mrvedit

I would like to log MPG, not for tuning lean cruise, but for documenting major engine changes, like different camshafts. Finding the optimal advance for hwy cruising would be nice too.

soundengineer

iTrader: (16)

I used to do this years ago....
the true way to enable lean cruise, is to change some bits of code in the Binary file of the ECU....
basically, there are multiple lines in the code that check to see if the vehicle type is Holden....
if its not holden, it bypasses the lean cruise routine all together...
if it is holden, then it uses the lean cruise.....
but there are like 20 spots it needs to change in to make it work correctly and not lose anything else...

however... there is one bit of code that tells the ECU to do the check for Holden...and if you just make it not do the check with a "No operation" command placed in just teh right spot... it works perfectly

then you fix the checksum, and do a write entire of the file to the ECU...and whammo!... lean cruise works

Thats the only way it works 100% correct on a non holden vehicle
you can try to put a Holden file into a car(like a GTO), but you usually lose a bunch of stuff in the process...gauge cluster stuff or cruise control...or ABS....
which is why the only way to retain everything is to change a couple of bits in the binary file using a hex editor





and FYI.... I do not do this any more....
its time consuming to do...and just not worth it...
not to mention that you have to have a binary file to begin with, and most modern EFI software has gone to a compressed and encrypted version of the file to protect their own R&D of making it possible to read and edit the file in the first place.
Do not ask me if I will do one for you... you dont have enough money....

kingtal0n

iTrader: (1)

Thanks

Heres an example of how to calculate MPG on any car, I wrote a small program in VB to do this in real-time while I drive with my laptop to find the best speed/timing and a/f, but as you can see you can accomplish this with just a pen and paper, while you drive write down the numbers you see then come home and do the math like this



You might notice that at leaner A/F ratios the engine pressure drops, I find this is consistent for most engines, and it means the injectors are going to flow more when you start leaning out the a/f ratio due to more fuel pressure. That is why we must calculate fuel pressure when doing this measurements. My previous post contains all the other info you should need for calculating and finding best economy. One more thing I need to mention, there is an error% at every rpm due to the difference between the computer's calculated injector on-time, and the actual on-time, since there is always an injector delay based on voltage, and this is known as "latency" and deviations from actual latency will cause you to derive, as I have just done above, "theoretical mpg" from "theoretical duty cycle". This might sound very confusing and complicated at first, but I will share you the simple way of fixing it, that is, once you have calculated your theorectical MPG, simply take the vehicle out and fill the tank to the absolute max, then drive at the speed you calculated your fuel economy for. Once you go 20-50 miles simply re-fill the tank and calculate you actual mpg using miles/gallons, then compare your theoretical with your actual number. In a car with a stand-alone capable of adjusting injector latency, you can go one step further and manually adjust that latency to make your theory and actual mpg units match, but this isn't necessary.

mrvedit

I understand your calculations; I just wish there was something more accurate than Injector Duty Cycle. As you mention, this will depend upon whether the fuel pressure is fixed or varies according to MAP. And as you mention, injectors are very non-linear at low pulse widths (low throttle).

I wonder if a calculation of the MAF's g/sec and the AFR might give a reasonable estimate.

kingtal0n

iTrader: (1)

The maf is accurate for airflow, but this has nothing to do with fuel. A/F ratio is never reported accurately enough with a wideband for fuel mass estimate; 14.7 could be 14.681475 or 14.741118457 there is too much variation to get any kind of accurate estimate of fuel flow this way, and many other small details such as exhaust leaks pre-sensor, even tiny ones, or misfire % (poor ignition performance) which will allow un-reacted move the sensor voltage more lean as well as fuel that is lost past the cylinder rings, fuel that is left on the valve, as most injectors provide fuel to closed intake valves where it sits for a moment. The fuel exposed to rapidly changing pressures there, partially evaporates and wets the runner of the head due to it's volatility.

The only way I know of to determine fuel mass injected is to ask the injector. Information provided by flow-test results can be used in conjunction with engine vacuum to determine mass of fuel injected per unit time, plus or minus accuracy problems due to voltage related latency delay and resolution of duty cycle report/log (does it tell you 14.4% or 14.4152% duty cycle?)

