| Author | Message | ||
     Xb9 |
I was hoping maybe we all could get an anonymous tip on this - where does the DDFI-3 ECM learn on an 1125r? What is the best practice based on the ECM programming? To put it simply, what range of steady state RPM cruising is the learning area? I noticed on the XB board that there are some '08 xb owners also that are experiencing poor low RPM performance and poor gas mileage as some 1125R owners. So is appear's to be a global DDFI-3 issue. What really got me thinking about this is I tried something different on my 1125r that has seemed to improve the FI lower RPM and surging. And it could explain why some say theirs are fine, others not so good. I raised the RPM that I normally cruise in to 4500 or higher. Soon after I did a bunch of interstate cruising at higher RPM's my rear LFV changed from 100 to 105 for the first time in 2600 miles of ownership! We all know from experience that the earlier XB's required steady state cruising to get a good learn. So one would assume Buell's DDFI-3 systems need to learn also. (It's obvious with the AFV values in diagnostic mode). And if all things are relative one would assume that since this engine revs higher that it's learning cruise RPM would be much higher. I'd be curious if the owners that say their bike runs great practice higher RPM cruising in their normal riding routine. I'd also be curious if the owners that say their bike runs poorly rarely get on the interstate or steady cruise RPMs are generally less than 4500 RPM. Maybe we can 'learn' something here.  Post up with: 1. Bike runs poorly or good 2. Do you steady state cruise regularly at 4500RPM or higher? It would be great if someone 'in the know' would just pop in a little post here as to what cruising RPM the 1125r learns best? A little information could go a long way! | ||
| Spectrum |
Bike runs good! I cruise at steady state above 4500K, but no AFV learning. AFV values only change when I thrash her. (e.g. trackday riding) | ||
| Thurstonbuell |
1. "Bike runs poorly or good ? " Poorly 2." Do you steady state cruise regularly at 4500RPM or higher? " Yes , I commute to work daily 30+ miles on way .I also pack on alot of miles on back roads , thruout the entire range of rpm's , steady and all over the place thruout the rev's , most of the time its interstate , but some afternoons it's nice to take river road all the way home. So she has alot of miles doing both styles of riding "I raised the RPM that I normally cruise in to 4500 or higher ?" Yes , I try to run 5000+ rpms all the time now. "What really got me thinking about this is I tried something different on my 1125r that has seemed to improve the FI lower RPM and surging." No improvements for me , In fact this last weekend was blazin hot , and I developed stalling issues that makes my motor sound like somethings gonna break next time she does this. (Message edited by Thurstonbuell on June 09, 2008) | ||
| 01xjbuell |
1. Bike runs very good 2. Steady state 3500-4000 depending on what gear 3. AFV's changed to 102-103 f 105 r No stumbling at all -Nick | ||
| Zac4mac |
I've been trying to get this info from BMC for 3 months. The only thing that I've done that SEEMS to have an effect is to apply gentle pressure and vacuum to the baro sensor. Ever since I did that, Loretta runs MUCH smoother. No effect on AFV but "running" is better. I have tried steady state for 5-10 miles at 3, 3.5, 4, 4.5 and 5 k rpms with no change in AFV. Once I rode the 10 miles to work and stayed in 2nd the whole way. no change. 8200 and change on the ODO. Z | ||
| Thurstonbuell |
Zac , I'm gonna give that a try , I'll try anything at this point , plus I think my mileage slipped down from 30 to about 24mpg , dont know if thats due to evaporative gas fumes blowing out from the vent tube or running around everywhere at 5000 rpms and above no matter if Im carving up some country roads or getting thru town trying to do 25mph , just to try and avoid her tourettes syndrome. | ||
| Markrd500 |
Cruising 4500-5500 Twisty roads with plenty of full on stuff. F-99.5 r-100 DR.Stutter low down in rev range. | ||
| Bigdog_tim |
Weird. My reflash was done at 2200 miles. Stumbles bad between 3-4K. AFV was 100/100 for at least 400 miles. AFV only changed after a VERY spirited ride (20-30 minutes of 4-8K RPMs). AFV changed to 100/105. Still stumbles - but I THINK it is better. Never a stall. | ||
