| Author | Message | ||
     Downhilldanny |
I recently noticed that my 05 CityX was getting less mileage on same tank than usual (I commute to work every day, so easy to notice changes). I took the bike in to local HD dealer for a check. Plugging the bike into their computer they confirmed that it was running rich, and reset the ECM (I think - not technical) and told me that I should ride the bike below 3500rpm for the first 4 miles while the bike reaches operating temperature, and then ride above 3500rpm if required / traffic conditions allow. Fuel economy improved back to "normal" for the route. A week later I had the bike back at HD dealer for an accessory fitment, and while the bike was there, they plugged into the computer again to see how the ECM (?) was doing. This time round it was running lean, and was reset. My question: Does the ECM (?) have a learning mode ?, and if you rode the bike in such a way as to throw the ECM (?) out, is it not supposed to re-teach / adjust itself back to best operating mode if you rode the bike in a "proper" way after some throttle twisting ?? These bikes tend to bring out the hooligan (fun side) in a rider, so why would the ECM (?) be so sensitive ? | ||
| Lovematt |
The ECM does in fact learn while you are riding. I believe the RPM range is around 3200 for a number of minutes...I think it is 20 minutes. I can confirm this does occur with elevation and temperature changes. For example I took a ride last year into the mountains where the average elevation I was in changed from ~500 to always over ~5000 for about a week. At first right after climbing the bike ran a little different (probably rich as there is less air up there but the fuel delivery was still the same). I rode around up there for about a week and after the first day it smoothed out and returned to how it was before (mostly). Then when I returned back home to lower elevations, the bike again rode a little different (this time possibly lean as there is more air now) for about 1 hour (popping more). But then it smoothed out again after that first hour and ran good again. I also had a slightly rough running bike when I first had my Race ECM fitted. I got the bike back and it ran okay but the acceleration and Idle were a little off. I rode at ~3200 RPM for about 45 minutes and after that ride the bike ran much better and stayed that way. It is a nice feature in our bikes but it also presents difficulties for trying to "alter the chip" by adding on devices. But I understand there is a way to do that now which is being finalized. | ||
| Thomas_lindemann |
How can the ECU learn? If you rode it at 3500 rpm for 20 minutes the only thing it can sample is 3500rpm at that throttle position. Maybe the one wire O2 sensor is so slow to respond in open loop it takes that long to adjust? | ||
| Lovematt |
Maybe "learn" is not the right word...it is not artificial intelligence per se but the ECM does change how the bike runs based on the air temperature, humidity, O2 Sensor, and other factors. I know this is intended and I even spoke with a buddy of mine who works for PowerCommander and he mentioned they had a lot of difficulty with the Buell chips since it tried to compensate when their device made changes to the fuel maps... My understanding is the chip does not necessarily have static air/fuel maps but rather compensates based on input from a variety of sensors as the bike is run to try and maintain the optimal A/F ratios. | ||
| Buell920 |
ADAPTIVE FUEL VALUE The Buell DDFI system has the ability to “learn” the engine fuel mixture needs. When the motorcycle is running in the closed loop mode and operated in a specific engine speed and load range, the system will compare the feedback from the O2 sensor to the base programming stored in the ECM. If a difference in these values is detected, the ECM will recalibrate the system program to compensate. This correction is termed the adaptive fuel value or AFV. This compensation value allows the system to adjust to different altitudes, air densities, and to some degree engine variations and wear. The normal AFV ranges based on altitude are between 85 and 115. The higher values are found at lower altitudes and the lower values at higher altitudes. The AFV correction is only applied to the fuel mixture during OPEN loop operation. During closed loop operation the O2 sensor signal is the primary compensation method. The AFV will be learned when the bike is operated at engine speeds between 2500 and 3500 rpm at road speeds in the 40 to 60 mph range under a steady light load (no down grades or steep upgrades, decelerating or accelerating) for 2 to 3 minutes. The AFV can also be reset to 100 by using the scanalizer at any time. The AFV value can be a valuable diagnostic tool. Compare the AFV values of