| Author | Message | ||
     Anonymous |
BusDude, Not trying to bust your chops, sorry if I came off that way.  Just trying to improve the accuracy of Buellers understanding of this system. I guess I missed the reference to open (non) loop systems on other brands in there some place. My mistake. Some open (non) loop systems use a baro sensor for ambient pressure change adjustments. The loop remains open because there is no sampling of the final product to measure accuracy of changes made based on the baro sensor's readings. Other open (non) loop systems use a MAF (Mass Air Flow) or MAP (Manifold Air Pressure) sensor to adjust for changing atmospheric changes. I still say you, Mr. Mushman :0, are one of the best reads on BWB. (BTW, love the Harvey Mushman name, I wonder how many posting here make the connection.) For those that don't make the connection, shame on you for not paying attention to "On Any Sunday". Watch it again!! And if you don't own a copy SHAME ON YOU! MikeJ, Not arrested, but fined if the EPA decided to come after end users. It has happened before!! The fines are something on the order of $10K per item tampered with. A simple Dynojet kit installed in a carb means 5 items tampered with or $50K worth of fines. 1. pilot jet changed. 2. slide drilled out. 3. Spring changed. 4. Needle changed. 5. tamper plug removed and idle mixture adjusted. | ||
| Benm2 |
I thought closed loop systems still used a base map, but the entire range of values would be adjusted up or down based on feedback from the O2 sensor. In addition, O2 sensors are designed to provide a tight feedback range around stoichiometric conditions. Typically, they are used to keep an engine slightly lean for emissions. When accelerating (HARD) or at WOT, the system still uses the map. The systems discussed are more based on whether they're feed-forward systems, or feed-back systems. The "mapped" systems use a feed-forward system, ie they gather data regarding the state of the engine (air temp, rpm, throttle position, intake air pressure, etc) and pick a value off a chart for the injector timing. Depending on the system, this setup can adjust for air density & temperature. If the feed-forward is mass-air, it can also (sort of) adjust for engine condition; when an engine pulls less air at a given rpm (due to wear or anything else) it will use a leaner pulse value from the injector table. Feed-back systems collect data on events that have already occurred and make a recommendation on what to do in the future. Since the data is received late, it is not neccesarily of immediate value to the engine. Hence, it is used to make small full-map adjustments while the engine is in operation. Feed-forward systems work better than feedback systems. Think about it, would you rather be able to predict what's going to happen, or only react to it after it did? The better the data the feed-forward system produces (hence the shift to mass-air based systems on cars) the better the engine will respond as its responding to operating conditions before it injects the fuel. The O2 sensor lets you know how close you were. At least that's my understanding of the differences. | ||
| Imonabuss |
BenM2, Pre-mapped feed-forward systems can only adapt to those upstream items measured, and only move to a pre-set map. Lots of things happen other than what happens to the ideal engine around which the map was built. O2 feedback systems tell you what just happened, not what might be happening if everything is OK. And, they do not have to be set at stochiometric, they just need to be offset appropriately. Given the sampling rate and speed of the microprocessor on the XB system, believe me, things are adjusted at a blindingly quick rate. This is not an adjustment every couple of minutes, guys! It takes a reading virtually off every time the engine fires. At 6000RPM, that is one hundred samples per second. Sound fast? To something mechanical (like a valve!), yes, but to a fast microprocessor, you must be joking to think that is fast. So it can sample and adjust the map accordingly. And, it does not adjust the map so much as the appropriate adjustment available to the ECU. It does take a more powerful processor, but