moto

Having the ECU in the modes where the O2 sensor output exerts the least influence on the map will be best, especially if the base map is close. "I" won't be happy, but "one" might be. Again, it depends upon one's criteria for happiness. Irrespective of that, the O2 sensor inputs will damage performance less if the basemap is off substantially to begin with, especially toward the lean side. Depending on what voltage the ECU is looking for, it might even help. Does the 15% you mentioned refer to pulsewidth, or something else? You can't drop a prefab chip in and expect the engine to run right. It might, but fuel pressure, injector flow rates and the mapping itself (!) preclude that from being fool proof. Hey, that's great! Do you have any before and after dyno runs to show us? As far as your engine running reasonably well in closed loop mode, it is possible, especially if it was running very lean to begin with. Remember the example from before, with a .4% oxygen content and 1.9 to 11.4% CO? Remember how below ~1.0% CO engines can begin to misfire? Say the engine was running at somewhere around 1.0% CO or near but not at misfire. This figure could be somewhat lower for an engine with big displacement cylinders, like a Guzzi. So lets say we are under 1% CO. If we then use .4% O2 as our target, and barring any anomalies such as Cliff and I have uncovered, the CO will likely come up to around 1.9% before the ECU stops adjusting, which will yield a significant improvement in performance, but not as much as if we were to bring the CO to between 3 and 4%. It should be noted that this example is not likely work out exactly this way in real life, because I have not established the relationships between CO and HP nor between CO and O2 for a modern Guzzi, and therefore have to resort to using figures from another engine. I think my example nevertheless makes the point...

That's all fine and good, but he himself stated that he had not read nor understood all of the posts. Huh? Why would you want to skew the facts? And then to announce it so brazenly? Tsk, tsk. Well, on the other hand you are being honest, which I commend! Certainly in a rich mixture, there is little to no oxygen. But I can't recall anyone advocating the use of an O2 sensor in a rich mixture. There is definitely measurable CO during a lean condition. As stated before, you can get meaningful CO numbers all the way down to lean misfire. No, they can't measure mixture, but they might measure oxygen content. Then what is it measuring? Quite true! True. That depends on how many throttle positions are tuned, how many rpm steps are tuned, whether ignition timing is tuned, if the table positions are overlapping for timing and pulsewidth, whether cylinders are tuned individually, if any interpolation is used (if so, how much), and whether the tuning is CO or HP; if to CO, how the target is derived; if to HP, to what resolution (1 HP? .5 HP? .2 HP? Some other number?). True. Maybe we have widely disparate ideas of what the word optimal means in this context. Can you define what it means to you? Really? Where? What is this happiness based upon? I would like to see this quantified and qualified. Agreed. I think that if folks had a better undertanding of what they were really getting, didn't just buy DynoJet's "A/F ratio" related marketing hype hook, line and sinker, and knew that there was a better alternative, it would not be.

For performance (hp and/or economy), pretty much.

Did my reply explain sufficently?

Just to clarify, the stated 3-4% range referred to CO. Once again, if CO looked ideal to slightly high and O2 looked high, and blanket changes to pulsewidth made little difference to HP, a stagger issue would be highly probable. If that was not the case or solved, and O2 was still high, the ignition timing would likely need to be advanced. O2 will not tell anything about overadvanced ignition.

Well, Cliff has been touting the benefits of his ECU mainly upon the supposed merits of O2 sensor driven full closed loop operation, but, as far as I can tell, so far things have not progressed/deteriorated enough for anyone to threaten anyone else... I think you might not if you had actually at least read all of the posts, even if you did not understand all of it. In any case, the usefulness of sensors depends entirely upon the relevance of their outputs. How can O2 sensor output be relevant when it can be the same for a 1.9% to 11.4% CO with 1% CO approaching lean misfire, 11% approaching rich misfire and 3-4% being ideal for power, and when it can also read all the way from .2% to 4% O2 at best power. Right, but how do you "interpret" a signal that, at best, bears a loose relationship to mixture stength? Don't get me wrong, as stated before, it can be usefully interpreted, but not by an ECU, which can't monitor other gas contents, stagger, or HP. What knowlege would that be? As mentioned before, ECUs correct just fine by monitoring throttle position, rpm, pressures, temperatures, etc., certainly better than they could based on O2 sensor voltages.