bikes in your area which are running fine during routine services. When you are working on a bike which may have a DDFI problem, compare the values. If the AFV is higher than normal, the system is trying to correct for a situation which is causing the mixture to be too lean. Look for intake manifold or injector O-ring air leaks, incorrect ignition timing and TPS zero setting, low fuel pressure or a fuel line restriction as well as a sensor malfunction. If the AFV is lower than normal, the system is trying to correct for a situation which is causing the mixture to be too rich. Look for incorrect ignition timing and TPS zero setting, high fuel pressure or a leaking injector as well as a sensor malfunction. The O2 sensor can also cause the AFV to be set incorrectly. Internal shorts in the lead, poor electrical contact with the exhaust system or contamination can all affect sensor performance. | ||
| Al_lighton |
During closed loop operation (roughly 1500-4000 RPM, light load, STEADY STATE) the bike runs entirely on the O2 sensor, and it will run 1t 14.7:1 AFR because that is the only AFR that the sensor has a resolution at. If you slowly open the throttle, it senses it getting leaner, and commands more fuel. If you slowly close the throttle a little, it senses it getting richer and commands less fuel. But if you move the throttle a little faster or more, it kicks the bike back into open loop where the internal map and all the sensor inputs are used to determine the required injector duration. A subset of the closed loop domain is the learn domain. While in learn mode, the ECM compares the injector duration obtained from the O2 sensor operation to the injector duration calculated from the map for the same operational point.(i.e., throttle position, RPM, temperatures, etc). It creates a correction factor (called the Adaptive Fuel Value) that is used as a scalar against the open loop injector durations. Buells do not have manifold pressure sensors to adjust for ambient barometric pressure/air density changes. It uses the closed loop learn mode to adapt for these changes. The AFV is a constant scalar applied to the open loop map. It is not capable of selectively raising or lowering the fueling by different amounts at different RPM/throttle settings. This is why it does not really compensate well for exhaust or intake changes unless those changes have a very linear response. The AFV is a dynamic number generally between 80-120%, though it can go further than that range. The AFV will change every time you ride the bike in the learn mode, depending not only on the ambient conditions but also on which operational point you've put the bike in when it was learning. The map values are not so precisely input that the comparison at every operational point will yield the exact same AFV, even if the ambient conditions are exactly the same. But they will be close, and close is good enough. After all, they didn't map YOUR bike. They mapped a similar bike and pushed the same map into every bike made. But there are differences between every bike. | ||
| Aeholton |
Does that mean at light throttle cruising along the bike is at 14.7:1 AFR? Isn't that a pretty lean condition? Wouldn't that create excessive engine heat? | ||
| Teeps |
Al_lighton & Buell920, Thanks for the insightful dissertation on the Buell DDFI fuel system. The Ulysses is my first F.I. motorcycle so it's always a pleasure to learn something new. Al, I received the ECM yesterday. I hope to have it installed today or tomorrow. So I should know, by the weekend, if it helps with the surging problem, we discussed last Friday. | ||
| Teeps |
Aeholton Posted on Thursday, June 01, 2006 - 10:08 am: Does that mean at light throttle cruising along the bike is at 14.7:1 AFR? Isn't that a pretty lean condition? Wouldn't that create excessive engine heat? Check this out: http://www.taftan.com/thermodynamics/COMBUST.HTM | ||
| Downhilldanny |
Cool, thanxs Buell920 & Al_lighton for the input, definitely taught me a few new things. Also explains the 100% value the HD dealer kept referring to. When my bike was running rich, the value was 86% - I stay at high altitude. Since I have become aware of this learning mode, I have been able to link it to how smooth / rough my bike rides, especially in 1st or 2d gear in traffic with minimal throttle. When the AFV value is out, the bike feels very rough, and sometimes requires a bit of clutch to ensure smooth riding. | ||
| Thomas_lindemann |
Isnt that how any closed loop injection system works? | ||
| Al_lighton |
