moto

Right, but to be precise the PCIIIUSB uses 500 rpm increments by default. 250rpm increments become available when one clicks on "Maximize Map Resolution" under "Tools". I am interested to know whether Powercommander Authorized Tuning Centers automatically tune at 250 rpm when the option is available, or if you have to ask for it. If they don't do it automatically, I wonder how much extra it costs. A word of caution: If you have the tuning done at 500 rpm increments, the software just fills in the gaps in the map with an average. My opinion is that this is not a very good idea, as adjacent cells sometimes require radically different values for optimum running. I don't think it makes the adjustments as it's letting the engine rev between steps, but rather during the steady state exactly at the steps. I think that's what's accounting for the pauses you hear. My version of TuneBoy allows me to tune in realtime. Granted, it's an alpha version, but I'm fairly certain Wayne intends to release a realtime version to the public at some point in the future. If there are indeed two methods (see below), it could have been either way. Could you ask them? I am very curious about this. There may be a method (which you alluded to) wherein the "A/F ratio" is measured during an inertia run at a certain TP, the map is modified, and then anther inertia run is performed to verify the validity of the changes. Maybe the varification runs are not usually performed, and that's why no one has any to show. I also wonder how the software is set up to function during realtime "autotuning" via step tests. Can you save the data from the beginning of the test in order to compare to the result, or do you have to perform one set of step tests beforehand? So maybe the before runs would cost more money? If so, I wonder how much? They should at least be able to provide the after runs for all throttle positions. I can't imagine the data taking up too much space on a drive. Even so, they could erase it after printing. Except that everyone might not do it the same. The ability to perform controlled sweep tests (at however many ft/sec/sec you want) has been a standard feature on Factory Pro eddy current dynos for at least 7 years now. However, sorting out problems on a dyno is a lot easier with an EGA (which doesn't really work during sweep testing). Say you have a bike that craps out after freeway riding for a while and you want to diagnose the problem. If you get the problem to occur on the dyno and it's fuel starvation related, you'll see the CO fall and fall until it quits. If it's ignition related, HC would go way up as it was crapping out. Or if you had a misfire, HC would go up at the same time as O2 and flux. I can't imagine that you couldn't manually tune during step tests on a DJ dyno. Why not?

Huh?

Well according to the articles, some of the issues that cause problems with a WBO2 sensors are abrupt changes probe temperature (on off throttle transitions for instance), minute changes in voltage/voltage offset (turning lights/accessories on and off, changes in charging voltage due to rpm), and changes in backpressure. With Tuning Link changes in voltage are likely eliminated by the presence of a regulated power supply, changes in pressure are likely eliminated by a pump, and abrupt temperature fluctuations are likely eliminated by the fact that the sample will have given up some of its heat by the time it reaches the sensor. The drawbacks of a probe on the other hand are slow/fluctuating response time and, if you can't get the probe in far enough, sample contamination. The former is not a big problem in steady state testing. Although this has nothing to do with a screw in sensor vs. a probe per se, sensor output/logging device (or ECU) input impedance mismatches must also be compensated for. I would ass-u-me that DynoJet has adressed this, whereas with closed loop ECUs this is an unknown, especially if the sensor controller does not come from the same source as the ECU and/or the people manufacturing the ECU are unaware of this issue. (I'm not trying to imply anything, Cliff!) That would be great! But I'll bet other Factory Pro EC997a equipped facilities with the appropriately skilled personnel would be willing to assist as well.

In thinking about this further, the TuningLink probably doesn't have any of the problems mentioned in the articles, as it doesn't use a sensor that screws into the exhaust system.

