front or rear header pipe. The WB signal wire has good length and feeds from under the seat over the shock area into a cutout in the frame then plugs into the stock ECM harness plug for rear O2 sensor.

I can compare the enrichment tables of stock VS race EEPROM and take a screen shot.

lowkey once I'm done traveling I'll get the set up out and check the max lean and voltage it puts out. It will probably be some time this weekend.
I dont have and AFR guage but other then that sounds about the same to yours.

View attachment 11336

Here is race vs stock EEPROM's enrichment tables(BUEGB vs BUEIB), I have no way to verify if the data was changed or not as I got them from the internet and not an ECM. The ECMID checks out though... I opened a stock XB9 EEPROM and all values for enrichment are the same except WOT is 105% instead of 110% on the XB12 EEPROM's.

Those tables look the same in my bike as well. Seems like if your bike is running good, with out jerking, and stumbling, maybe acceleration enrichment is not the culprit.

How ever I do think mine is running "lean" during idle. Tuning guide recommends 14.7, and as low as 13.2 for crisp idle. I did fattened up the idle area a little and will have to see what it looks like after weather gets a bit better.

 

[Cooter]

FYI, Enriching the idle to as little as a 14:0 target makes for a stinky stop light.

 

[yugo42]

It finally stopped raining enough around here to get some logging in! Unfortunately the bike behaved pretty well the whole time so I'm not sure what troubleshooting I will get out of it. I had corrections off and weather was a perfect 68 degrees so I'll probably have to try to keep logging in different situations. Haven't had a chance to bring it in to MLV yet.

I was thinking as I was riding - one of the things that makes this cruising stumble so hard to deal with is how jerky the throttle is when going from idle/cruising/coast to a very slight throttle application. Slight adjustments at lower speeds result in lurching. Can this be dealt with in the acceleration enrichment tables to some degree rather than fiddling with the whole map?

 

[cossack84]

It finally stopped raining enough around here to get some logging in! Unfortunately the bike behaved pretty well the whole time so I'm not sure what troubleshooting I will get out of it. I had corrections off and weather was a perfect 68 degrees so I'll probably have to try to keep logging in different situations. Haven't had a chance to bring it in to MLV yet.

I was thinking as I was riding - one of the things that makes this cruising stumble so hard to deal with is how jerky the throttle is when going from idle/cruising/coast to a very slight throttle application. Slight adjustments at lower speeds result in lurching. Can this be dealt with in the acceleration enrichment tables to some degree rather than fiddling with the whole map?

Here are the things that made jerky throttle better on my bike.
I found out that intake seals leaked and spent couple month battling that issue. I ended up using Hylomar blue sealant. Also found that some times when, either unloading or downloading to ecm, some cells can get corrupt. Downloading a "fresh" tune that never been used before, cured some of the jerking. And the last thing that I think made a difference was taking time to adjust throttle cables. There are few data logs I have that show throttle jumping from 10 ADC count at closed to 20 ADC, with no readings in between. After the above changes data log slows throttle positioning at 10 and 12 and 14 and 15.
Hope this helps, I'll try posting few screen shots from mega log viewer.

 

[yugo42]

Thanks for the input! I'd love to see the mega log viewer screen shots. I replaced the intake seals on the bike a couple years ago after I first got it on the road - they didn't show signs they had been leaking and the new ones didn't change things. That said, I didn't use additional sealant on them. The throttle cable adjustment thing is interesting - I'm guessing the cables had too much tension between them? I just got my logs onto the PC last night and played with the MLV HD free trial to get a feel for how to set up the plots before I buy a license. I'll be on the lookout for TPS jumps when I dig into it.

 

[cossack84]

Thanks for the input! I'd love to see the mega log viewer screen shots. I replaced the intake seals on the bike a couple years ago after I first got it on the road - they didn't show signs they had been leaking and the new ones didn't change things. That said, I didn't use additional sealant on them. The throttle cable adjustment thing is interesting - I'm guessing the cables had too much tension between them? I just got my logs onto the PC last night and played with the MLV HD free trial to get a feel for how to set up the plots before I buy a license. I'll be on the lookout for TPS jumps when I dig into it.

Cant find a log with choppy load data in it. Here are couple of screen grabs of the latest ride and slow/light load.