"Elite performance engines" could be construed by someone to mean "racing". To clarify, I actually have reservations against any O2 sensor controlled full closed loop application. What I meant to imply is that in instances where the O2 sensor is used to optimize cat efficiency/longevity (as in most) the sensor is either not a player or only an incidental player in controlling pulsewidth for the purpose of producing HP or mileage (or "performance"). If that means you would like to withdraw your withdrawal, I will humbly apologize for the confusion...

If the customer's previous run was corrected, and the one presented by the tuner was not, because he realized that he lost the customer some power and wanted to cover that up, I think that's scandalous. The customer deserves to be able to compare apples to apples. If both runs are corrected he can do that. As a matter of fact, for both the tuner and the customer that's about the only way to reasonably compare data. What else can you do? To expect that somehow all runs and tuning can or should be complete before any changes in ambient conditions occur is a bit of a stretch. Not quite, as the correction factors work pretty darned well on carbureted engines as well. Once again I can provide as much data as is required from as many bikes as you like, and you can see for yourself. Well, on an injected engine, that depends on the ECUs temp compensation, which seems to be quite good nowadays. On a carbureted engine you can certainly tune to a particular temp, say the average you might encounter. That doesn't mean that tuning a carbureted engine is not useful. It just means that under some of the conditions the tuning will be suboptimal. Does that mean one should not tune?

The answer to all of this is catalytic converters. They need the O2 content to be in a particular range to function. Most closed loop ECUs only change the pulsewidth from the basemap based on O2 sensor input when the O2 content begins to go out of the range of where the cat will work efficiently as well as survive. The rest of the time, the ECU is just working off of the basemap with pulsewidth corrections based on TP, rpm, temps, pressures, etc. The map is never permanently modified by O2 sensor input, nor is the ECU ever running in a full closed loop mode, where pulsewidth is fully controlled by O2 sensor voltage. O2 sensors have to be reasonable accurate to do this job. That's why I am, aside fom the anomalies Cliff and I have uncovered, not essentially challenging the accuracy of O2 sensors. Once again, I am challenging the validity of the use of O2 sensor output-voltage in being sufficiently related to mixture strength to warrant using it to control pulsewidth in a full closed loop ECU for a high performance application, or in "Autotuning".

What's FUD? I'm sorry if I'm being dense, but I don't know what that is. I think in the case of DynoJet you would be correct. In the case of Factory, I would say less so, especially on this particular topic. If you are referring to Factory, they do not make the bench used in their gas analyzer. As mentioned before, that is manufactured by Andros. However, Factory's pump, filtration, sample cooling and water separation make it superior to anything else I've seen so far. I think gaining a "competitive advantage" is a byproduct of Marc's trying to solve some of the problems he encountered as a tuner. I see nothing wrong with producing a superior product equating a "competitive advantage". I think it should be that way, although it certainly isn't always. It seems that hype and repetition give the best "competitive advantage" in this day and age. Good! This is astep in the right direction. However, I don't recall anyone saying that O2 readings (or O2 sensors) are useless. I did say, and I think Marc is saying, that it's not nearly as effective to tune pulsewidth to an O2 content as it is to tune to HP or CO content. Could be true, but the readings I and everyone else who is using an EGA is getting are not "anomalies". The so far most effective tuning methodology would simply not work if all or most of the O2 data was anomolous. Once again the methodology, as follows: You can use CO to rough in. You tune pulsewidth to best power. If the gas readings are for more than one cylinder at once, you look at CO and O2 content next. If the CO and O2 look high, you might have a stagger issue. You then fix that by whatever method you have available. Once that is fixed, CO and O2 will have dropped, while HP will have gone up. If O2 is then still high, then you likely need more advance. If you had one EGA takeoff per cylinder you can go straight to looking at timing, as (obviously) O2 will not tell you about stagger in this instance. If HC is high, you might need less. Once again, HP is the arbiter. Once you've actually got fuel and ignition right, CO2 will look good. At this stage, you are done, no matter what the O2 readings are. Of course this all assumes an engine in mechanically good order with no ignition problems, etc. As you can see, there would be a problem if you went straight to O2 content. I am firmly convinced that the readings I am getting are the actual O2 content. I'm also conviced that the readings most people are getting with their wideband, control box equipped, 5-wire O2 sensors are accurate as well. What I am arguing is not the accuracy of these devices (although the anomalies are disturbing), but whether or not O2 content is a proper indicator of mixture strength, to the extent that it has a place controlling an ECU in a performance application. Again, is something untrue until everyone and their mother is talking about it? Very cool. There was some interesting reading here.

