Phil you can read lambda direct from the ECU using an OBDII reader, or install an O2 sensor just after the exhaust port. (this is the method used on the race bike). I any event there is little, if any reversion at full throttle and the test is a comparative one anyway.
Are you referring to using Leaded or Unleaded fuel? We have found that 12.4 to 1 is as lean as we can go with Unleaded fuel and sustained WOT. That was backed up by the fuel maker. !2.6 to 1 burns pistons. 12.8 to 13.2 is safe with leaded race fuel. This is with a naturally aspirated 13 to 1 small block chevy. Just my .02 Rick
"Patrick O/T, but how does the EFI Race 4.0 compare to the Motec, pricewise?" About the same but you get knock control, traction control and dual NGK lambda included.
I asked that question because I've seen guys sit on the dyno leaning off engines while seeing power gains. They then get to a point where they seem to have good HP so they stop. These guys are usualy dynojet tuners so they just keep doing run after run. What they aren't realising is that most of their "gains" are really just the engine/drivetrain warming up. they don't believe you until you put the AFR back to where it was and get the same power!
Leaded/unleaded is not relevant, it is the quality of the fuel you need to concern yourself IMO and the accuracy of fuel distribution. Given a good combustion chamber design and accurate fuel distribution, running 13 to 1 CR I would start with 104 Octane, otherwise you are using an excessively rich fuel mix to combat excessive combustion temperatures. When unleaded race fuels were first mandated in local racing we had to drop compression ratios until better quality fuels became available.
So basically, operators make errors, dynos lie, conditions vary... tests are not perfectly repeatable. It all comes back to the stopwatch. Good tuners earn their money with reliability and reduced laptimes / 1/4 miles.
Let me be clear, I am no expert on the subject, just spent many years hands on tinkering (no jokes thanks!) Just learning how to get repeatable results on the dyno was a discovery in itself. When tuning, I learned that controlling water temp, airflow and most importantly AIT were the most important factors in achieving truly comparable and accurate readings.
Most of the development by the top tuners is done on a dyno, not the track, which results in reduced lap/quarter mile times.
..... what about transmission temp, tyre temp, tyre type, tyre condition, tyre size, gear ratio, humidity, barometric pressure, ambient temperature... and lets not forget about cam timing..... ... and alternator loading through battery condition and loadings..... engine dyno (not chassis) in a controlled cell, or stopwatch. All else is interesting and fun, but not definitive proof of result.
Don't get me started!! Modern dynos correct for the atmospheric differences, but for these sorts of tests, they are generaly run on the same vehicle on the same day. My procedure is dyno room with at least 40,000CFM capacity (BM Fuel Systems only has 12,000) Run the car until oil/water all at optimum readings. Three "dummy runs", engine idling with dyno fans running for 5 minutes, or until oil and water temps come back to spec, then actual power run is commenced. I also only use one dyno - Dynomotive in Bayswater, who have sufficient fan capacity ( I make them use both of their large fans), up to date software and regular calibration from DynoDynamics. I understand Vinny at STI Docklands facility is as good or better. APS has the best facility, but don't allow customers on their dyno anymore. At one stage we were using a new slick for every dyno session as the torque of the Ducati would just destroy the tyre and leave us with inaccurate readings, also the Ducati was VERY sensitive to cam timing, we tried aorund 20 different settings before we got the optimum combination!
What kind of Duke do you have? Would the cam timing sensitivity be related to the Desmo system? Oops, getting OT.
Road and race. Colin Edwards at Phillip Island road bike Eastern Creek Aust Superbike Championships 2003 Image Unavailable, Please Login Image Unavailable, Please Login Image Unavailable, Please Login
does any one know, do any OEM ecus run closed loop under power? I don't know of any that do. I also don't know of any that run a wide band O2, so I don't see how they could. All the ones I know of run a narrow band senser and use it to keep the mixture right for good gas milage and good emissions results. To accelerate they drop out of closed lood and use the maps. That's the why I set-up my car too and it seems to work out good. Which brings me to the real point, I don't see the slightest reason why you would want to be closed loop under power. It's nice to rough in the fuel maps quickly as someone else mentioned, but once the maps are written and tested, what the O2 reading is doesn't seem to be of any significance. On a general tuning note I have seen fuel make much more than a 1 or 2% change on the dyno...but not by itself. What I generally do (naturally aspirated) is incease the fuel until the hp stops increasing. Then increase the timing to peak hp. Then I go back to fue and if the timing changed, then usually it will want more fuel again. If it takes more fuel, usually it will then want more timing. It usually take 4-6 times back and forth before nothing changes anymore, and depending on the state of the original tune, the hp increase can be pretty significant and it's ready to race.... But that's on a drag engine where heat build-up isn't an issue. For roadrace I add at least 1 point usually 2 to the dynos A/F result (which reduces hp 1-2%) before hitting the track, then watch the temps and plugs and adjust accordingly.