I must point out the application here is towards saving fuel and not determining mileage. To get mileage perfectly accurate you need to fill the tank, drive, re-fill and divide two numbers. That is a solid, realistic, accurate way to determine mpg in any vehicle. What we are actually interested in is adjusting for better economy coupled to a reduction in EGT is a double guarantee, and it does not matter how inaccurate your measuring device is, but rather, the consistency with which it reports numbers (precision of even erroneous values is useful for comparisons). In computer modelling this would be called "module order modelling", and by comparing "non-sensical" data we are still able to measure for improvements.

Finally, there are standard improvements you can make without every doing a single calculation. Tire pressure, alignment, lower vehicle weight, wheel bearings, driveshaft weight/balance, flywheel, internals, axles, differential type, deflection in any components, even motor mounts that are weak will result with a disturbed transfer of power to the vehicle and should be addressed. The drivetrain itself needs to be as light as possible for maximum economy, and as balanced as possible for minimal disruptions and unwanted energy transfers. Then the vehicle itself should be easy to roll as possible, and as light as possible. None of this has anything to do with fuel injection efficiency or tuning; and they are constants, i.e. lower vehicle weight always results with better economy, all things held constant.

LETZRIDE

iTrader: (6)

The has to be something for it because my aeroforce gauge calculates it and I can see instant mpg on my gauge as I drive

That's why I know I am having issues with LC enabled- I can see my commanded vs my AFR in real time without looking at the laptop, then cycle to my mpg and see I have dropped now from 26mpg to 15mpg with LC enabled.

But I am leaner than commanded and am using the stock monaro spark tables.

I am still diagnosing why this is

mrvedit

Based on all the inputs and the tuning tables, the PCM must surely calculate how much fuel it needs to inject at that moment, lets call it "need-xx-g/sec". Then based on the Injector tables and RPM, it calculates the injector pulse width needed to inject that amount of fuel.
Too bad there is no PID for need-xx-g/sec.

It would be interesting to know if the various MPG gauges, such as the Aeroforce, calculate from the input parameters (e.g. MAF and IAT) or from the output parameters (.e.g. Duty Cycle or Pulse Width).
I realize it is only an estimate, but it would still help tune e.g. the spark table.

I'm still recovering from surgery (prostate removal) or I would be experimenting with various PID formulas right now. Maybe in 2 weeks or so, unless Michigan winter comes too early.

joecar

There is a pid GM.INJFLOW that reports the current injector flowrate in g/s or lb/hr...

take INJFLOW (lb/hr), scale it with injector DC%, convert to gal/hr (using average density of gasoline), and divide this into VSS (miles/hr).

joecar

Ted, get well soon.

mrvedit

That sounds very reasonable. So let me take my first shot at the HPT formula:

[SENS.20]/[PID.6210]*8*[USER.9002]/(100*6.073)

[SENS.20] is Speed in Miles/Hour
[PID.6210] is Injector Flow rate in Pounds/Hour
8 is the number of cylinders
[USER.9002] is the Injector Duty Cycle in %
100 converts the DC % to a fraction
6.073 converts Pounds to Gallons of gas

In case formulas cannot use e.g. [USER.9002], it is predefined as:

[SENS.112] * [SENS.70] / 1200

[SENS.112] is Injector Pulse Width
[SENS.70] is Engine speed

So, the MPG formula using only sensors and PIDs would be:

[SENS.20]/[PID.6210]*8*([SENS.112]*[SENS.70]/1200)/(100*6.073)

Again, this is untested. I will update this post once I or someone else confirms, corrects or improves it.

I hope this isn't too far off-topic, but the purpose of Lean Cruise is to improve gas mileage and monitoring your current MPG (even if only an estimate) will be useful while experimenting with AFR and Advance.

Obviously completely off topic - A biopsy in March diagnosed me with very early prostate cancer. (I am 60Y.) Doc said it should be removed this year and I scheduled it for a convenient time, which was 2 weeks ago. If anyone is interested, I am happy to share details in the "Racers Lounge" section. I am doing very well. Thanks and sorry for the Hijack.

kingtal0n

iTrader: (1)

this portion, [SENS.20]/[PID.6210]*8 is pounds per mile total, which only awaits conversion to gallons by using, "6.073 ". The rest of the equation makes no sense (Why would we want to know duty cycle if we already have pounds of fuel per mile?)