I can say from my own experience that SAE correction factors work quite well. I have many different runs taken on different days on the same motorcycle with the only change being in atmospheric conditions, and the corrected graphs are nearly indistinguishable (I'm willing to post as needed). Atmospheric conditions can easily change from one run to the next, and if there were no correction factors you could not tell at all what you were doing. You would be hard pressed to tell whether or not the horsepower you gained or lost was due to the parts you just bolted on or due to the change in atmospheric conditions. I used to do a fresh run in the morning to get a new baseline whenever I had to stop tuning on a previous day and I usually still do, but how well runs repeat with the correction factors has made it something I do because I'm persnickety, not because I have to.

The author of those links is not hiding anywhere and has no desire to remain anonymous. I simply don't know how to get the pages to come up within the context of the website, as it uses frames. I first looked at the page properties and copied the URLs. Just now I tried to use the main page adress in front of the extensions instead and it works just fine, although the pages are still not brought up in context: http://www.factorypro.com/dyno/dyno_error.html, http://www.factorypro.com/dyno/4gasEGAvso2sensor.html, and http://www.factorypro.com/dyno/dyno_locs.html. To say that Marc Salvisberg doesn't know what he's talking about and/or has an agenda to mislead is a bit of a stretch. Sure his writing style sounds agressive. I don't agree with it either, but that does not make what he is saying here untrue. Think about the magnitude of what he is up against. People tend to believe the falsehood that is repeated most frequently and with the most marketing glitz. Think about how difficult it is to change people's minds about anything, no matter how wrong it might be. I can undertstand Marc's frustration coming out in the written word and I can forgive him for that.

Are there really people who have had great results with a tuning link?

You can't expect much for $200, no matter who the tuner is or what they might claim, unless they are for some reason willing to work for pennies on the dollar...

Even ignoring what was written there, one only needs to look at the charts provided to come to one's own conclusions. Excellent! If you spend enough time with a dyno, with either an analyzer or a wideband O2 sensor, you will find them. You may be able to mount a CO meter of some kind on a bike, but it will not end up being very useful, as you need a certain amount of time at a given throttle position/rpm for CO to stabilize. What you are once again missing is that you can have mixture strengths that vary drastically at the one and the same O2 sensor output voltage, and corrections to the map won't happen until the voltage actually changes. That would definitely be a step in the right direction, but will still not solve the problem. You said that your ECU was making only minor changes and that they are directly related to O2 sensor voltage. I'm assuming that this is what you mean by saying that the ECU is producing a stable output. Once again, none of that is in dispute. But say that you determine what output voltages should be present at a particular throttle position/rpm. You then let the ECU look for that voltage at that throttle position/rpm while operating the motorcycle. As soon as that target voltage is reached, the ECU will make no further adjustments there until the voltage changes again. It will be just as you say, "stable". The problem is that the ECU can recieve that voltage at a myriad of mixture strengths (not pulsewidths, which as you say are directly related to sensor voltage). Say the target O2 content as established on a dyno is .4% (I don't know what that is in voltage)with a CO of say 3.2%, and the O2 content is currently .5% which is leaner and the threshold for getting to .4% O2 from the lean end is at 1.9% CO, the ECU will stop adjusting when it gets there. Or, lets say we currently have an O2 content of .3% and the threshold for getting to .4% from the rich end is at 11.4% CO, the ECU will once again stop adjusting once it gets there. It will not adjust again until it recieves some other voltage. Will the pulsewidth be related to voltage? Yes. Will the ECU appear "stable"? Yes. Will we have the mixture we wanted? Not likely. That you on one hand argue for the use of closed loop systems on the basis that pulsewidth and mixture (or how an engine runs) do not always correlate and therefore need correction, then on the other hand defend the use of O2 sensors by saying that pulsewidth and mixture (or how an engine runs) are 100% correlated is a bit contradictory. 41486[/snapback]