Mind that mine is still not as smooth as I think it should be. considering that cursing at ~75 mph load is ~45, so going from 10 at idle and jumping to 20 would make it very jerky. It also could be a mechanical/linkage problem.

 

[yugo42]

Cant find a log with choppy load data in it. Here are couple of screen grabs of the latest ride and slow/light load.

Mind that mine is still not as smooth as I think it should be. considering that cursing at ~75 mph load is ~45, so going from 10 at idle and jumping to 20 would make it very jerky. It also could be a mechanical/linkage problem.

Thanks for the visuals! Does your AFV/EGO value of 115 at cruise concern you? That seems like it must need to add a lot of fuel consistently.

I finally logged a couple more rides - one with corrections turned off, and one with them on so I could see how the system was correcting. The riding felt a little smoother running open loop, but the data consistently showed idle being lean and cruising being rich. Since I'm just using a narrowband, it was hard to tell how much. Turning everything back to default EGO/AFV settings showed that most often the EGO correction was taking away fuel at cruising (89-95) and adding fuel at idle (105 or so). There were areas in the data with exceptions though. I didn't have enough rides in to establish AFV but previous times I checked it with ECMDroid it always hovered around 105. The first ride was at about 70 degrees F, the second about 77.

I'm guessing this means my base fuel map needs work, but I also wonder if it is worth replacing the o2 sensor just to eliminate a possible issue there. It seems to be behaving normally, bouncing rich to lean with corrections in place and outputting expected voltages.

 

[Cooter]

Thats a pretty minuscule sweep of adjustment.
If you are looking for the corrections to stay at 100 exactly, I think you misunderstand their purpose.

 

[cossack84]

Thanks for the visuals! Does your AFV/EGO value of 115 at cruise concern you? That seems like it must need to add a lot of fuel consistently.

I finally logged a couple more rides - one with corrections turned off, and one with them on so I could see how the system was correcting. The riding felt a little smoother running open loop, but the data consistently showed idle being lean and cruising being rich. Since I'm just using a narrowband, it was hard to tell how much. Turning everything back to default EGO/AFV settings showed that most often the EGO correction was taking away fuel at cruising (89-95) and adding fuel at idle (105 or so). There were areas in the data with exceptions though. I didn't have enough rides in to establish AFV but previous times I checked it with ECMDroid it always hovered around 105. The first ride was at about 70 degrees F, the second about 77.

I'm guessing this means my base fuel map needs work, but I also wonder if it is worth replacing the o2 sensor just to eliminate a possible issue there. It seems to be behaving normally, bouncing rich to lean with corrections in place and outputting expected voltages.

I have intake leak that drives my AFV that high. Not worried about it at the moment.

I doubt there is anything wrong with your base map. If you change base map, O2 will compensate for it with AFV and you will be right back at square one. If you are dead set on adjusting AFR in closed loop you would have to change the set point for O2 target voltage and O2 low/high voltage. But I would think very hard about why you want those changed. I don't think what you have described above, is the reason for choppy low load that you have.
What does you load data looks like when you apply small amount of throttle? Would be nice to see a screen grab.

 

[yugo42]

Cooter - I understand how the EGO correction works and I don't expect it to be at 100 all the time. The ECMSpy Tuning Guide Version 2 states that the further your fueling is from stoich based on fuel maps, IAT correction, and engine temp correction, the less smooth it will be. I guess -12% isn't that far off but I have no smooth DDFI2 Buells to compare to. Having the bike cruise more smoothly in open-loop mode is just strange to me and I'm grasping for ideas at this point.

Cossack84 - I attached two screen shots - one from my open loop ride and one from closed loop riding. I did notice as I was taking these that the pulse width 1 (front?) was longer on the closed-loop ride (outside of a couple random high spikes where PW2 shot up for 1 datapoint) and pulse width 2 was longer on the open loop ride. My engine was fully warm so engine temp correction should be 100 for both cylinders I would think.

I know these are only snapshots, but does anything jump out at you guys?

Open Loop:


Closed Loop:

 

[cossack84]

yugo you care to send me those logs, I can't really see anything in those pictures, they are kinda small.

 

[yugo42]

Sorry, I tried to embed the PNG images so I could label them but it automatically reduces them to small JPGs I guess. I will definitely send the logs when I get home. I really appreciate your input on this!