I'm convinced that currently TuningLink will work better for a given throttle position/rpm step than a closed loop ECU using a WBO2 sensor. The WBO2 articles I referenced earlier (http://www.megasquirt2.com/PWC/LSU4.htm, http://www.techedge.com.au/vehicle/wbo2/wblambda.htm & http://www.bgsoflex.com/pwb/0.95/PWBV0.95_QandA.pdf) indicate that with the current controller technology, the sensor works much better at steady state than it does installed on a vehicle and attempting to respond in real time. I would encourage people to read the articles, but the two main points seem to be that abrupt changes probe temperature (on off throttle transitions for instance) and minute changes in voltage/voltage offset (turning lights/accessories on off, changes in charging voltage due to rpm) cause innacurate sensor outputs. According to one of the articles, this is being adressed with a much more sophisticated controller technology. I was not able to derive whether or not this technology is currently available and if so how much it costs, but it could make full closed loop much more feasible. Maybe Cliff could look into this (if he is not already doing so)? One of the other drawbacks that was cited was that the WBO2 sensor needs to "know" how much backpressure is present because of extreme sesitivity to pressure in general. This definitely makes sense when one considers how the thing works. I'm not sure if controller technology adresses this, but if not, it can be remedied by developing individual sensor output targets per rpm/throttle postion/cylinder for a given motorcycle with its particular combination of parts. Of the three drawbacks mentioned (besides O2 content not relating directly to mixture strength, which, admittedly, has bean beaten to death), the TuningLink only suffers from one of them. This brings me to my next question. I'm wondering if it is not possible to get results that are as good or better than without it at a comparable or better price point. So far what hampers direct comparison is that I have not found anyone who has had tuning work done via TuningLink who was willing/able to tell me what throttle positions they had tuned at what rpm increment. The only thing anyone seems to discuss is price. Can anyone here fill in the blanks?

To all of those who I may have offended, please accept my apology. To offend was not my intent. I am not the type of person who thinks that anything positive could come from it. My intention was to be rigorous. I can't see how the truth can be revealed if falacies are allowed to stand. This goes for myself as well. I have welcomed (whether it seems so or not) the challenges presented to me, because I have had to examine further whether the knowlege I have or think I have is correct. My only interest is to do the best job possible, which I can't do if I don't continue the push the boundaries of what I know, and if I am not willing to reexamine what I have held to be true. I had hoped everyone would benefit and that the bar would be raised. I had hoped that empowered consumers would go out armed with a more exact idea of what they wanted, how they wanted it done, and the ability to specifically ask for it. I may be completely off base here, but it seems that especially in the United States, no one wants to be bothered with any details. Unfortunately, the truth happens to be in the details. Unfortunately, potential for progress happens to live in the details. Some of you may find this to be tedious, masturbatory and whatever other language was used. I don't think it would be possible for us to have motorcycles that perform as well as they do if someone had not concerned themselves with dissecting these details to what seems like a mindnumbing extent. We certainly would not have this internet forum do discuss these matters in. Without getting too philosophical about this, people put their blood and sweat into the fruits we are reaping today. Should we be betray them by being advocates of entropy? Without thinking on things deeply, and being awake, I don't see any reason to live. Maybe it was not my intention per se, but the unskillful way in which I went about things that is the culprit. Again my apologies. If anyone wonders why I have been anonymous so far, I would like to quote a couple of posts from earlier in this thread: and etc. Now, I've really got nothing to hide, so I suppose having been anonymous was misguided. I don't know and don't care if I'll ever tune another Guzzi. I won't be retiring on any island soon. I have dedicated my entire life since I was a child to doing what I am doing and I am not making a lot of money. I am just concerned with doing the best job I can. If I can't do that because there is no market for it, so be it. I would rather live in a cardboard box than do a job that I cannot be proud of. Best Regards, Derek Capito http://www.moto-lab.com

Good news. What are your criteria? The only evidence I have is that you didn't mention anything about it... I was suggesting that the pickup be moved circumfrentially around the rotor, which doesn't necessarily imply that the gap would change. I know you were estimating a 360mm circumfrence, but without knowing the exact size of the rotor, hard to say. From the picture, it does look rough because the pickup seems to be located in its hole and has those screw bosses on either side of it. Is there normally oil in this cavity? If the rotor is keyed to the shaft, you might be able to use an offset key. You're welcome. I'd like to check that out, do you have a link handy?

