| Author | Message | ||
     Blake |
Paul, I have photos of "The Destructor" in action. Thanks for the generous donation.  | ||
| Steveshakeshaft |
This is an interesting topic, a Google search on "oxygen sensor response time" brings 000's of results. I was surprised how "slow" the resonse time of an automotive Oxygen sensor apparently is. Around 0.1 sec is about right. From- http://www.forparts.com "A properly functioning oxygen sensor will exhibit a rapidly fluctuating voltage signal alternating between approximately .2 and .8 volts in response to varying residual oxygen content in the exhaust stream. Look to your scope's time reference line for a desired lean-to-rich and rich-to-lean time of less than 300 milliseconds. A response time greater than 300ms. means that the sensor should be replaced." And- "The Bosch LSU 4 wideband oxygen sensor has a response time of less than 100 milliseconds to changes in the air/fuel mixture" Regards Steve www.ukbeg.com steve_s@ukbeg.com (Message edited by steveshakeshaft on February 04, 2005) (Message edited by steveshakeshaft on February 04, 2005) | ||
| Court |
>>>"Where IV art thou?" Now THAT's funny. | ||
| Spatten1  |
MAF is without a doubt the most accurate system, but I don't know how it would work with the pulses of a V-Twin. If memory from my previous life as a mechanic serves me correctly, the basic concept is that a "hotwire" is supplied voltage to keep it at a specific temperature. The more air that flows past the hotwire the more voltage is required to keep it at the specified temperature. The mass air flow system measures the mass of air, not the volume. It is very accurate regardless of barometric pressure, temperature, or anything else. What is so wonderful about the system is that if you know the actual mass of air then you know the actual mass of fuel required, before the combustion event. Most modern automobiles would not meet their strict emissions standards (much stricter than motorcycles) without them. In fact, by the Early 90’s most of the fleet vehicles that I worked on had made the transition to MAF. I saw the post about the dirty MAF and can’t help but wonder whether he had a K&N or an air leak upstream of the sensor. I haven’t personally heard about that being a really common problem. Again, I can definitely see the complications to using this type of system in a V-Twin, especially with a short intake tract. That does not detract from the fact that most auto manufacturers moved beyond speed density 10 years ago. One more question, do we know how many other existing motorcycles already comply with 2008 standards? | ||
| Wyckedflesh  |
MAF is without a doubt the most accurate system, but I don't know how it would work with the pulses of a V-Twin. If memory from my previous life as a mechanic serves me correctly, the basic concept is that a "hotwire" is supplied voltage to keep it at a specific temperature. Gee, that sounds an AWEFUL lot like what the IAT (Intake Air Temperature) Sensor does. | ||
| Reepicheep  |
Wow! It's going to be nuclear powered? Or is it just that "50 thou a year will buy a lot of beer". Aaaack! Make the 80's music in my head stop! | ||
| Spatten1 |
Why would it not use a similar technology as the IAT? It does, however provide different information than the IAT. Can we please stay on topic rather than just look for an opening for ridicule? I'd love to hear something constructive. | ||
| Madbandito  |
For those of us who don't speak "Jive", I hope that all this translates as sumthin in our future thats just "Gooder N' Gum"  | ||
| Spatten1 |
I don't get the inside lingo either, but having the first Nuclear powered bike come from the US would be killer. We could get the fuel from N. Korea or Pakistan. You would only need to refuel every couple of years. | ||
| Sarodude |
Blake- The .1 seconds was WRT response of the sensor and not of the systems reading it - if that's what you were wondering. Again, my memory is hazy. I can't recall if this is what O2 sensors generally do or if this was a special O2 sensor or what. Anybody here read Racecar Engineering? I used to lots a few years back but haven't really loked at it recently. They would have lotsa neat little gems in there. It was in the early 90's when they had a little blurb on electromagnetic valves. Back then, the prototypes were fetching something like $10,000 per valve. Man, I loved that magazine. Too bad it was over $10 an issue. -Saro | ||
| Court |
>>>>It's going to be nuclear powered? The word, boys and girls, is ATOMIC. Standby for more..... | ||
| Mikej  |
Will "it" be ready for test rides in April when the hat=>slingbag promotion advances? | ||
| Wyckedflesh |
And just how different do you think the info that both sensors provide is? Both read changes in temperature within an airflow. Nothing more, nothing less. Just because the MAF itself is generally within the direct airflow and the IAT is not "as far as we can determine" in the direct line of airflow doesn't mean they both don't provide the same information. Without a clear airbox and a smoker to see how the airflow within the airbox moves its hard to determine if the IAT sits in the normal line of airflow or if it sits in an eddy of current that effects a temp change from the airflow past the eddy. Either way, it reads a temperature change and the ECM acts upon that change. | ||
