| Author | Message | |||
     Mmcn49 |
On page 61, in Motorcycle Fuel Injection Handbook, Adam Wade makes the following comments about narrow-band 02 sensors: The vast majority of the time, the ECM is not paying attention to the 02 sensor. When they are in use, it is not uncommon to get surging issues and sharp transitions to roll-on, (common problems on big twins). 02 sensors are installed for the same reasons as catalytic converters, to cut down on emissions. Narrow band 02 sensors have no other use. In ECMSpy we can turn OFF the 02 sensor. Will turning off the 02 sensor put the fuel map’s Closed Loop area into Open Loop? What is the Up-side of doing this? What’s the Down-side? NOTE, to the moderators: The publishers allow brief passages to be quoted from the Handbook. The comments are not quoted word for word. | |||
| Mnbueller |
Yes it put the fuel map into open loop. The bike will run rich or lean, depending on other environmental conditions, because the bike will not be able to adjust the AFV to compensate. In the ridding that I do, I do not see any reason that I would want to shut off the O2 sensor.... | |||
| Swordsman |
"The bike will run rich or lean, depending on other environmental conditions" Like every other carbed vehicle on the planet. I wouldn't expect it to be a problem unless you lived in a mountainous area where the big elevation changes might cause it to act up. Even then I don't think it would be a threat to the bike itself. Turn off the O2, and lock the AFV at 100, and I'd expect it to be perfectly safe using stock maps. I've been tossing this same idea around... I'm sick to death of the surging, and I'll do anything to be rid of it. That includes selling it.  ~SM (Message edited by Swordsman on September 15, 2008) | |||
| Xl_cheese |
Do a datalog and then watch the ego corr number. If it's ~120 in CL region then if you disable the o2 your bike will be really lean. I think the o2 sensor plays an important role. | |||
| Swordsman |
Xl, that's not a bad idea to datalog first, and then you could lock the AFV at wherever your "normal" hits, and shouldn't have to worry about it. 'Course, I'm totally talkin' out of my ass here. I've never done it, and have very little experience. Just guessing based on stuff I've read. ~SM | |||
| Mmcn49 |
Been dealing with stumbling and surging all summer but have made progress. The first improvement came from Mike Cobb, (Xopti). Mike said the stock 0.49 volt 02 sensor mid value gives closed loop a lean 14.9 to 1 ratio. According to him, raising mid voltage to 0.52 changes the ratio to 14.65 to 1. He also suggested opening the other maps tab and setting Open Loop enrichment to 113 for12’s and 108 for 9’s, and to change the other line to 0 from 3, (lower left box). These changes made a big difference. Data logging and map changes also helped, but I still have issues. I went for a long ride on Saturday. AFV was set to 0 and the bike ran great starting out. The bike still ran pretty good all day, but as the day wore on performance dropped. When I spied the ECM on Sunday, AFV had dropped to 90. After resetting to 100 the bike ran better. I’ve not been able to find a way to prevent the AFV from changing like this. I have limited the AFV range from 95 to 105 and found the bike ran well consistently. Occasionally I ride in the mountains 2000-4000 feet and don’t know if limiting AFV to this extent will cause problems higher up. For this reason only, I’m a little hesitant to make that a permanent change. On the next data log run, I’m going to change the closed loop ratio in MegaLog to 14.2 to 1 and turn off the 02 sensor. I may just say screw it and leave the 02 sensor off and see what happens. If I can figure out the 02 sensor voltage value for 14.2 I’ll plug that in and see what happens | |||
| Swordsman |
One of the things I'm curious about is the different ways ECMSpy supposedly allows you to disable the closed loop. In one tab, you can turn off the O2 sensor altogether, but on the last tab you can disable closed loop "learning" separately with a checkbox. I wonder what the difference is? ~SM | |||
| Xl_cheese |
Swordsman. You mean open loop learning? I understand that if in open loop a prolonged lean condition is seen then it will raise the AFV. I don't think it really does anything. Mmcn49 - Try to adjust the lean threshold of the o2 sensor up from .41 to .42. That may help get your AFV up. I have been running with similar adjustments except I only raise the lean and midpoint up one point. I notice my AFV is up ~5-10 higher than before. | |||