I think that given the extreme temperature difference between the previous run and this one, I would have opted to do another baseline with the original configuration. SAE corrections work well in general, but I have also found that I do need to do another run if the weather has changed drastically since the last one. I think a delta of 28.82° is what I would call drastic. Even if the starting engine temp was closely monitored and exactly identical, it would have climbed much faster on the hotter day, especially with the poor cooling arrangement evident in the photos. This could possibly have been mitigated somewhat by doing step tests instead of inertia tests (so you could measure every rpm at the same starting temp), although DynoJet dynos have the nasty tendency to take a long time to stabilize at each step due to the high inertia of the rollers they employ, which again cancels out some of the benefits of that strategy. You need to be careful with inertia runs, because simply swapping rear tires can erroneously show you a horsepower gain/loss. It would be great to see the before/after runs for all other throttle positions and just for shits & giggles with the "A/F ratios". I'm also suspicious of this "CF:SAE Smoothing:5" business. What's that all about? Waddaya need that for?

Shouldn't you be comparing quality as well? Makes sense, although if the tuner does his job, you shouldn't have to fuss with it again, at least until the next time you decide to make some hardware change/modification. I think that depends on how you define "dyno tuned". I think it would be good to know what exactly which throttle positions/rpm were tuned for the $500. We already know the methodology (hitting "A/F ratio" targets) that was used, so you could ask the FIM guy/gal how much he/she would charge to do something at the same level. The added expense might only be due to additional precision/thoroughness he/she normally employs, which he/she might be willing to forgo if asked nicely. I know for sure that Doug Lofgren could tune to "A/F ratios" using the FIM software if so desired, but he could also tune for power and change advance as needed. I don't know how the pickup fasteners are arranged relative to the crank axis on a Guzzi, but what I had imagined is that they were parallel, which means you might be able to slot the mount. Is the rotor splined or keyed on the shaft? At the risk of stating the obvious, have the person doing the baseline also do the tuning. That'll learn ya! Makes sense to me, except then you won't know if the gains came from the exhaust or the tuning. You could just optimize 100% throttle and ignore all other throttle positions for comparative purposes. Perhaps, but see above... Yes, I would think so, exept maybe at the very top. If you shorten and install a less free flowing muffler design, you might loose everywhere. How do you know what the power potential really was without optimizing fuel and ignition properly? Mufflers don't do that unless they're not really mufflers. Agreed, getting rid of a crossover means that you must have much, much more free-flowing mufflers to compensate. Yes. Don't do it. I generally agree that crossovers are a good idea on a vee twin, but for two into ones to work, the design is very, very critical. In other words, you can't just expect it to work because it happens to be a two into one. No, it just needs to be the right size and in the right place.

I have not yet seen a dyno that acts as a wind tunnel. I suppose one could put a dyno in a windtunnel and then test with airspeed below, at and above roadspeed. Yes, but you would be surprised at what some dyno facilities have, or don't have, I should say. If you didn't have a wind tunnel handy, you could put pressure transducers in your airbox and record with a datalogger. You could then try to duplicate various airbox pressures, up to the max of what the datalogger recorded, on the dyno. This is very difficult to do, as a major amount of CFM are required. A magazine tried doing it a while back and found that they were not able to find a suitable fan. Just because they couldn't do it doesn't mean it can't be done, but you get the idea that it's nontrivial. What they did find to sort of work was a huge industrial compressor with a massive hose. With this thing, they were able to create enough pressure for just long enough to do a run. In the case where the motorcycle had two inlet snorkels, they had to block one so the pressure wouldn't leak out. I think they were just trying to show how much more these engines put out at speed, and I don't think this type arrangement would be suitable for development work, as controlling the pressure would be too difficult. Ideally, you would test at all combinations of throttle position, rpm and airbox pressure the engine might encounter. Then we arrive at the next problem, which is that there isn't any way to access (so far) the air pressure correction in the ECU. If we were lucky, we might find that there is no error, and that the ECU corrects just fine. If there were errors, we could tailor the fuel delivery to the most common or most important (as deternined by datalogger traces) airbox pressure for a given throttle position/rpm. Or we could average the error over the given throttle position/rpm so as to loose the least power at either end of the pressure range. Or some combination of these two approaches. Maybe we could construct a lying box that intercepted the signal from the ECU's airbox pressure transducer and tweaked it so the ECU would think there was a pressure that happened to make it correct exactly the way we wanted. To start, in the case of the Guzzi, I would just see that there is a pressure transducer in the airbox stock, see whether it has the range needed to provide the appropriate signal to the ECU, and lastly see if the ECU has enough range built in to respond to the signal . If only wideband sensors didn't have the problems they do, this, with the development of proper targets, would be a perfect application for a closed loop ECU.