14.7:1 is the stoichiometric A/F ratio for gasoline. IF the combustion chamber was perfectly shaped so that every fuel molecule found an air molecule, any more or less fuel would yield less power. More fuel would displace useful air, less fuel would yield no combustion for the excess air. But in reality, not every fuel molecule finds an oxygen molecule. You have some unburnt fuel on the left side of the combustion chamber longing to get married to that O2 molecule on the right side, but before they can get together, along comes that piston and pushes them out into the pipe. If enough doesn't get burnt in the combustion chamber, they get married in the exhaust pipe and explode there (muffler popping). This is compensated for by putting more fuel than necessary into the chamber, so that all those lonely O2 molecules find some fuel after all. But if you put too much in, you lose power because you displaced O2 molecules to get more fuel in. 12:1 AFR will NOT yield more power on a Buell than 13:1. 13.5:1 is close to optimum. Narrowband O2 sensors are highly non-linear 0-1VDC devices that are designed to allow an ECM to hold a 14.7:1 AFR. They do this by bouncing the voltage back and forth between approximately .4 and .6V. It isn't coincidence that they have all their resolution at 14.7:1, they were designed for that. A gadjillion automobiles run on them. Modern automotive FI systems use both narrowband and wideband sensors, as well as knock sensors and a host of other sensors, to optimize their fueling. I think the Buell system is a nice simple system that works pretty well overall. One could get better fueling with a more sophisticated system. But I like simple. I believe the info posted by Buell920 was written by Dan Hurda, the Buell powertrain lead engineer, so you can believe it. | ||
| Badbuell82 |
"Al_lighton & Buell920, Thanks for the insightful dissertation on the Buell DDFI fuel system." AMEN! We should really take up a collection and pay these guys for their time! I'm also a new Bueller and these guys consistently supply the most accurate and valuable Info on here! | ||
| Mikemax |
What controls the AFV, the ECM? I was wondering what changing to a race ECM would do to features like this? Or does the O2 sensor work closed loop with just the valve? | ||
| Mikemax |
I have about 650 miles on a 12 Long and my mileage last tank dropped about 20% (from just over 50 MPG to just over 40). For the first 500 miles I was varying the speed and staying below 3500 RPMs for the most part as a break in. This tank I commuted three 60 mile legs at pretty much a constant 70 MPH (pretty much 3500-4000 RPM) and would expect this learn mode to help. I haven't had any popping indicating a real rich condition as described above though so maybe wind resistance is that much more? I'm disappointed since one of my purchasing decisions was based on expecting around 50MPG. Any ideas? I'm sort of scared if I want more power and go the pipe/ECM route it will get even worse. | ||
| Buell920 |
Mikemax The O2 sensor, sends a signal to the ECM proportional to the amount of fuel remaining after combustion.as stated by al_lighton. The ECM then uses this signal as feedback in order to modify the amount of fuel delivered per engine cycle. Thus, the DDFI system maintains the optimum air/fuel mixture for complete ( or darn near ) combustion. The 02 sensor allows the system to compensate for changes in barometric pressure or altitude. so the need for a manifold pressure sensor is not needed. The 02 sensor is still utilized with the race ECM to get the best combustion in the cyl. | ||
| Buell920 |
BadBuell82 thanx for the kind words. | ||
| Mikemax |
Thanks, missed that in Al's post. And presumably a race ECM with no other changes would yield basically the same "best" combustion settings? Do the race ECMs retain the learn feature as well? I'm trying to learn as much as I can about the new bike. | ||
| Al_lighton |
Race ECMs are very similar, they do run the same (or similar) learning algorithms. It is possible that some of the set points in the race ECM have been changed. We know the differences in the fuel and spark maps, but we aren't as certain about the other parameters that may be different. For instance, the XB12 race ECMs like to hang on high idle when they are hot more than the stock ECMs do. There is a temperature disable on the decel function, and the temperature set point may be different on the Race ECM than the stock ECM. Or it ma be a temperature dependent timing function that is different. We aren't sure, we don't have visibility of the internal algorithms and other set points. It wouldn't surprise me if the skip spark rev limiting functions on the race ECM have different parameters as well. But when we tune a race ECM or a stock ECM with DirectLink, we can get the same performance from each. Al (Message edited by al_lighton on June 01, 2006) | ||
| 99buellx1 |
920 answered that in his last post, but it must have been missed. The esay answer is yes. I think you are getting a little mixed up with the purpose of the exhaust servo in your bike. It is not a variable in the ecms AFV. | ||