My dyno will measure hp and record EGA data at any throttle position/rpm combination the engine will run reasonable well at. I could see there being a problem at very low RPM and large throttle openings. Still, I was able to tune the Futura (for example) down to 1900 rpm at full throttle. What kind of dyno where you using?

On the Futura I was able to make minor repeatable changes in mixture that produced in the range of ±.1 to ±.2 hp. That's as good as it gets. My dyno will simply not repeat any better than that. Of course this tuning was done in real time. I think if I had to shut down, make changes to the map, reload and then re-test, the repeatability would not be as good.

How do you know where the stoiciometric ratio is? There is measurable CO produced until well below the point where engines start lean misfiring (somewhere around 1%). Could be, but if anything that just serves to further illustrate my point. Does that mean that the earth was not round until everyone agreed upon it? In any case, I'm not the only one who thinks that tuning to an O2 sensor voltage is a bad idea. See http://66.47.68.116/dyno/dyno_error.html and http://66.47.68.116/dyno/4gasEGAvso2sensor.html. For a list of tuners that are likely to agree, see http://66.47.68.116/dyno/dyno_locs.html. Whether or not the ECU makes accurate and repeatable pulsewidth adjustments based on the O2 sensor voltage is not in dispute, but whether or not these adjustments are appropriate is. You can say that you have a "monotonic and reproducible relationship between voltage and pulsewidth" and I will not argue one bit. But you can't say that you have a "monotonic and reproducible relationship between voltage and mixture" unless you have some way that is independent of O2 sensor voltage to verify that. If it's possible to have CO readings that vary from 1.9% all the way to 11.4% with a 0.4% O2 content, I can certainly see how your ECU would be prone to making infrequent changes confined to within 1% of a mean. I never suggested that there was some other better O2 sensor available that would solve the problem. Again, the same voltage can exist at many different mixture strengths. I'm not assuming that, but a modern ECU's compensation will be more valid than what can be expected with an O2 sensor driven closed loop system. I'm definitely a believer in native tuning and avoid the use of PCs whenever possible. You would adjust tappets & TPS and balance throttle bodies prior to tuning. Subsequent service would serve to get the engine to once again run optimally with the existing map, not cause it to run worse. Again, if so many different mixture strengths can exist for a given O2 sensor voltage, what good is it? Again, how do you know how lean it really is and how lean is appropriate? Why would you want to do that? Won't you loose power? How do you know how rich you've made it? Now that sounds like a plan! The fact that closed loop ECUs use O2 sensors is the problem, not how they do it. What are your expectations of an ECU?

True, but the results will be just as lacking in usefulness.

First, you tune for best power at the chosen throttle position/rpm step. If you then have a situation with a multicylinder engine with a common muffler and no individual EGA takeoffs where changes in pulsewidth common to all cylinders don't do much of anything to HP while the CO content is a somewhat high and O2 content is unusually high, you have a stagger issue. If you have control over it, you then try adding or subtracting pulsewidth on one cylinder at a time (if you don't have control over the map for each cylinder individually, it's worth flow & pattern testing and matching the injectors). If HP goes up, CO goes down, and O2 goes down, you are going in the right direction. If O2 is still high once you don't get anymore gains from adressing stagger, the engine can use more advance. Conversely, if HC looks high, you might have too much. As with pulsewidth tuning, the EGA readings help to point you in the right direction, but the final arbiter should be HP. You can use CO2 to check your work, as it will only look good when the proper pulsewidth and timing exist on all cylinders. I think it's a bad idea to try to tune advance without getting fuel right first.