| Jpfive |
From the Electrical Diagnostics Manual: 'The adaptive fuel value, learned during closed loop operation, is applied to open loop operation to adjust fuel and spark maps for optimal performance. 'During closed loop operation, the O2 sensors provide input for an optimal air/fuel mixture resulting in reduced emissions, good fuel economy, and smooth power. In order for the system to enter closed loop operation, certain conditions must be met: - O2 sensors must be at the normal operating temperature of the engine. - Operating below 4200 RPM with engine under steady or light load conditions. By using both open and closed loop systems, engine performance is continuously tuned to compensate for changing conditions and providing maximum performance.' Hope this is helpful... Jack | ||
| Jpfive |
FWIW, and to add to my previous post... MY UNDERSTANDING of the barostat function is that it does NOT contribute to the learned AFV figure generated during closed loop operation. The barostat corrects from a separate table, based on its own readings, which is applied during closed AND open loop operation, and the learned AFV corrects for everything else. AFV values reported by owners from different altitudes seem to support this 'conjecture' on my part, which I defend as supported by past experience with EFI systems with barostats, and a careful study of the manual. I submit this as a contribution to the discussion only, and hope that an Anony will pop in to confirm or correct. Jack | ||
| Id073897 |
WD0J8, RPM/50 vs. TPS %*256 LCL region upper boundary: 60 - 40 80 - 50 90 - 54 110 - 58 LCL region lower boundary: 60 - 18 66 - 20 72 - 20 80 - 21 90 - 22 100 - 24 110 - 26 Regards, Gunter | ||
| Jpfive |
Dang, Gunter...where do you get this stuff...? Thanks for the info, and for you earlier post of the barostat table. Jack | ||
| Zac4mac |
Loretta seems to be running almost as good as she used to - Ever since I measured and checked functionality of the baro sensor. When I check MAP voltage , in Diag Mode engine off, it's the same as what I measured on the baro output, 3.3 volts. My baro sensor was right-side-up, but I think the diaphragm was stuck. Z (Message edited by zac4mac on June 11, 2008) | ||
| Xb9 |
Gunter - THANK YOU!!!!!!!!!!!!!!!!!!!!!!!! Can't wait to use your software to fix my 3500rpm mapping! damn the EPA! | ||
| Id073897 |
It's questionable if these bottle warmes will ever be supported. But we know a bit about them nevertheless. Regards, Gunter | ||
| Xb9 |
RPM TPS Low TPS High 3000 7.03% 15.63% 3300 7.81% 3600 7.81% 4000 8.20% 19.53% 4500 8.59% 21.09% 5000 9.38% 5500 10.16% 22.66% | ||
| Cataract2 |
So what does all that stuff mean. Someone please explain. | ||
| Xb9 |
LCL = Learning closed loop (region) TPS = Throttle position Sensor In my last post I calculated what Gunter posted as the actual values in the ECM EEPROM. What it means (example)? The ECM learns: at 3000 RPM w/ a TPS percentage of 7.03 to 15.63 etc. Putting it simply it learns at steady state cruise light load between 3k and 5.5K RPM What I can't figure out is all the surging occurs when it is in closed loop - when it should be reading the O2 sensor feedback and adjusting real time. Doesn't appear that that part is functioning properly (speculation) or the map is so far off in that region that that the ECM won't overcome it (limits) (Message edited by xb9 on June 11, 2008) | ||
| Slypiranna |
...or if the actual air/fuel mixture target within the closed loop regions is in fact, 14.7:1??? I'm finding much leaner mixtures than stoic within closed loop and now on more than one 1125 example...which makes sense. Most catalyst equipped machines target 14.7...how then...could you begin to pass epa w/o a cat and still operate at stoic? It is my opinion that the actual design target is much, much leaner than we all might think. Being that o2 feedback is the main input, that is where I'm now testing for further improvements but only in their accuracy and speed of switching, NOT forcing a richer condition. I agree 100% with Jpfive's post. The atmosphere correction would be in a separate table. I also now believe it inaccurate, based upon the data thus far, that AFV's being used to compare elevation differences. AFV's are mainly looking to the o2's, not the baro for fueling corrections. Example, my 11 was at sea level in florida since new in Jan until trailering up to ohio early this month. When I left, AFV's were up and down since the reflash from 100-105 but finally settled on 100FR. After getting here in ohio, approx +900' elevation, they took some time but have settled in on what the wideband says is more in sync with F&R cylinder, 105F 100R. I think this supports that the baro is separate from AFV's. It should also be noted that the stock spring loaded throttle assembly COULD/can/and do(on some!)