Exactly! I do it, and I get finished. Your spark plug wire thing may not, but comparing hp and EGA data does. The standard reference point is how well the engine performs on the dyno as compared to previous runs. If you are trying to say that we can't compare readings from different dynos against each other, I agree to a point. In testing the same machine on my dyno vs. Wheelsmith Racing's (a subsidiary of Factory) I found the numbers to be repeatable within better than .5 hp. However, trying to compare hp numbers from a DynoJet dyno to Factory Pro numbers is mostly useless. Are you suggesting we can compare percentages of gains across various platforms? I agree that it's silly to strive for a specific horsepower number, but I also think that under ideal circumstances one should add/subtract fuel or advance/retard the ingition until no more hp or mileage can be found by these methods. I don't recall anyone talking about aiming at a specific horsepower, only about achieving the maximum possible at all throttle positions/rpm. Also the differences in performance between premium fuels is minimal my experience. I often use up what is in the customers tank (various brands) and then finish tuning with Chevron, which is what I keep around. I have not been able to detect a difference that would lead me to believe that the tuning I did for one pumpgas would somehow be invalid for another with the same octane rating. Maybe this varies more seasonally, and certainly this becomes much more of a factor when we start to get away from pump gas. Where did you get the idea that I would not want to "optimize the FI system to accomodate all the myriad things it deals with and handle them well" or that I would be interested in "skewing the system to acheive some specific result at the cost of reducing overall useability"? Though, I do have to admit that I'm not sure what is meant by"skewing the system". 41693[/snapback]

It follows then, that the engine needed more fuel (at least in some areas), and that the use of the "tuning link" provided more fuel. How much it provided, where it provided it, and whether the amounts it provided are optimal is unknown. Without looking at dyno runs that include EGA data and/or changing pulsewidths, stagger and timing to see if there is an improvement, I would say that's unknown. It would certainly also depend on the definition of "messing up". Apparently it's not. You would have to define "crucial". Even so, I can't comment right now, as I would want to tune natively and Wayne has not yet released TuneBoy for Guzzi. I would need to see how many table positions there are, and whether there are individual tables for the two cylinders.

Imagine we equip a vehicle with a fuel flow meter and a mass air flow sensor. We then add or subtract fuel so that we have a stoiciometric mixture (as evidenced by the fuel flow and air mass measurements) at a particular chosen throttle position/rpm. We then screw a wideband "lambda sensor" into the exhaust. Do you think it will show a lambda value of 1.0? We then taylor the amount of fuel to yield some desired BSFC or HP. Do you think the A/F ratio as measured by our fuel flow meter/mass air flow sensor combo and our "lambda sensor" will match, or even be offset by the same amount? Say we try this experiment at a different throttle position/rpm. Do you think the results will match, or the offset from the previous comparison will be predictable, and therefore transferrable to other throttle positions/rpm? Say we map this all out so that we know what these values are at all table positions. Will it carry over to another engine of the same model? What if there have been modifications? As the above example illustrates, we are not. True. Here are some links that point toward where the errors with wideband sensors can come from:http://www.megasquirt2.com/PWC/LSU4.htm, http://www.techedge.com.au/vehicle/wbo2/wblambda.htm (scroll to the bottom) and especially http://www.bgsoflex.com/pwb/0.95/PWBV0.95_QandA.pdf. These people seem to have done their homework and have/are poised to release a controller that adresses many, if not all of the common wideband sensor problems (besides proper targets, obviously). If anyone still has doubts about the problems with narrow band O2 sensors, see http://vehicle.me.berkeley.edu/~markw/efi/SAE920289/. The first little graph is very telling. I suppose one could do that, but I don't generally believe in that sort of thing. What comes around, goes around, ya know...

