| Author | Message | ||
     M1combat |
As I understand it... The O2 changes from low to high voltage (and back) as the AFR crosses a specific threshold. This signals the ECM to add or remove fuel. Sounds like a good system to keep the AFR where Buell wants it. I don't want it where Buell does  . Does anyone know of an O2 sensor that will change from low to high at a different AFR? Preferably about 14:1. If we could source one of those then many of our tuning problems would be solved. | ||
| M1combat |
Ooops ... Could someone please move this to the proper area? I was checking my torque spec thread and got all excited and just posted this one where I was at . | ||
| Blake |
Done. Interesting idea, operate on the sensor rather than the logic circuitry. If only the darn sensors were linear, you could just bump up the voltage a bit to make the logic circuitry think the engine was running a bit leaner than it actually is. | ||
| Alessio66xb12r |
"this signal the ecm to add or remove fuel..." it signal the ecm to add or remove fuel on the whole map ("learning mode" during closed loop) the 02 sensor during closed loop operation ,2000 to 4000 rpms and light throttle % , will change ( if in need) the AFV to compensate rich or lean conditions.this method is very approximative because a rich condition ( for example) at 2500 rpm 10% throttle does'nt necessary means rich condition at 5000 rpm full throttle.to change 14.7(race kit ecm) with 14.0 does'nt make it better just richer. for me the perfect 02 sensor is the "Removed" one , also the "Unplugged" works fine on my set. to solve a lot of tuning problems : 1) get ecmspy and the cable 2) set the AFV to 100 3) unplug the sensor 4) play with the fuel map and find your optimal set-up to tune your bike you can use ecmspy to collect data ,while riding (GREAT), from the sensor (when connected) so you can see where add or subtract fuel. bye Alessio | ||
| Id073897 |
I would agree with Alessio. Why fiddle around with sensor data, when it's so much easier to change fuel map bins? AFV is calculated from fuel map values in the learning mode region, which is small compared to the full map. If this region is set up for an AFR 14.7, the AFV should be okay in open loop. If not, either OL enrichment or adjusted OL fuel map bins should fix it. Regards, Gunter | ||
| M1combat |
What if you live in the mountains? I need the ECM to adjust like it wants to... but be tuned like I want it. As I understand it... when you add fuel in an area where the ECM will adjust the AFV it will pull fuel out to get back to the AFR it likes to see via the O2 sensor. If that happens then you'll be lean when you go into the open loop portions of the map. It all works off the V from the O2 sensor. I'm sure a good, stable tune can be had if you unplug the O2 but then it won't adjust for elevation changes. Another issue I had with adjusting the map (via a PCIII) is that the bike would stumble as it transitioned from close dto open loop (like if I was riding near 4500RPM slowly crossing that RPM boundary. That said... It was a PCIII... So... How can the closed loop portion of the map be adjusted without having the O2 sensor signal the ECM to adjust the AFV up or down to get back to where the O2 wants the AFR? That screws with the rest of the map. Or is the recommendation still to unplug the O2 even if I live in the mountains? | ||
| Id073897 |
You're right, the use of the O2 sensor is a bit of dilemma. You will need an O2 sensor for monitoring mixture, but you will not want the ECM adjust fuel because this is what you want to do manually by adjusting the map. AFV is calculated from O2 sensor transitions, so what needs to be done is to adjust the learning mode region to a stoich mixture that less transitions occur. So the usual approach is to disconnect/disable the stock O2 sensor and put the bike on the dyno to monitor fuel with the wideband equipment provided. Or to swap the narrowband sensor with a wideband sensor and connect it to the ECM but disable closed loop. Thus the ECM monitors and logs O2 voltages, but closed loop calculations are disabled. As soon as the tuning is finished, you will connect/enable the stock O2 sensor again, that the ECM will get it's ability back to cope with air pressure changes. On the track, where maximum control of fuel is desired, keeping the ECM in open loop might be a better approach. All that needs to be done is to check air pressure and then adjust AFV accordingly. IMO thats a very elegant and easy way to "rejet" the engine. Now, speaking about piggyback systems. All systems that modify fuel in learning mode area without providing as suitable O2 signal have to fail given the way AFV is calculated. On the other hand, systems working in open loop only, might succeed. Regards, Gunter | ||
| M1combat |