Mark, I think you are correct in that the O2 sensor is used(closed Loop)only at idle-steady state driving conditions. Then when under power the manufactures engine management maps for that engine are used. The way I understand this is that disconnecting the O2 sensor would only effect the mixture at that idle or steady state, what would be the point of disconnecting it? Usually when it is not fuctioning properly the mixture is slightly richer which is the reason most cars get slightly less fuel mileage when the O2 sensor isn't doing its job. Am I correct on this guys? Please correct me if I am not. Love trying to understand computer operated engine systems. Regards, Vern
yes, the motec ECUs support narrow-band and wide-band operation. Abosutely, wide band can be used under all efficiency points. I am using it now with the M48pro on my countach. The m4x ECUs use a bosch lm-11 sensor. This is a "bin-ed" narrowband sensor that can be used for wide band operation because the temperature induced error is known, and can be compensated, resulting in a fairly accurate lambda. The m800 supports the more accurate lsu-4, which is a 5 wire sensor (this wide band sensor is used in many cars worldwide). The m4x also supports the lsu-4 in conjuction with a PLM, which is what I upgraded to. The plm is the controller for the lsu-4 and provides the lambda signal to the m4x. As for the wide band operation, I agree, in theory it shouldn't be necessary. If the motor is tuned properly on an engine dyno and fully mapped, then there's nothing else to consider, right? Problem is, not every motor has had the benefit of being formally mapped in this manner, so there is "best estimating" for the initial map based on engine displacement and injector size. This will get the motor running. Then a better map can be developed with a chassis dyno that can hold an efficiency point (not a dynojet, it can only do WOT runs) and the fuel table entry modified based on target lambda. Motec has a "quick lambda" feature which is useful for tuning. Alternatively if you are brave, you can use the road dyno and modulate the brake/throttle to set a few key table entries (a passenger can do this), then the rest can be interpolated. Also, wide band is useful because it will compensate for changes not considered by other sensors. i.e. changes in VE due to exhaust/intake manifold configuration, or air filters are clogged over time then they flow somewhat less etc.... The lambda sensor will correct for this. Also, it is not completely accurate to say lambda measurement works "after the fact". What really happens is the ECU knows when it can sample the sensor signal and apply this as a trim. yes, there is a lag, and this is part of the algorithm with some user config parameters. Also, there are short term and long term trims maintained by the ECU. Once the ECU uses a lamda measurement as a "trim", it is considered a short term trim. This is folded into the short term table. The short term entries eventually migrate to the long term table, per user config knobs. So in theory the short term table tends to zero out, as the long term table contains the full correction. Again, these tables are not user accessable. My experience with the wide-band, and motec in general, is very satisfactory. The closed loop system works excellent, and with data logging, you can see the corrections taking place. For a goof, I threw bogus values in the fuel map and openned up the correction band of the lambda feedback. Sure enough the motec learned how to correct this and the car ran fine (not that the this could be considered tuned). That's my nickles worth.
That's what you would thinl logically. under power modern ECU's use their maps, but also make corrections to the maps from inputs from the sensors. When the O2 sensor in my 550 shorted, the car ran lean, not rich as I would have expected.
assuming the 550 uses a narrow band sensor, if you shorted it to ground, then true, this simulates a lean signal. However, the ECU may not trust the lambda sensor if it does not see expected behavior. Narrow band algorithms modulate the injector pulse width and expect to see the narrow band signal toggle between .2 and .8 volts. This is how the software learns where stoich really is. If the toggling is absent, then the ECU stays in open loop mode. The fuel map is probably stoich, so that expains why you saw lean operation.
So what you are saying is that the map for a particlar engine is set, lets say, at 13:1 AFR at WOT (the ECU being open loop at this time) the 02 sensor is still contributing info to keep the mixture at the set 13:1? I thought the air flow meters info was doing that job at WOT. Regards, Vern
All Bosch, at least from 5.2 onwards, write a set of correction tables, which the ECU "learns" over time. When the correction inputs get out of the programmed range you will get an error logged or CEL set.
They're using a narrow band O2 to re-write the WOT maps, not just low power maps? Are you sure about that?
Hmmm. I did not know that. I haven't paid much attention to the 5.2 OBDII, I have a 355C with the 2.7 system which I don't think has that capabillity. Thanks for the info. Regards, Vern P.S. Great Duc, I have a 749S love that thing. I'm not sure whats more entertaining the Ferrari or the bike. HaHaHa
You just contradicted yourself. If the control system is working properly, you don't see the corrections. ;-)