the XB has a much more powerful processor than the one on say, a Ducati 748. So, what does this mean to a rider? A pre-mapped sytem may merrily be leaning out the bike at a slow rate as he is climbing up a mountain road, based on the mapped values for the barometer settings set with that ideal engine back at the lab in Bologna. Sounds fine, right? Not inside the engine, which happens at this moment to be wishing it was in ideal conditions back in Bologna, drinking some wine with its sandwich. It is unhappy as hell, and pissed off at the stupid ECU that thinks it is back on the dyno. Because, actually, the cheap ass rider decided not to spend the $800 on his recommended desmo valve job, and now one little desmo shim is slightly the wrong size, and the intake valve is not completely closing. The rider also refueled at the bottom of the mountain, and decided not to buy the super-premium, overpriced Saudi swill. And he is getting frustrated that your XB is killing him in the corners, so he is redlining it in every gear on the straights. Now, you who are riding with this poor sucker (and stuffing it under him with delight) have sort of the same situation, but your XB DDFI system is noting the fact that you also bought the cheap gas, and that you put on a pipe and were to damn cheap to buy the race ECU. Are you about to be penalized for your stupidity? Nope, the closed loop DDFI is reading the exhaust as getting hot, and will therefore not be leaning it out so much. Result? Pre-mapped sytem feeds forward a bunch of outdated information not applicable to the individual situation and gives your import-riding buddy a burnt piston. Your brilliantly developed XB DDFI gets you home without a hesitation. God Bless American technology!!!! | ||
| Bud |
Bussman, I now, the Xb DDFI is for me the best thing since sliced bread. I’m very happy it ain’t a magneta marelli FI This will explain the wide offer after market software applications to alter the mappings. Its not a luxury, but bare necessity ! But the main question remains; will the race ecm coop with 1200 cc ?? With a minor adjustment like a higher pressure fuel regulator So far I heard that the inj. & throttle house are the same as the X1’s ?? The TFI box, is not a option so far. there still working on one for the XB, the one that the got will only work in the open loop  en will give error code’s in the ecm. And I hate all things with those pinch wire connections. c.b. James is offering re-programmed race ecm’s , ill ask him what he is doing to make it work, perhaps more duty cycle ??, or taking the IAT signal and make it reed – 30 C in the ecm ?? hmmmm. Still have till next winter to decide what and if I going to big bore it. For now,10 march my fist dyno run, the bikehospital is going to try out a new exhaust ( Hsa ) Ill post it. Gr,m | ||
| Davegess |
Imonabuss, very interesting information. Sounds like Buell is out toward the leading edge of motorcycle design on more than just the chassis. How does this system compare to the better automotive systems? Or to the systems in use on MotoGP bikes and F1 cars? My little knowledge suggests that for a bike system to do the same things a car system does it needs to be smaller, lighter and tougher, so for the same result the bike system has to be “tricker” than the automotive version. Dave | ||
| Sarodude |
El Bussman Grande- How does the DDFI see the exhaust gas getting hotter? It can see it get leaner or richer - but how does it see exhaust temp? Also, does the XB FI utilize a knock sensor? This is good stuff! -Saro | ||
| Hootowl |
The only major difference between cars and bikes is that cars have room to use MAF sensors, and bikes usually don't. This means that most EFI on bikes are speed density type systems. They don't measure the amount of air coming in, it is calculated from input from the other sensors, specifically the TPS, IAT, and engine speed sensor. | ||
| Blake |
Saro, Leaner exhaust is hotter exhaust; they go hand in hand. Hoot, Aint't acronyms great? Back in my stealth bomber days with fellow engineers, I could carry on an conversation entirely in acronym. MAF = Mass Air flow? IAT = Intake Air Temperature? | ||
| Sarodude |