| M1combat  |
I thought it was Nukeular?  | ||
| Spatten1 |
The MAF uses the voltage variations to determine how many molecules have contacted the wire. The shape of the MAF and the wire placement are critical for calculation of the entire mass of air molecules passing through the device. It is calculated, but extremely accurate. This is opposed to the "trap door" type system use on the 1980's GPZ1000s that measured the volume of air pushing the spring loaded door open. It could not differentiate the density of the air, thus the mass. | ||
| Wyckedflesh |
I understand that Spatten, but at the time those MAF systems were run the ECM couldn't do as many calculations per second as the newer ECM's can. So "back in the day" the extra precision was vital to accurate fuel management. With the current ECM's ability to process the data, the variations in the IAT are plenty sufficient. Your right about the Trapdoor systems, Audi/Porsche/VW and their JetronicFI which was the basis for the GPZ1000's had problems dealing with altitude changes and temp changes until they later added extra sensors. My last Jetronic FI system I dealt with was on a 1980 Audi 5000 and getting the tension on the platten "just right" was a major pain. And that car hated going from sea level at the beach up to 4000 feat in the span of the same afternoon. | ||
| Steve_a |
Mass-airflow-measuring injection systems are wonderful for easy calibration, and for the ability to compensate for many tuning changes. After all, why not measure the variable you're most directly interested in (intake air mass) rather than try to infer it from other variables? GM switched in large part because they didn't have the engineering man-years to calibrate every car/engine combination in their line using the speed-density systems of the time. However, actually measuring mass-air flow can be difficult, particularly with a small number of cylinders and a highly tuned engine. Multiple pulses and flow reversals past the sensor confuse the matter. A really big airbox is very helpful, if not essential. In any case, racing applications are almost never MAF-based, as alpha-N gives all the power that can be had with a simpler installation. Inference of air mass is good enough for racing and a lot easier to implement. I'm not aware of any current motorcycle applications that use MAF-based systems. | ||
| Opto |
it's really easy for an O2 sensor system to adjust for altitude. If the air density is lower, the mixture will get rich. The O2 sensor sees this, and reduces fuel until the mixture is right. This is true if the bike is running in closed loop, but if you are going hard with a bunch of mates the ecm is mostly running in open loop and the O2 sensor is not used. And believe me this happens at an insanely fast pace...remember that at 3000 rpm, the O2 sensor is getting a mixture reading 50 times per second! Much faster than most O2 sensors can respond. The O2 sensor system can also adjust for anything else in the engine that might be causing an imperfect burn, which a speed-density or MAP system can't do. A speed density system can also run an O2 sensor, many do. On an engine that is solidly in the center of the adjustment range of the ECU, you can get tremendous accuracy of fueling on an O2 sensor system with a high level ECU and the right sensors. That is an accurate statement. Since it has tach and speedo inputs and a timer, it certainly can tell whether it (the engine) is loaded or not. Some export models are shipped with the vehicle speed sensor disconnected. Craig, My take is that you are becoming emotional and combative. He displayed some sarcasm and said sorry and apologized, the rest was him stating the truth as he sees it. Please consider that the great thing about this kind of discussion is that right or wrong in our understanding of the issue at the outset, we all hopefully learn a lot or at least a little in the course of the discussion. That is an accurate statement. | ||
| Craigster |
Opto, Quite the level head you have there. Thank you for clearing things up in a factual manner. | ||
| Reepicheep |
Lets also not get stuck in static analysis. You not only know what the engine settings are right now, but you know where they were before. The ECU knows the throttle was just whacked, because it just saw the throttle whacked. It knows the bike is trying to deliver more power, because it it was just asked to deliver more power by the operator. That covers the "accelerator pump" part of the equation, and gives you plenty of margin to give the O2 sensor time to joing the party. You can't change temperature, pressure, and humiditiy instantaneously unless you get dropped from a C130. I don't know why you would not just calculate these things from instantaneous and average baseline measurements of other inputs available (temperature, 02 sensor behavior, even internal dyno plot curve fitting, whatever). If you have the CPU cycles to spare, why mess with the unnecessary sensors... Atomic huh? Are we going to be running technology derived from two stroke's or diesels? I love *my* tractor motor. Whatever it does, it works great, aside from the fact that my xb9sx kept wanting to wheelie on the way into work today. That thing sure loves cold air. | ||