Agreed! Right. Ideally, you would optimize the fuel and timing maps on the dyno to establish a baseline, and subsequent to every change. If you didn't do that, you couldn't tell whether all of the changes were actually beneficial. The large benefits of one change could be masking the small loss from another, or a large loss could be masked by the cumulative effects of many small gains. You might also think that parts you installed make more power, when the power really comes from the fact that the existing pulsewidths/timing are suddenly more appropriate. Or... Well, you get the picture. In any case it allows you to zero in on the combo that gives the best result, rather than just guessing. What do you mean, "even"? Are we talking about an existing system, or adding one?

Right, but whoever did do the tuning should have provided you with printouts that include "A/F ratio" traces for all throttle positions that were tuned and for both cylinders if they were tuned individually. If not, they should certainly be available on request. How much would it have been, what would have been done, and by what technique? Agreed. Do you think it has to do with a reduction in air density away from the Ocean? I imagine that shortening the headpipes will yield reduced bottom end and midrange performance while not doing very much for and possibly hurting the top end as well. For valve jobs and porting, I reccommend Kinetic Analysis in San Rafael, CA. You're very welcome. Why not? I'm sure the pickup could be moved slightly, or the rotor could be moved on the shaft. I did find some threads having to do with ignition timing, but none with regard to changing it mechanically. Agreed!

I don't recommend it. You could do it before with the FIM software. There is also nothing you stop you from doing it mechanically by either moving the pickup or the rotor. Yeah, they've supposedly been working on it for some time. If you had other throttle positions tuned, you should have been provided with the graphs to go with them. They should certainly be available upon request. Understood. I would be interested in seeing that. They should give you the files from the runs if you ask for them and there is a downloadable run viewer available here:http://www.powercommander.com/viewer.shtml There should be no special hoops to jump through to get that. Right. See above. It may just be that he wasn't thinking about how to solve the pinging problem outside of the PC paradigm. Wow! Does it come with a free Ginsu knife set? Does it slice, dice, and puree? "If" is the operative word. The owner of the Futura I keep mentioning had a pretty bad experience with the PC thing. He installed one and took the bike to an "Authorized Tuning Center". He got the bike back in what he describes as "unridable" condition. He took the bike back and on the second try there was an improvement, but not enough to be acceptable. They did not offer him his money back, nor did they offer to try anything further. The theory we have is that they used the Tuning Link for the first try and that when that didn't work as planned, they had neither the skill nor the patience to be able to figure out what else to do or how to do it. The customer then took the bike to a DynoJet facility that was not "authorized". He said that they took a little longer and that allthough the bike was not up to the performance he expected, he was willing to settle for what he got. Within a short while after that the PC failed, leaving him stranded on the side of the road. He was a few days out if warranty, and DynoJet refused to help him in any way. He then investigated other alternatives and found out about the TuneBoy software. His search for someone willing and able to tune with it led him to me. After I completed the tuning, I was interested in his stamp of approval before paying me, so I sent him out for a testride. When he came back, he could not wipe the smile off of his face, saying that the bike was performing better than he had ever dreamed possible. He called me later to report increased gas mileage and the cessation of an oil consumtion problem he had had (I think we finally got the rings to seat properly on the dyno). It's likely so be all, some, or none of the above (depending on dependencies of course).

As far as I know, the FIM software has been able to do this just fine.

I'm not aware of any such thing and I don't see any use for it. Even so, you can't make assumptions about the original mixture strength nor the appropriateness of the theoretical additional amount of fuel that would have been delivered at various table positions/manifold pressures/airbox pressures. To that comes that optimized settings for, say, a pipe and filter might end up at or near the same mixture strength (not pulsewidth) as was used with the original parts. It would also depend on whether or not the theoretical additional fuel that was delivered was a percentage of the main table position, or some absolute.

Why not try pulling the timing back? Do you mean the graph? Again, why not try pulling the timing back? Hard to say what happened with using blanket targets and with tuning to "A/F ratio" in the first place. Where and on what basis does he recommend this? Understood...