| Xoptimizedrsx |
you actually are setting it to a number beyond recognition. its still on but not affecting anything. what you want to do is this. ride the bike in cruise read the afv. ride the bike in the open loop area. 5000 up to wot doing a really hard ride in low gears.read this afv asap. now the hard ride. should be a lower afv by a few points. if its opposite you nee3d the fuel map adjusted before you disable the o2.. if you want a safer way? get the fuel map set so the upper afv is aprox. 5 points less than the cruise afv. then disable the open loop learn. mike | |||
| Turk |
what you want to do is this. ride the bike in cruise read the afv. ride the bike in the open loop area. 5000 up to wot doing a really hard ride in low gears.read this afv asap. now the hard ride. should be a lower afv by a few points. if its opposite you nee3d the fuel map adjusted before you disable the o2.. This will not yield a good tune. You cannot appropriately tune for wide open throttle with a narrow band O2 sensor. Period. It can only read rich or lean of stoic and wide open throttle should not be anywhere near stoic. | |||
| Id073897 |
ride the bike in the open loop area. 5000 up to wot doing a really hard ride in low gears.read this afv asap. now the hard ride. should be a lower afv by a few points I'm not sure if AFV will change there except the engine runs way too lean in OL. "Too lean" means no rich reading from the O2 sensor for the given period, which will not happen with the stock map. Taking into account that AFV has to cover air pressure changes primarily, and these usually do not exceed +/- 3%, then adding that to the AFV monitored in cruise mode should be safe. Changes in altitude might lead to a larger difference and some adjustment of the AFV might be necessary. I put my XB9 in OL running a slightly rich map for about 5k mls. now, but I'm not sure if I would recommend doing so. The ECM does pay attention to the O2 signal, otherwise the AFV would stay constant. Regards, Gunter | |||
| Turk |
The AFV should stay constant for a good fuel map unless you're changing altitude. | |||
| Id073897 |
That's the problem with reality. It does not always corroborate theory ... | |||
| Turk |
Why would you want the AFV to float? Get the fuel map right and don't rely on an AFV to correct anything. | |||
| Buplaux |
Installed o2 on the front cylinder. Did a lot of logging. Ran fuel map ve analyzer on the rich side to create my maps. Disabled o2 sensor. Best thing I have every done for the bike!!!! If you plan to disable the o2 after logging I think you should change the open loop enrichment from 105% to 100% while logging to get accurate maps. Because if you log with it at 105 open loop enrichment then disable the o2 you will no longer get the 105% increase because there is no longer a change from closed to open loop. Thus your maps where created with the 105 increase and now you will not get it, so you will be 5% lean in what used to be the open loop area of the map. Make sense?? Love the way the bike runs with no o2 sensor, period....Way smoother!!!!!! | |||
| Xl_cheese |
Unless of course Ve analyser takes open loop enrichment into account in it's calculations... | |||
| Id073897 |
you will no longer get the 105% increase because there is no longer a change from closed to open loop ... or, in other words, you will always have the OL enrichment activated. It's the region that triggers enrichment, not the change. As the enrichment is constant, the ve analyzer doesn't need to take it into account. Regards, Gunter | |||
| Buplaux |
I may be wrong but after I disabled the o2 I noticed that the open loop enrichment did not happen. I think I might go out and log a little to verify this. BTW I am pretty sure the ve anal. does take into account the open loop enrichment. This is why it would be a problem when you no long get the OLE. You will be 5% lean. What are your thoughts Gunter? | |||
| Id073897 |
I may be wrong but after I disabled the o2 I noticed that the open loop enrichment did not happen. I think I might go out and log a little to verify this. BTW I am pretty sure the ve anal. does take into account the open loop enrichment. This is why it would be a problem when you no long get the OLE. You will be 5% lean. You are wrong. I've got more than 50 logs proving my statement. The veAnalyzer doesn't have to care about OL enrichment, as long as it's constant. The essential part is O2. Whatever enrichment might apply, as long as it's constant, the cell value will be adjusted according O2. Regards, Gunter | |||