| M1combat |
I've gained >2K' elevation in a matter of a minute or less multiple times at a highly opened throttle position and 4-6K RPMs. No problems at all. NONE. The bike never felt rich for a moment. I lost power of course, but I think more due to the elevation than anything wrong with the ECM. "After all, why not measure the variable you're most directly interested in (intake air mass) rather than try to infer it from other variables? " Why do you figure we're most interested in air mass? Aren't we more interested in the resultant a/f mixture? I think most of this discussion is splitting hairs anyway. When riding up a hill do you REALLY figure that it's a common occurrence to not drop into the learning mode (I can never remember if that's open or closed) in a matter of miles? You don't think the ECM is working on adjusting the AFV at this point? Has anyone ever actually had a problem with the way their ECM works (aside from being a four year old and revving a cold engine until the plugs foul I mean)? Even if the bike DOES get rich "er" going up in elevation when you're "having fun", do you really figure that it surpasses stoichiometric mix? Wouldn't the fact that the mix is getting CLOSER to stoichiometric be a GOOD thing? Wouldn't you actually end up with MORE power (compared to a system that retains it's EPA controlled mix) at the top due to the NONADJUSTMENT while climbing and being closer to the optimum mix? | ||
| Glitch  |
Set parameters = Open Loop "Learning" = Closed Loop Never a problem. As for your other questions, I'm still learning this stuff. This is the first FI bike I've owned. | ||
| M1combat |
Thanks. I doubt I'll remember it though. To me it should be the open loop that learns because only an open mind learns... If you have a better way to keep myself reminded about which way it is, I'm all ears . I suppose thinking about it like an electronic tech would give me the required help. An open circuit won't let anything through, whereas a closed on will... There I have it . | ||
| Bud  |
Very interesting reading indeed. Now for the future, I’m thinking blank ecm’s with a vin dedicated software package And dealer downloadable “race ecm” software packages Fast and simple updateable ,just hook it up to the digitech And download a upgrade Bmc able to see where and how every bike is doing ( counters in the ecm ) To get ahead if any problems occur on a regular base Just thinking out off the box | ||
| Reepicheep |
You are dead on Combat. Open look is just the controller attached to an actuator, end of story. Closed loop has added a sensor on the other end of the actuator, and another wire that goes back around to the controller. Thus the wires have "closed the loop". And climbing thousands of feet in minutes is not enough. Even the crappy O2 sensors get more then one reading a second. How much altitude can you change in a second, versus the ability of the O2 sensor to detect mixture changes? I don't know, but i suspect that unless you are worried about how your buell does if dropped running from an airplane, you are good. | ||
| M1combat |
"How much altitude can you change in a second, versus the ability of the O2 sensor to detect mixture changes? " Well, The time doesn't matter if you are climbing/descending in the open loop mode. The A/F mix will change. I just think that it would be a VERY rare occurrence that you would be going through enough elevation change w/o dropping into the closed loop mode every once in a while for it to be an issue. I say that even when I live in the mountains and gain/lose alt. VERY regularly AND quickly and have had NO problems what so ever. "I don't know, but i suspect that unless you are worried about how your buell does if dropped running from an airplane, you are good." And dropped at WOT no less... Dammit Erik... When I want to drop my Buell from an airplane at WOT I want the damn A/F mix just right when it hits. Re-design the damn inferior ECM logic. What WERE you guys (and gals) thinking? Oh... Probably that I wouldn't be doing that... never mind. My bad. Carry on then. | ||
| Cataract2 |
Question for you guys. At what RPM is the bike in a closed loop vs. being in an open loop? This is of course assuming the bike is fully warmed. | ||
| M1combat |
I "think" it depends on throttle position. I'm sure at a slow and steady throttle opening that it would indeed change over to closed loop at a certain RPM, and I would GUESS that RPM to be somewhere around 4500. | ||
| 99buellx1  |
Idling - Open Loop Accelerating through 3rd to 5th gears - Open Loop Steady Load above 60mph(4,000+rpm)3rd to 5th gears - Open Loop Cruising, light load, 40-60mph(1500-4000rpm)3rd to 5th - Closed Loop Decelerating from 5th to 3rd gear - Open Loop If that helps. Craig (edited for closed loop rpm misprint) (Message edited by 99buellx1 on February 04, 2005) | ||
| M1combat |
I "think" it's in closed loop mode above 3K RPMs and light load. I say this because the procedure to re-calibrate the AFV calls for something above 3K. I'm probably wrong though... I don't have access to HD-Net . |