I think "all" is a bit overstated. If you will recall, I asked you for your definition of optimal. I did not ask you to tell me what you thought mine was, to give your interpretation of it, and then your opinion of your interpretation of it. How can you or I know what person would or would not be riding a Guzzi? I think that what is important in terms of performance is a very personal thing and that stereotyping, regardless of the arena, should be avoided as much as possible. You must mean within 5% of best power at every throttle position/rpm. Is that correct? It's absolutely none of my business to meddle in what will or will not make you happy! I do think that you could raise efficiency by getting away from using this "A/F ratio" thing as an arbiter. Good. I don't think that peak efficiency can be achieved through super lean mixtures alone. I don't think you need closed loop spark advance to get it.

I think that you could more properly say "will". Steady state testing with a low inertia roller solves this problem handily. A good tuner will initiate every throttle position/rpm step while the engine temp is falling and within ±.5°F or so of each other.41671[/snapback]

Cool! I'd like to see the before and after runs for the other throttle positions on this one as well. Just curious, why aren't there "A/F ratio" traces for this one?

What is a "bewilderingly small numerical difference" to you? 5 hp, 3 hp, 1 hp? Something else? It is impotant to note that many minute differences add up to make large gains. By paying attention to the small stuff, a good tuner can make considerable gains where someone less rigorous would not get anything. This is what distinguishes a good tuner from a mediocre one. Some of the most difficult to tune engines have the most potential for hp, and I sometimes make progress by 1/2 or less hp at a time. Sometimes it's the only way to massage things into place, but when I am done I look at the baseline and realize that the gains I made are massive. If I had simply ignored the small gains, I never would have gotten there. This is especially pertinent on engines that I am tuning individual cylinders and/or timing on, trying different parts combos on, or on carbureted engines. There are also sometimes areas on maps that are terrible in terms of hp and it seems that nothing will fix it. Sometimes those spots just need something really, really specific to work. The pulsewidth sometimes literally can't be off by more than a half of a percent or less of CO or there's this massive loss in HP. How would you figure that out and fix it without paying attention to the small stuff? Or other times when the CO just climbs and climbs because of reversion to a point where hp starts dropping. I then have to make a judgement call about where during that climb I am going to tune. How do you think "autotuning" or full closed loop deals with that? I would generally say that optimum system behaviour is when the most possible horsepower is available at every throttle position/rpm combination. I could also see someone wanting the best mileage at every throttle position/rpm combination, but to me that has little to do with why most people ride in the first place. I could also see someone wanting to strike some type of balance between most hp/best fuel economy at all throttle positions/rpm. However the latter two are quite a bit more difficult to achieve accurately and consitently throughout a map, regardless of the methodology used. If this means that the best performance at all throttle positions/rpm is not a useful goal, then I disagree. If you are saying that high peak numbers at full throttle/redline have little to do with the real world, then I agree.

That's not quite correct according to the well written article toward the bottom of this page:http://www.innovatemotorsports.com/resources/news3.php. It does however make clear that, as Cliff says, in certain ranges the sensor is not really measuring O2 content, even though it is called a "wide band O2 sensor". They claim it is measuring "A/F ratio", even though this is not possible, aside from measuring and comparing the mass of air being consumed to the mass of fuel that is being consumed. Regardless of what the thing is actually measuring, this still leaves the problem of supposed "A/F ratio" having a tenuous relationship to how an engine runs, the repeatibility errors Cliff and I have uncovered, as well as the O2 measurements in the areas where it does measure O2 being semi-useful for direct pulsewidth tuning. I can't totally get my head around the latter, as I don't know how broad of an area this would be in relationship to CO numbers.

What throttle position was this run for? Does the pinging happen at this paricular throttle position? Can you post the runs for the other throttle positions? Well first off, A/F ratio is not measurable by an exhaust sensor. We can speculate about how DynoJet attemptes to relate "A/F ratio", by simply looking at their readings as O2 sensor readings on an unknown scale. Second, we have already established that various mixture strengths can exist at a given O2 content. Third, a delay of some inconsistent duration always occurs between the time a sample is produced and when sensor readings are recorded. Whether or not the amount of delay that DynoJet applies within their software is appropriate at any given time is unknown.