I would definitely recommend tuning natively instead of using a PC.

How do they do that? Define "work well". If they don't provide data about whether the mixture is leaner or richer than it ought to be vs. close to optimal, then what good are they by themselves? If I can have CO readings that vary from 1.9% all the way to 11.4% with a 0.4% O2 content, how can O2 content be useful for the determination of mixture strength? Using the term "A/F ratio" is quite misleading in the first place, as it implies that it's relevant to something and that we can measure it. We can't measure it with an O2 sensor, but we might be able to measure it with a mass airflow sensor in conjunction with a fuel flowmeter. Even though this would be interesting data, how the engine is running is what is relevant, not what "A/F ratio" it supposedly has. Just to clarify, the Futura was tuned to within .1 to .2 hp of best power, not to within 1 hp. Assuming O2 content was actually meaningful, you could certainly tune each table position to max power via pulsewidth adjustments on each cylinder individually, then similarly tweak the ignition maps, and then try to use the final resultant O2 content at each table position as the target for the ECU. The problem is that this defeats the major percieved advantage of a closed loop ECU being able to tune itself without lengthy development on a dyno. Unfortunately it's a "garbage in, garbage out" scenario. What makes you think that an open loop system will lose 1 hp in six months? In my opinion, just adding an O2 sensor driven closed loop feature to an ECU that has had every fuel and ignition table position tuned to within .2 hp of best power stands to compromise performance more than leaving said ECU untouched for years. How?If we want to tune for economy rather than power, how do we do that? I suppose we could tune for power to establish where the engine runs best and then subtract a particular percentage of pulsewidth across the board. Or do we look to reduce the CO by some amount, or perhaps increase CO2? How do we then know that what we did is optimal? Can we do this across the board, or do we once again have to work on each table position individually? Wouldn't we also have to refine the ignition map to match our new fuel map? If we do subtract some fuel, we will certainly be loosing power. The rider has the throttle open commensurate with how much power he needs. So now he has less and opens the throttle further. Is he then still using less fuel than before? How much less? I suppose that could be determined on the dyno with a fuel flow meter. A closed loop ECU will not be able to do that by itself.

This would be a more valid question to ask if the engine was never accelerating at or passing through 12% throttle during acceleration. True, but how do you propose to measure fuel economy on a dyno? I suppose a flow meter could be used to measure how much fuel was being used at any given time. Is this akin to what you had in mind? It will take into account the clogged filter, as the airbox pressure is usually measured on the outlet side of the filter. The Futura ECU monitors the pressure in the airbox on the inlet, or theoretically wrong side of the filter, but also monitors manifold pressure which makes that ok. I think the oil slipping past the guides should mostly be irrelevant unless it's a large quantity, in which case you should be getting it fixed anyway. A closed loop ECU is also reactive to its inputs and not to how well the engine is actually running. You are ignoring the fact that a particular O2 sensor voltage does not automatically equate the best running engine. Sorry, that should have read "barometric".