...flutter outside of the LCL TPS % regions under certain operating conditions. Think about this for a moment, as the butterflies can and do move independent of the stock throttle's actuator. With that cat now out of the bag, I believe that there is NO problem with the ecu or it's programming at all. At least non of the examples I've been involved with. Post flash/Pre flash, this same condition has occurred in the problem spots that many of us have experienced. To end this long post, I think we are dealing with a combination of very minor issues that are adding up to one seemingly large one...making diagnosing very tricky. If we all keep posting data as we get it though, I'm sure all negs will be resolved soon. We CAN figure this machine out! There should be no secrets in making this great product the best it can be and dare I say, still meet EPA standards. (Message edited by slypiranna on June 16, 2008) | ||
| Jpfive |
Great post, Slyp. I think your analysis is correct. Thanks for working so hard on this. Jack | ||
| Xb9 |
MM, agree with everything in your post. Narrowband O2 sensors in their design are only made to be accurate for stoic (14.7:1). If the rich/lean/target values in the ECM where shifted to run leaner (or richer) than Stoic with a standard O2 sensor, I would think you would have some serious accuracy problems. I always wondered if the O2 manufacturer could build an O2 sensor that was calibrated to something other than Stoic. Never heard of one but seems possible. | ||
| Mustangturbo |
Just a thought... But, I think that it has something to do with the throttle body design. I've been around some race cars that couldn't run in a lower rpm range because of stacked/down draft intake manifolds. If you tried to hit the throttle in the lower rpm's they would "load" and have to be cleared. It seems like airflow may be so free that it is causing an air fuel problem. Who knows maybe altitude would help, because of the lower atmospheric pressure. Less pressure less air better mixture? Just a thought... | ||
| Id073897 |
Narrowband O2 sensors in their design are only made to be accurate for stoic (14.7:1) NB sensors are mostly made for detecting a cycle (from lean to rich or v.v.). This is the way the ECM seems to work too, it's reacting on these cycles. A stable voltage emitted from an NB sensor is usually interpreted as a sign of failure. So if the mixture is stoich, it's just stoich in average, but not at every point. Regards, Gunter | ||
| Slypiranna |
Excellent information Id073897...and the more adjustments over and under per unit of time, the closer the resulting averages will be to the target mixture. Keep this thread alive and learning! | ||
| Id073897 |
and the more adjustments over and under per unit of time, the closer the resulting averages will be to the target mixture. Yes, but this might be all theory only. First the O2 sensor has a noticeably dead time, up to 200 ms. At full spead this means 20 revs "out of control". Not really, but reaction is delayed (which is not always bad). Secondly take a look at this image: http://ecmspy.com/download/gunter/WOT.png It shows a typical WOT situation, WB monitored. Around the cursor position you will notice the lean mixture although no corresponding change in injector pulsewidth pw2 occured. This can be watched very often with the XB models, so fuel supply is not as steady as theory states. Probably this is caused by the single manifold, creating interference effects. Regards, Gunter (Message edited by id073897 on June 16, 2008) | ||
| Slypiranna |
Awesome info Gunter! Keep it coming! I've been busy today with other issues like making $'s but managed to rewire this 11's stock o2s into the wideband's single narrow output... Unbelievable results after only running/riding for less than ten miles! AFV's went solid balance, as previously thought, based upon a single source, but this quick? Accuracy!? The heated function? The sensor's placement in the exhaust sampling stream???!!! ...will post more info tomorrow after making sure the data is correct by reversing and retesting. This is all very interesting and a sure means for one to stay outta trouble...or get deeper into it! None the less, it should prove to be more useful information in our quest for learning more. mm | ||