Blake- Leaner is hotter only if you're too rich. Leaning out too far will cause EGT to drop - though that may partially depend on one's definition of "too far". Ask any of our pilot friends or anyone else who's ever run an EGT gauge. Experience (and some dangerously minimal assumptions) have shown me that peak power tends to be darned near peak EGT and peak EGT tends to be at the "power rich" area of carb operation. -Saro | ||
| Anonymous |
Saro, Buell systems do not use a knock sensor. EGT is not monitored. | ||
| Blake |
Saro, I disagree with your statement that "leaner is hotter only if you're too rich." I agree that leaning beyond a certain point of leanness will begin to reduce exhaust temperature. However, that point is an extreme and unlikely condition. Within the realm of reasonable operational conditions in an engine, leaner is hotter. | ||
| Rick_A |
Doesn't the ECM also receive data from the TPS? What about the intake air sensor, and the tachometer? Aren't all those feeding data to the ECM too? Blake, all those sensors in all open loop systems that I know of are there to figure cylinder air volume and cylinder air density to deliver the proper mixture. As it's been noted some systems use a slightly different mix of sensors to gather the same information...closed loop only adds, obviously, the exhaust O2 level. I didn't see the point in elaborating...and others already have done so quite well. I think anyone who owns a Buell has at least a basic idea of how DDFI works...is there a great mystery I'm missing? Don't some tuners use engine temp sensors in air cooled engines to figure out the maximum safe fuel mixture? | ||
| Jim_Witt |
I own a couple of Buells but they don't have FI if that matters. (smile) -JW:> | ||
| Skully |
Blake, I agree with Saro. Is there a way I can post a .pdf document here? Keith | ||
| Hootowl |
Blake, I know what you mean. A typical conversation between two avionics techs in the Navy would go something like this: "I put that Y coded 2745 up on CASS and it ran ETE RFI, so I A799'd it back to O level." By the way, the above action will start a major pissing contest. | ||
| Mikej |
Skully, click on the BBS Formatting button and find the attach function. There are size limitations, but I think it takes almost anything. What is the pdf file about? | ||
| Mikej |
I still think these efi programs need an over-ride function that once the low fuel signal is activated the system stops trying to compensate. If my carb'd bike runs out of gas I just refill and go, a f.i. bike should be the same, and this is one of two main reasons I don't want a f.i. bike. The other reason is home maintenance and roadside troubleshooting simplicity. Hmmm, I wonder if I can add this test to a test-ride bike. Drain the gas to a quart or less and ride until dry. Then add in a quart or a gallon and continue the ride. | ||
| Sarodude |
Mikej- I'd change that to override if there's a low fuel PRESSURE condition. I could just imagine making a serious elevation change right as the reserve light comes on and having a poor running bike when all is said and done. Just an opinion. Also regarding EGT and power... We must keep in mind how the tuned length of the exhaust changes with EGT and just how much a particular motor may depend on an over-rich mix to keep things cool (and keep making power). Aircraft EGT observations: Get inside a puddle jumper with an EGT gauge. Find your desired altitude, level off, and lean the mix while paying attention to EGT and engine pitch / RPM. Max RPM (max power?) coincides with max EGT. Good practice has the pilot then richen the mix by a squirt to keep the engine a bit understressed and perhaps more responsive to sudden throttle change... -Saro | ||
| Blake |
So Keith, are you agreeing that anything leaner than theoretical stoichiometry reduces EGT? Cause that is what I interpret Saro as saying. Can you post a pdf? What mike said. Read the BBS Formatting documentation. Lots of good formatting tags available. Don't make me tell you again. | ||
| Sarodude |
Blake- A couple of things. I never mentioned 14.7 or stoichiometric. Also, all this assumes that no detonation occurs. -Saro | ||
| Skully |