Right, but by what criterion is the CO on the charts high? As far as I am concerned if I am producing the best power I can at that CO, then it is not high. Most motorcycles produce somewhere in the 3% to 5% range when tuned for power. High by what standard? According to the thousands of runs in my database for every type of motorcycle you can imagine in various states of tune 464ppm is on the low to average end except for when a cat is used. Right, that's why I'm wondering why you would talk about the CO shown in a chart for 100% throttle indicating rich mixtures during sedate driving. Also, see the CO for the 12% throttle run. The CO is comparable. I agree with you about most riding happening in the 20% and below throttle range. That's why the Futura's ECU was mapped on the front and rear cylinders individually to within ±.1 to ±.2 hp of best power at 100, 92, 82, 73, 62, 49, 39, 29, 22, 16, 10, 6, 4, 2, and 1% throttle (all throttle positions the ECU interpolates from) at the rpm steps shown in the charts (to where the engine would rev out at the smaller throttle openings). I'm more than happy to post the final runs (and the prior runs) for whatever throttle position anyone is interested in. I do think that at the larger throttle openings are where people have their fun, and that it's anticlimactic for the rider if the engine does not work well there. Plus, you've got to tune 100% throttle for bragging rights, no? Again, by what criterion is the mixture you see on the charts rich? And again, what are we tuning for? While it may be possible to improve mileage from best power by tuning specifically for that, in my opinion the increase in mileage will be small enough to make it a waste of time. I think most people ride a motorcycle primarily to have fun, not to achieve a certain fuel economy. An open loop ECU is certainly a reactive engine control. Modern open loop motorcyle ECUs monitor rpm, throttle position, engine temp, ambient temp, airbox pressure and relative humidity at least. With a proper map to make adjustments from, that's more accurate than using an O2 sensor. One thing that people forget is that O2 sensors were not introduced to help performance. They were introduced to assure the proper amount of oxygen in the exhaust gasses to keep a catalytic converter functioning optimally. Also, if stagger issues and timing have not been adressed in the base map, the ECU will meter the wrong amount of fuel in response to O2 sensor inputs. I'm not sure I follow. Can you clarify? In my opinion the lack of correlation between the proper amount of fuel and O2 sensor output makes this fairly irrelevant. The max brake rating is sort of a soft limit. The problem is heat. The cooler the brake, the more HP it can contain. What ends up happening is that at really high HP, the brake can't get rid of the heat quickly enough and then it will gradually let go. I think mine is probably good for 200 and some without additional cooling. How much more it could contain with extra cooling depends on how much was provided. Factory has models available for various applications and with various HP ratings. I'm sure that if they didn't have exactly what you needed, they would build it for you. For example, BRG's model MD-750 dyno is not listed on the Factory website, but will contain 1000HP. Tire slip has not been a problem for me, as the drive roller is slotted and the geometry of the setup causes the tire to "dig in" proportianately to how much HP is being produced.

My EGA responds to CO in ~2 seconds. However this relatively unimportant, because the delay is just set appropriately in the software and the particular rpm step is held until the traces level off. Agreed. I might expect that as well, except that some engines that you would not expect it of run best with seemingly high COs and that I find the correlation to "A/F ratio" to be irrelevant. Futura data including HC and CO2 for 100% throttle with pulsewidths tuned to within ±.1 to ±.2 hp of best power: Rpm hp ft/lbs CO HC CO2 O2 1904 15.97 43.91 4.2 668 11.3 0.5 2353 18.76 41.82 4.2 156 11.4 0.3 2655 24.75 48.90 4.2 182 11.5 0.3 3005 31.54 55.09 2.6 427 12.0 0.7 3306 36.13 57.39 3.3 647 11.4 0.7 3606 39.62 57.68 3.6 618 11.3 0.7 4004 44.01 57.68 3.7 596 11.3 0.8 4407 47.01 55.99 4.0 743 10.9 0.8 4906 55.59 59.48 2.7 459 11.7 1.0 5408 63.17 61.28 4.1 599 11.0 0.8 6007 67.56 59.08 4.7 174 11.0 0.2 6406 75.25 61.68 4.2 189 11.5 0.2 6810 85.08 65.66 3.8 596 11.1 0.4 7108 91.18 67.36 4.1 592 11.0 0.3 7408 95.49 67.66 5.1 599 10.6 0.2 7808 98.69 66.36 4.4 469 11.1 0.1 8309 103.29 65.26 3.2 428 11.7 0.2 8809 105.39 62.86 4.3 435 11.0 0.1 9309 107.60 60.70 3.4 306 11.9 0.3 9807 106.40 56.93 4.3 405 11.4 0.4 I use a Factory Pro EC997a.