| Xb9 |
Oh, cool idea for testing! So you took one ZT2 and replaced one of the NB 02 sensors with the wideband, then wired the WB controller's simulated narrowband output to both the front and rear O2 sensor leads going to the ECM? I would eventually put a dedicated controller on for each cylinder. The widebands are so much more accurate. Did it fix the lean running condition in closed loop? | ||
| Slypiranna |
NO ZT2's on this one, YET! Trying my best to keep within street legal rules... I simply wired in the FAST's NB output function...something that has eluded me for over two months now to test...just hit me to try out of sheer determination I guess. While I've gained a whole lot of usable data with the widebands addition, I started to think that this might be looking with horse blinders on...as to what might be the simplest of issues. Nothing fixed, per say but another step in the right direction in understanding what is going on here and more importantly, what this ecu's learning curve and parameters are. It has always been a question in my mind as to why the AFV's take so long to adjust, in real time as feedback has always been present to reflect upon. If I had to say right now though, butt dyno and feedback points to +'s! More info to come! | ||
| Slypiranna |
Reasons why AFV's take so long to adjust? I'm not yet convinced that the ecu's parameters warrant an adjustment from the way we are looking at it. Meaning, the time element. There seems to be a whole lot more to this Adaptive Fuel system than what information we've understood thus far. I'm starting to believe that this system is much more advanced than previously thought. Comparing this system to even aftermarket programmable race units doesn't even come close. It's starting to grow on me in a positive way! True, much of what we are trying to understand here is still questionable theory. Until one of us trips upon via trial and error methods and/or enlightened by factual information from someone in the know...the testers march on. It should also be noted that while I've posted some off the wall tests and or voiced arguable opinions, I am in no way recommending that anyone fiddle with their own 11. My warranty, due to many mods and tests, might be void at this point. I would hate to hear that someone else's warranty be voided in replicating as such. If in fact, by my sharing on this website seems at all questionable by others, please inform me of such and I will not post anything from this point on. No need to pm me, post it for everyone to understand why. Back to the thread's concerns, I've ordered different o2's for continued testing. XB9's post, keeping the F&R cylinders separated, has been burning in my thoughts and will be done so in order to further evaluate any credible improvements/changes. I threw some codes with the wideband's NB output...they did self clear by hooking back up the stockers. My interpretation of resolution though, went backwards. More time and more testing to follow. Who else has info in an effort to keep this thread alive and learning? | ||
| Bob_thompson |
Slypiranna, I have really enjoyed your input here and on this particular thread and hope it stays alive. I am really stupid on these ECM's and have enjoyed learning from it. I'll help keep it alive with at least questions, which brings me to one: if you have disconnected the throttle solenoid actuator mechanically only, have you had a fault code yet, has it stayed and just how much benefit have you seen? I am completely stock yet without apparent problems. Many think its like night and day with throttle response. Good job and keep up the fine research. Anybody else? Bob On a similar note, my ECM has seemed to adjust to at least altitude changes quickly as on aa average ride I go from 4200 ft. to over 9000ft. and never seem to see a difference in performance going either up or down in elevation and I ride pretty aggressively in the twisties. (Message edited by bob_thompson on June 19, 2008) (Message edited by bob_thompson on June 19, 2008) |