I know from personal experience tuning carbs over the years that most blue pipes are caused by too rich a mixture, not too lean and have an article that backs me up that I would like to post. | ||
| Jim_Witt |
Keith asked: Is there a way I can post a .pdf document here? Do you mean actually post it or link to it? If you have Photoshop you can open the .PDF and convert it to a .JPG or whatever then save it. More than likely it will be larger than 50k, so you'll have to run it through a compression programm like xdat. -JW:> | ||
| Blake |
Keith, I think I see where you are headed. An overly rich mixture, way overly rich, will due to its slow burn propagate a flame into the headers. The flame is hot! The pipes turn blue. I agree. The scenario I thought we were addressing is a bike that is running properly (not overly rich) that then starts to go lean. I know from my experience over the years of not tuning carbs, that a lean running engine will cause even a water cooled motorcycle engine with an electric fan to overheat. What a nightmare!  Does that mean the exhaust gas is also hotter? Seems like it would. | ||
| Sarodude |
Blake- It seems like what you're saying is that a hotter engine causes hotter EGT - all other things (including fuel mixture) being equal. What I'm saying is that hotter EGT generally makes more pwer - all other things (including CHT / Cyl Temp) being equal. Would that possibly move us more toward agreement? -Saro PS - Nothwithstanding the ludicrously rich scenario described above... | ||
| Benm2 |
While the map is built around "ideal conditions" in an open (or closed) loop system, the ability of the system to adjust to things such as engine wear, carbon buildup, reduced intake flow, or a dirty intake filter can all be accomodated in feed-forward (open loop) systems. Lets say the engine starts to get a dirty intake filter. Depending on the feed-forward device (let's say Manifold Air Pressure), for a given rpm (and load) there will be a corresponding drop in MAP. With the other values available, that lowered parameter will result in a lower fuel value being used from the table. This holds equally true for elevation changes, as MAP is measured downstream of the throttle plate. In the event the engine is accelerating, the conditions for acceleration call for (a) richer conditions than stoich and (b) every successive engine firing is different than the one before it. The data the computer receives before the firing event is more useful than the O2 sensor data in these situations. Also, the O2 sensor has the worst job of all the engine control devices. Its in a hot, dirty environment, and has a response time that is likely SLOWER than the processor. Also, it produces a step-change voltage around conditions that only vary slightly from stoich (from around .98 (lean) to around 1.02 (rich), or so) This, of course, excepts wide-band O2 sensors, which (so far) are not in common use. For steady-state operating conditions, engine control computers try to keep the A/F bouncing around stoich, so the signal received back from the sensor can be seen as a frequency (ie watching for the step change). My toyota truck, for instance, develops offsets of the stock table based on the feedback from the O2 sensor. Gathering data from steady state conditions, the system develops both short-term and long-term adjustment factors for the offsets. The difference is in the amount of data collected to make the offset, and the time that the computer remembers it. My truck forgets the short-term one every time the engine turns off, but the long-term one requires the computer to be reset (to be forgotten). Again, this offset is applied to the whole table. Lastly, the lousy environment that the sensor is in ensures that the data is not entirely accurate. As the sensor collects material (exhaust carbon) its response time gets slower. This hinders the ability of the processor to keep the mix around stoich, as it takes longer for the system to "see" that its still around that value.
Do you know for sure that the 748 has no mechanism to adjust for barometric pressure? Some systems have a small pressure sensor right inside the computer for just this reason. I think someone above already addressed the knock sensor comments.