| Mmcn49 |
This morning, prior to reading posts from the 18th & 19th, I made another Data Log run. Before starting, AFV maximum & minimum were set to 100%, unchecked open loop learn enable, and o2 mid point set to .54V, rich .58V and lean .50V. Did the VE analysis with Closed Loop set at 14.4%. Entered the new rear map values, calculated %age differences for front map and entered those values. I also decided to leave the square idle enrichment values in place, 72 rear and 76 front. After making the map changes I left the AFV, o2 voltage and OL learn enable changes in place, and took a 40-50 mile ride. The bike ran well, (very smooth and plenty of power). I live about 200 feet above sea level. The ride I took never exceeded 1000 feet. Tomorrow I plan on taking a ride into the mountains and see how the bike performs at higher elevations. Where I’m planning to ride is over 5100 feet. Based on how the bike performed today, I have high hopes that climbing to 5100 feet will not be an issue. Questions for those who know more than I, (everyone): Will the ECM read the new o2 sensor voltage range, .50-.54-.58V? If so, with AFV min & max set at 100%, and OL learn enable unchecked, even if the ECM can use the new o2 readings, will it be able to apply changes to the front & rear fuel maps in Closed Loop? | |||
| Buplaux |
yep I was wrong. You still get open loop enrichment with the o2 disabled. You get it over the entire map minus idle and wot zones. This is actually good from my point of view. makes it a lot easier to get smooth maps with the o2 disabled. After a lot of data logging to get consistent maps I disabled o2 sensor & set the open enrichment to 110% & wot enrichment to 115%. The bike is really happy with these settings and very smooth. Jardine, kn, swiss cheese airbox. Btw for those that are making the front mapped based off of % of rear, my actual logging on the front came out a lot different from the % method I initially used. It was worth the $25 o2 sensor and bung. Regards, Jeff | |||
| Turk |
If so, with AFV min & max set at 100%, and OL learn enable unchecked, even if the ECM can use the new o2 readings, will it be able to apply changes to the front & rear fuel maps in Closed Loop? I don't think you should be disabling the AFV like that. It's there for a reason, to account for air pressure differences among other things so let it do that. Also, open loop learning is there for a reason and if it's learning in open loop you've got serious problems with your map that need to be addressed, not just turning off the learning. It doesn't work anything like Xoptimedrsx seems to think it does. The internet is a bad place to come to try to learn how to tune a bike properly... | |||
| Buplaux |
open loop learn enable is a % richer or leaner applied to the entire open loop area of the map based on what is actually sensed or learned by the lcl(learning portion of the closed loop). So if the lcl portion of the map was to rich and the open loop perfect the ecm will lean down the closed loop instantly with ego correction and after enough steady riding in lcl will lean down the open loop portion by applying a % less than 100% to the open loop portion. This will make a perfect open loop portion of the map lean. This is how I understand it anyways after literally hours and hours and sleepless nights thinking about this system. HTH | |||
| Mmcn49 |
I’ve made many data logs and ECM changes over the past 7 or 8 months trying to find the perfect fuel & timing map. This all started because my new 07 Buell ran like crap most of the time. WOT and 5500+ RPM were fine, everywhere else it ran rough. I’ve been able to pretty much smooth things out, but have/had two remaining issues: 1. The ECM applying corrections to the new maps. 2. The new front and rear cylinder fuel maps are not balanced. It was recently explained to me, that if changes are made in closed loop, and those changes don’t yield a 14.7:1 AFR, the ECM will scale those changes out. It will remember how much it scaled the new values and it will apply those changes across the whole fuel map. We all know our bikes run much better with a closed loop AFR of 14.4:1 or 14.2:1. Until I started limiting the AFV range, my bike would start running somewhat rough within 20 or 30 miles of making map changes. It never ran as poorly as it did with the stock maps, but if would not run as good as it did before the ECM started applying changes. Today I went