Understood . So you retune the closed loop to give you an AFR that the O2/ECM are ok with so that it doesn't pull the open loop areas off with the AFV, but you can still tune the open loop areas to whatever you need them to be. Of course on the track you run a map that's "fully tuned" and "rejet" by manually setting the AFV. Excellent. Sounds like it'll do what I want then as long as I can get the full closed loop area tuned to provide a stoichiometric mix. Do you know roughly how much elevation each point of AFV change corresponds to? I would assume you can't say for sure due to local barometric conditions but do you have an estimate? I'm asking because I'm at 5K feet. If I just reset the AFV to 100 and start tuning I'll have a ways to go in each cell. I'm sure it wouldn't be an issue if I was closer to 2000' (that's where an AFV of 100 is calibrated at yes?). Thanks for your answers Gunter and thanks a LOT for the work you've done. All that said... Would an O2 that changed voltage at a different AFV solve the issue as well given the volumetric efficiency curve was close to stock (especially through the closed loop area)? "On the other hand, systems working in open loop only, might succeed. " And that's exactly what I found with a PCIII. As soon as I zeroed the closed loop portion of the map, let the AFV settle, then tuned the open loop portions it ran well in the open loop areas. The trouble was that a power difference was quite noticeable when it would transition from closed to open loop slowly by just rolling on enough throttle to get it to creep past 4500 RPM. (Message edited by m1combat on March 14, 2008) | ||
| Alessio66xb12r |
for me there is something wrong at the origin : buell's "learning mode" enrich or not a whole map based only on datas coming from a small part of the rpm/throttle combinations . it works with all stock components (stock map is developed for a stock bike).when you start to make mods ...air filters , mufflers , air boxes or big bores you must adapt the system to the new set-up. Gunter ..."you will need 02 sensor for monitoring...but you will not want.." RIGHT for this i think the best solution remain to remove the stock 02 sensor and fit in his place a sensor connected to a display ( not connected to the ecm ) more accurate as possible so you can read on display your conditions while riding. i know there are some aftermarket company selling this sensor/display set-up. M1 "how can the closed loop portion..." no 02 sensor = no "closed loop" portion you will run only on the map from 0 to 7000 and from 0% to 100% throttle. once you find the right map for your set-up you can use ecmspy to handle the AFV and enrich or not the whole map ( like the learning mode in closed loop does) depending your air density , your weather , ....your mountains exc. exc. bye Alessio | ||
| Id073897 |
for me there is something wrong at the origin : buell's "learning mode" enrich or not a whole map based only on datas coming from a small part of the rpm/throttle combinations . it works with all stock components (stock map is developed for a stock bike).when you start to make mods ...air filters , mufflers , air boxes or big bores you must adapt the system to the new set-up. But you have to notice, what this was meant for originally: to handle air pressure changes. These apply for the whole map. The learning mode region does make sense to me: it's mid-load and mid-rpm and comparably small. This prevents you from O2 sensor "misreadings" as O2 voltage changes with sensor temperature. The second constraint is that you don't want the closed loop interrupted by other enrichments, e.g. acceleration enrichment, to set up some "long-term" data (long-term in comparison to short-term as would be a single crankshaft rev.) And the third reason is IMO, that you need a to find a region, that is used frequently in everyday ride. The learning region as currently used, applies to all these constraints. The smaller the learning mode region, the less work has to be done to find the required fuel map values that lead to the desired results. So just a few map bins had to be adjusted (more) precisely to result in a stoich mixture (or whatever the developers targeted at). This makes changes in further development faster and cheaper. I'm sure it was never intended to support variant configurations (airbox, muffler etc.) with that learning feature, but just a side effect. How would alternatives look like? The standard used in many bikes is to install one or more pressure sensors (which BTW could be done with the ECM also). More sensors means more maintenance and is more error prone. More sensors that might fail, more plugs that might loosen. Another configuration would be changing the setup to speed-density instead of alpha-N. But what happens, if the MAP sensor starts floating? It's not as easily calibrated ("reseted") as the TPS. In my oppinion, once set up correctly, AFV is your best friend. Changing configuration always requires adjusting the map. This is valid for all alpha-N based FI. Relying on AFV, which is not intended to cope with that, will never give best results. Regards, Gunter | ||
| Bombardier |
There was a post about the relationship between the head temp sensor and the o2 sensor with regards to when the engine was hot enough to apply inputs from the o2 sensor. Is this a fact and if so how can the input setting be adjusted to allow o2 voltages at a lower temp? |