| Mikemax |
Thanks, I think I get it... just missed the references when speed reading at work I guess. The O2 sensor basically checks for off mixture in the exhaust and the AFV controls the mixture going in; I wasn't sure if those two were closed loop by themselves or if the ECM was a controller for these in between (as it is). Thanks again. | ||
| Mikemax |
Man, I am tired... I thought it said Valve not value. Duh. | ||
| Davo |
Can, and if not when, will we be able to read the AFV on the VDSTS? Currently I can reset it, but I have not figures out how to read the AFV value. Thanks. | ||
| Al_lighton |
It is very close....I spoke with Sandro earlier this week and he has located the hex address for the data. So now he just needs to imlement the display in the program. Al | ||
| Silverado140 |
I've just installed the Race Kit muffler and air filter. I'm waiting to install the race ECM until my appointment to get the TPS reset. Is it still OK to run the bike with the stock ECM, or will it not know what to do with the extra air and lower exhaust pressure? | ||
| Kurosawa |
Question, since a stock ECM tries to maintain 14.7:1 and best power is 13.5:1, can it be reprogrammed to the richer mix? Wouldn't that help my stocker run smoother, cooler, and a hair quicker overall? | ||
| Al_lighton |
Joe, You won't hurt it on a casual ride in cooler weather. But I'd be hesitant to go rail hard on it in hot weather in that configuration, and I would get the new ECM in asap. And do ride it steady state at 50-60 MPH as soon as possible so the AFV adjusts. Kurt, Yes, and no. We program the 68% and 99% throttle positions to 13.5:1 for max power. The 48% and 39% throttle position columns are blended at 14:1. But the bike will run at 14.7:1 when in closed loop because the sensor is incapable of doing otherwise, race ECM or not. If the tuner doesn't match the map to 14.7:1 over the learn mode, the AFV will skew, plain and simple. You should note that tuning richer down low is folly anyway. If you want more power, just twist your wrist and more power is available. You shouldn't tune your cruise range to 13.5:1 on ANY bike, you're just wasting fuel and polluting more for no good reason. Only the high throttle positions need to be tuned for a richer mixture. Now, fact is, it probably WOULD be a good idea to tune the cruise range to a point or two below 14.7:1 to account for real world issues that can cause some lean missing. But the O2 sensor and the AFV learning mechanisms on the Buell don't support this, so I would be hesitant to do so. I've seen the result it has on the fueling curves after the ECM has learned, and it isn't pretty. The folks tuning with DirectLink that aren't properly tuning the map in the learn-mode domain are making some pretty screwed up running bikes. | ||
| Brewtus |
Al, Could you please post and let us know when the new VDSTS software is available? I think the AFV reading will help us all out when tuning with the Power Commander or anything else we currently have. Thanks. | ||
| Badbuell82 |
"We should really take up a collection and pay these guys for their time! I'm also a new Bueller and these guys consistently supply the most accurate and valuable Info on here!" OK, since obviously nobody's taking up a collection, I at least, am going to be sure to purchase everything I need for my Buell from American Sport Bike in the future! It's the least I can do to compensate for such selflessness. | ||
| Brewtus |
Last time I checked people sign up for this site voluntarily. It is not that I do not appreciate all the help everyone gives, and I do support all the sponsors. As a matter of fact the people at my local dealer are looking into the Directlink software for the Buells and are asking where I am getting all this info. They are asking for contact numbers and web sites and I have certainly given them Al's site and have mentioned Terrys name. I must admit that it is strange that I know more than my Dealer does, but this is not just a job to me, it is an obsession. I want to get my bike to run the best it can without hurting my motor and my budget. | ||
| Silverado140 |
Al - Thanks for the input. I have the race ECM in my possession, and in fact will install it tomorrow morning prior to my TPS appointment. I definitely feel better about riding it to the dealership following your comments, since I have no way of getting it into my truck bed where it's currently stored. Also, please explain what you mean by "steady state at 50-60". Does this mean just cruising down a 2-lane blacktop for a while at those speeds without much accel/decel? And if so, how long should I do this? I love this bike and want to make sure I'm doing right by it. Thanks!! |