One other thing, "6.073 converts Pounds to Gallons of gas" ->

This may be closer to 6.2 or 6.4lb/gallon and may change tank to tank, therefore making any attempt at accuracy difficult. It seems like those MPG Gauges use a constant that must be adjusted on a personal basis to become calibrated.

And last, injector pulse width, as I must have mentioned, is only part of the story. First you must know, is your computer reporting actual pulse width or corrected pulse width (including latency?) Many computers only report theoretical injection PW (They chop off the latency portion) the actual value changes based on voltage and ECU calibration, which relfects the brand of injector used and even the fuel pressure (higher pressures may slow injector latency)

joecar

INJFLOW is what the PCM looked up in the IFR table... the 8 injectors are not flowing this constantly, but rather only INJDC% of the time.

so fuel gal per hour = 8 * INJFLOW[lb/hr] * (INJDC[%] / 100[%]) / 6.6[lb/gal]

then miles per gal = VSS[mi/hr] / (8 * INJFLOW[lb/hr] * (INJDC% / 100) / 6.6[lb/gal])

removing parens = VSS[mi/hr] / 8 / INJFLOW[lb/hr] / INJDC[%] * 100[%] * 6.6[lb/gal]

joecar

Actually, looking at my old pids, I see:

{SAE.VSS.mph}/{GM.INJFLOW.lbpm}/{GM.IBPW1}/{SAE.RPM}*15000*6.073 =

or

{SAE.VSS.mph}/{GM.INJFLOW.lbpm}/{CALC.INJDC1}/1200*15000*6.073

mrvedit

kingtalOn: Yeah, when I first saw "Injector Flow Rate" I also thought it was the instantaneous amount of fuel going through the injectors. Needless to say my MPG was then a nearly constant number. And everything you say about PW inaccuracy is correct and understood.

There seems no agreement on the weight of a gallon of gas. Google "pounds gasoline to gallons 6.073" and I will see I didn't make it up. I don't doubt the 6.6 number.

To me absolute accuracy is not important, just that any increases/decreases in value reflect increase/decrease in fuel consumption.

svede1212

iTrader: (3)

With a '04 GTO it's easily turned on with HPT. This thread is funny in that LC is what got VW in trouble with their software. It could tell if the car was moving and so when being tested it didn't enable it but did on the road under normal use. The issue is higher levels of nitric oxide (NO) being formed from higher combustion temperatures.

LETZRIDE

iTrader: (6)

ya except I don't like the EPA telling me what is more harmful to the environment... they say higher NOX gasses so we should lose gas mileage.

My opinion is you're probably helping the environment more than hurting it...less gas used means less burden on the system to produce and extract oil... oil extraction and refinement produces far more emissions and damage to the environment with oil spills than a little bit more NOX gasses. Just think, if every car was more fuel efficient because of LC and everyone gained at least 12% more mpg (which is what research has shown the LC when properly tuned to gain) then we'd cut our oil consumption down by at least 12%... my personal opinion is this would overall cause less emissions to the environment given you're introducing more emissions in one step (your car) as opposed to having more emissions through multiple steps (extraction, refinement, transportation of gasoline to stations etc)

in any event- VW got caught for tests that we won't get caught with...atleast not yet. And worst case...untune it for emissions- it's one parameter to disable and takes 30 seconds.

joecar

( while we're on the subject of environment, electric car drivers fail to realize that it's someone making the carbon footprint (i.e. the power station). )

Darth_V8r

iTrader: (4)

plus the efficiency loss via transmission lines, losses charging the batteries, losses extracting energy from the batteries, and the added weight. Like keeping your fridge open to cool the room. Due to less than 100% efficiency of the fridge, you actually heat the room.

Cleaner to just run a gas engine.

Maybe not mine.....

joecar

We're a minority, our contribution is insignificant.

gagliano7

I want to see what happens when all these electric car batteries go bad and they aren't recycled properly. They already have this problem with computers and cell phones.