for a 150 mile ride with the AFV min & max set at 100%. The bike ran just as good at the end of the ride as it did at the beginning. It appears that setting AFV min & max to 100% greatly limits or turns off the ECM’s scaling ability. I also climbed to 5,240 feet. The bike had plenty of power all the way up. It ran fine and showed no signs of running rich. Idling at the top I smelled the exhaust before shutting it off, it did not smell like it was running rich. After shutting it off for 45 minutes, it started right up. Next weekend I’m going to shut off the o2 sensor and see how the bike runs with the AFV min & max at 100%. The bike runs good, but after many front map %age changes, cell smoothing, and outright guessing I can’t quite balance the front and rear cylinder fuel maps. The next step will be to put an o2 sensor in the front pipe and data log the front cylinder. | |||
| Buplaux |
mmcn49 very well said... If you get the fuel maps right I see very little need for the o2. Put in the front bung and be done with it. It is worth every penny. Then disable the damn thing and forget it. | |||
| Turk |
open loop learn enable is a % richer or leaner applied to the entire open loop area of the map based on what is actually sensed or learned by the lcl(learning portion of the closed loop). No, that is the AFV. AFV is set through both closed and open loop learning. Open loop learn enable is allowing open loop learning to take place. With that disabled, AFV will be set purely based on closed loop learning. | |||
| Turk |
If you get the fuel maps right I see very little need for the o2. Put in the front bung and be done with it. It is worth every penny. Then disable the damn thing and forget it. I guess, they're your bikes so do as you see fit but you are severely underestimating the responsibility of an O2 sensor. | |||
| Buplaux |
I didn't think anything could happen on its own in open loop that closed loop learning didn't command and apply it to open and closed in the form of afv? I am by no means a expert on this system. When logging after leaving closed loop the ego correction "flat lines" on the last number that it was on when it left closed loop. It stays flat lined as long as it is in open loop. This tells me that it is not applying any correction in this zone and it doesn't appear that it ever would. This is not argumentative in any way. I love talking about this system and learning. Gunter any thoughts to add? | |||
| Id073897 |
Turk's description is correct, as far as I know. It stays flat lined as long as it is in open loop. This tells me that it is not applying any correction in this zone and it doesn't appear that it ever would. OL learn affects AFV directly, not EGO correction. EGO correction is limited to CL, whereas AFV is limited to OL (unless WOT). Regards, Gunter | |||
| Mmcn49 |
I guess, they're your bikes so do as you see fit but you are severely underestimating the responsibility of an O2 sensor. ------------------------------------------- Not according to Adam Wade. On page 61, in his Motorcycle Fuel Injection Handbook he flatly states that a narrowband o2 sensor's only purpose is to keep closed loop AFR at 14.7:1 for emissions purposes. It does nothing in helping the bike run smooth. Below is a narrowband o2 sensor output voltage vs. AFR graph from Musclecross Forum. Looking at the graph, you can get a sense of how limited narrowband o2 sensors really are. As mentioned in my last post: “It was recently explained to me, that if changes are made in closed loop, and those changes don’t yield a 14.7:1 AFR, the ECM will scale those changes out. It will remember how much it scaled the new values and it will apply those changes across the whole fuel map. It makes and applies those changes by measuring AFR in closed loop with the o2 sensor, comparing the o2 AFR reading to the fuel map cell values and adjusting injector pulse width in those cells}} to achieve 14.7:1. Like Buplaux I’m no expert. I’ve read, studied, asked questions, data logged and experimented to reach the point I’m at now. Based on the above, I’ve concluded: 1. The only way to get a smooth running Buell is to data log and input the new data log cell values. 2. Make Buplaux’s timing map corrections. 3. That limiting AFV range to 100% will not affect engine operation from sea level up to 5,200 feet. 4. AFV min & max must be set to 100% to prevent the new cell values and fuel maps from becoming skewed. Next week I’ll see what happens with the 02 sensor turned off.
}} |
Narrowband o2