Alot of current fuel injection technology was developed by Bosch. The first O2 sensor was on a volvo. And, are you sure the DDFI computer was made in the US? | ||
| Hootowl |
Ben, Nicely written. One little nit though, a MAP is a manifold absolute pressure sensor. | ||
| Anonymous |
BenM2, You are way wrong about open loop systems sensing engine needs better. Closed loop systems are significantly more expensive, which is a major reason why they aren't used more. Sure open loop systems get upstream information, but they have very little idea of what is going on in the engine. There is absolutely no way for an open loop system to adapt the way an O2 sensor does to so many situations. Closed loop systems are significantly more expensive for good reasons. Knock sensors are an excellent tool too, but different. Open loop systems don't even come close to competing with either of these feedback mechanisms. Here's some inside info. The XB has fantastic emissions numbers while running perfectly. Every magazine has given kudos to the perfect operation of the XB DDFI, and you won't find that ranking on many other EFI systems. Yet it passes '04 California emissions easily without a catalytic converter. You won't find many other V-twins doing that, maybe none. On your comment regarding the 748 story , yes it can sense ambient pressure change. But it would never be able to sense a slightly leaking valve in the engine...that would be way too small of a signal to show up in the airbox. But it will show up in the exhaust mixture. BTW, if you think Bosch is the number one developer of EFI, think again, and check the patents. Heck, they hung onto mechanical fuel injection forever. Sorry to burst the bubble of a Toyota owner like you, but yes, the DDFI ECU came from the USA, from Indiana, as a matter of fact. | ||
| Blake |
Saro, "hotter EGT generally makes more pwer" That makes no sense. The more heat that goes out the engine the less that is used for making power. However, that actually supports your point that EGT may be cooler for leaner mixtures. My assumption that a hotter engine means hotter exhaust may not be correct. You mentioned aircraft. I thought that the pilot of an IC engine powered aircraft would lean the mixture until the EGT reaches a peak allowable temperature; if the EGT climbs too high, the mixture is enriched. Is that correct? I may have to break out the thermo again. | ||
| Sarodude |
Blake- I see another area where our assumptions differ... I'm speaking from the perspective of carb tuning - no other changes. True - if you do something absolutely silly like decide to open the exhaust valve only 1/2 way down the power stroke you'll see a rise in EGT and a likely drop in power compared to a more properly timed exhaust valve. Same thing with a very retarded spark. Burn starts late and might propagate into the exhaust causing high EGT readings. However, if you're only tuning the carb (we were talking about engine management, after all) then generally leaning the mix to the point of max EGT will usually get you a nice result. As far as the aircraft thing goes, I was taught to tune to peak EGT (if it was available) or peak RPM, then back of a hair. What I'm talking about is reading EGT to tune the fuel delivery of an otherwise properly sorted out engine. The more heat you lose out the exhaust, the less efficient the engine is. However, does running a richer mix somehow magically improve the thermal efficiency of the design and tuning parameters? The EGT is simply an (imperfect) indicator that you got a hotter or colder bang based on the gasses passing by the sensor. It's obviously unfortunate that you lost that much heat out the exhaust, but without that, air cooled VW folks wouldn't have had heaters! (shoot me now) Look at it this way - let's go run an engine with no load at 7000 RPM, wait a while, and take whatever thermal measurements we feel like. Now, let's put some load on that puppy and run it at the same 7K. Things are obviously gonna be hotter. We lost more heat - but we were making more power doing it, weren't we? -Saro if an engine revs in the forest and no work was performed, did it make power? | ||
| Blake |
"I'm speaking from the perspective of carb tuning - no other changes." Me too.  "However, if you're only tuning the carb (we were talking about engine management, after all) then generally leaning the mix to the point of max EGT will usually get you a nice result." If you are going for fuel efficiency, yes. If you are going for max power, no. Two VERY different objectives. "As far as the aircraft thing goes, I was taught to tune to peak EGT (if it was available) or peak RPM, then back of a hair." Would that not indicate, since you are tuning for max fuel efficiency, that lean is hot? "However, does running a richer mix somehow magically improve the thermal efficiency of the design and tuning parameters?" No, it hurts efficiency. What does that have to do with leaner being hotter?  Tune your bike on the dyno for max power. Then reduce the jet size. The exhaust will become hotter at equivalent power output. "Look at it this way - let's go run an engine with no load at 7000 RPM, wait a while, and take whatever thermal measurements we feel like. Now, let's put some load on that puppy and run it at the same 7K. Things are obviously gonna be hotter. We lost more heat - but we were making more power doing it, weren't we?" If you are going to compare the effects of A/F on EGT, you need to do it for equivalent power output. After all, no matter the A/F, it takes the same amount of HP to cruise down the road at 70 mph. The bike running a 15:1 A/F will have a hotter EGT than one running 13:1. You agree? That is all I'm saying. |