 

[lowkey]

Yugo42, your screen shots don't even show the .msq file was loaded... pull the file with ECMSpy from the EEPROM that was run when you were logging. Load the rear fuel map into the top right, load the AFR map on the bottom right. ONLY run analysis on open loop log, delete the other CL log as it is no help. If only running a rear NB 02 sensor ONLY run analysis for rear cylinder fueling map with NB 02 enabled (TPS setup to LOAD). After analysis on the rear you will have to open tuning software of your choice and change each front cylinder fuel cell an equal % that analysis applied to the rear map (you can save the .msq and load it for the rear fuel map to save time). You will do this entire process over and over until corrections from analysis are within + or - 2-4%.

As for jumpy throttle response, ECMDroid live monitor would work best but you can put tape on your grip and mark TPS values against a tape mark on the grip control and look what TPS/RPM value it is occurring. Check the timing map as most likely there is a jump in advice timing. Or you can add resolution to the maps by scaling rows and columns into that area of the map.

Have you read up on using MLV with NB 02 on a DDFI2 bike?

 

[lowkey]

So I quickly looked back on the first page, you have an 04 XB12 correct? So you have the 0-RPM and 8000-RPM column that are not being utilized on stock mapping. Whatever your TPS value is at the throttle stop is what you can set the bottom TPS row to. Use best judgment if shifting TPS Rows around (I'd leave OL switch point value and WOT switch point alone and focus on the CL TPS rows).

For setting fueling at idle all you have to do is enable CL and watch GEGO value %, once the temp is 160C warm up enrichment is done so you take the % value an add or subtract it to the live cell within the tuning software example: Idle 1000RPM TPS14, cell is lit at 64 on the rear cylinder fuel map and GEGO is 107% (it will fluctuate but if this 107% is the center take 7% an add it to the 64 value of the idle cell, do this same 7% for the front same cell if you don't have a separate 02 sensor. You can screw in the idle stop screw and adjust up in the TPS/RPM. Now that you have idle dialed in you shouldn't need to mess with it again. You can take it a step further and a 2 degrees timing to the idle timing area of the map which has a value of 0 degrees on both front and rear timing maps, once RPM stops climbing adding the 2% it is dialed in, for reference I run 8 degrees in this area of the timing maps.

It's your choice to enable CL idle/timing ECM adjustment or not as this will be a timing multiplier during warmup and adjust fuel accordingly BUT if disabled it will go off the 8 degrees timing in my case and the 64 + 7% fueling in the above example + the warm up enrichment multiplier table within the EEPROM.

From the ECMSpy website: http://www.ecmspy.com/tgv2/guide2.shtml#I.%20Calculation

8.1.7 Tidying up of idle
As a result of setting up the Closed Loop areas, the idle mixture setting was complete, however the ignition timing was not optimised. Since the idle region is defined with a closed throttle, ignition advance was not going be knock limited. Timing in the idle region is corrected as a function of temperature , see 5.2.2, however the base map for DDFI and DDFI-2 is set for zero advance.
For optimisation, the base map was increased in steps of 2 degrees until engine speed stopped increasing. At this point, the idle screw was adjusted to bring the idle speed back down to 1050 RPM. The result with the fuelling changes was a more stable idle at steady state and on engine deceleration.

 

[cossack84]

wrong post

 

[cossack84]

Sorry, I tried to embed the PNG images so I could label them but it automatically reduces them to small JPGs I guess. I will definitely send the logs when I get home. I really appreciate your input on this!

No problem. I found I have better luck uploading to image hosting site first before posting on a board.

 

[yugo42]

Cossack84 - What's the preferred way to send the logs to you? The zip file is larger than the maximum attachment size for the forum. Email/Dropbox?

lowkey - Thanks for the input and tips. I downloaded the XPR map extractor the other night but haven't yet had a chance to convert my maps for MLV. The bike is indeed an '04. I have read through the ECMSpy tuning guide a few times and several forum threads related to the subject but so far have been going through the logs as a troubleshooting tool to see if anything obvious is going on before I change fueling/timing. I took the closed loop log to see what the level of EGO correction was, and try to figure out why the bike was much smoother in open-loop running. It sounds like I probably need to go down the path of improving the tuning overall.