Page 2 of 2 FirstFirst 12
Results 21 to 27 of 27

Thread: octane mod

  1. #21
    Senior Tuner
    Join Date
    Jan 2013
    Location
    Hawaii
    Posts
    2,101
    With the exception of startup, Idle, and the many limits, the basic principle behind spark control is: And this can be applied to nearly any vehicle if you tune other platforms, they just don't lay out the why as clear as ford did.

    Base timing is determined off two main things. Borderline or MBT which ever is lower is the timing that the engine will get. MBT is minimum spark advace for best torque, if you are at this value the engine will be making the most torque for the given amount of airmass its getting in the cylinders. Its your "Goal" not a limit, any more advance and you will not get more torque, and probably make less if way far past it. It is spark source "base" because the timing is at the minimum point for best torque. The only reason you would want less spark, would be to control torque output, idle control, or other limits and those are all different sources with their own defined criteria, but the same logic of the source with the lowest value is used.

    Borderline and MBT have "starting point" tables. Borderline has borderline corrections, MBT has MBT corrections. They are clearly separated into sections in the editor by underlines of the headers. If compared correctly you can find the crossing point where borderline drops below MBT, usually around .3-.4 load at low RPMs and a bit high as the RPMS go up.

    In the editor you will see a section underlined as base and base corrections. This has to do with cold startup timing and not MBT.

    The spark retard logic can both add and remove. It has enable values just like the knock learning, so its not active all the time, just most of it. It either adds in from the borderline and gets to the MBT value or until it finds the threshold where the engine detonates. Detonation isnt exactly a hard border, but has levels too it. Adding slowly allows the computer to not cross past the light threshold(pulling .5* per knock event) into the harder thresholds(pulling 1*+ per knock event). Adding 8* in a second is pretty fast. adding 2* in a second is much slower. So you want to rely on this system as little as possible, thats why there's the correction tables and this octane adjustment logic. This is also why I say add to your borderline tables, and reduce the speed at which your knock advance comes in so it gets you closer to the knock threshold, but doesnt jump you over it into hard detonation causing the retard to pull more timing out than when it had to add in 8*.

    If your borderline corrections are setup right they will help this logic not have to work as hard. If they are setup conservative or aggressive this logic can make up for it as the system is searching for borderline threshold or MBT, but instead of the spark changing because of air temp or coolant temp it will think its an octane difference thats doing it. Remeber the knock logic still has the ability to add and, most importantly, remove like 10* unless you changed that. The OAR(octane modifier) is just another borderline correction.

    MBT is only corrected by lambda, because it effects the flame propagation speed that potentially changes torque output.

    So be confident your system is safe working toward the borderline threshold of knock or MBT. The only exception is if the hardware gives out before the borderline threshold, luckily thats pretty high even for stock engines.
    Last edited by murfie; 06-18-2020 at 04:30 AM.

  2. #22
    Murfie thank you for taking the time to respond.

    That helps me visualize, and clarifies the table structure. I believe what I will be testing will work well for me. I kept the ecu ability to pull 10 degrees, but it never does. I have ran 93 and spent a great deal of time after fixing the SD to get my borderline tables setup to where little to no knock activities occur. When it does its typically between .25 and .5, not repeatable, and averages over a couple runs to less than .1 to .2.

    I cant do much concerning MBT tables or corrections, I have left those stock. A very well known tuner told me they were off, but I also think he wanted an excuse to get out of the remote tune as he was struggling, and I know he didn't want to touch the SD which is what actually helped the calibration more than anything else.

    My desire is to use OAR so the car detects better gas, or additives that add octane and uses best timing values for its initial values. This was so I could use relatively conservative values in the advance section and not have the car try so hard to ride the knock threshold. Then on a case where I might have older or bad gas pull some timing. I monitor knock during all drives so I shouldn't be caught off guard unless a sensor goes out.

    If I find my learned octane value on 93 is not in the center of the range I would then need to identify if the borderline tables needed adjustment, or the modifiers. The other option would be to disable this and calibrate the Borderline on a known fuel?

    I am worried I calibrated the borderline tables with this active, and it was adding timing at all times. All my logs show the learned value was -.5 . I will need to go back and review the values in the other tables when I was logging.
    Last edited by superman07; 06-18-2020 at 10:48 AM.

  3. #23
    Senior Tuner
    Join Date
    Jan 2013
    Location
    Hawaii
    Posts
    2,101
    Where the lambda correction is 0, at 1.0 stoic, that's where the ecu thinks the border line tables are calibrated to.

    Where the Iat correction table is 0, at 70f, that's where the ecu thinks the borderline tables are calibrated to.

    Etc correction 0 at 200f, same thing.

    To calibrate to a known octane, you would not only have to have a tested fuel octane, pump gas is a no let than x deal leaving room for inconsistency, but keep all the corrections in mind as well.

    It's also very common for 93 pump gas to get you very close to MBT, Ford knew people would be running that, so that's how they designed the engine physically. They calibrated it to 91, with the ability to go down to 87 or up to 93.

  4. #24
    Tuner
    Join Date
    Nov 2018
    Location
    Georgia
    Posts
    81
    Quote Originally Posted by superman07 View Post
    So if I understand correctly, and maybe Murfie will jump back on.

    The success of using this feature to me relies on really good base calibration and table values with correct modifiers for fueling and temps. Otherwise learned knock may be useless to harmful. If you start with a poor calibration, this is just one more thing to confuse you or make he car drive worse. Setting it up wrong could add in place of removing timing, and get you a trophy (Greg Banish).

    Your base timing will adjust based on Borderline, Lambda, and torque requests. If within your driving patterns the ECU sees knock, or is adding timing due to no knock it uses the knock learned correction vs Octane table. This table should be setup to ensure that the range of timing advance or retard properly resolves to a correct Octane Modifier. For my test I am scaling from 1 to negative 1. Because this modifier works with a table with negative numbers for retard a positive value will resolve in less timing commanded, and I assume visa versa.

    I set my max and min modifier values to 1 and -1 respectfully - to match the above table.

    I have the clip set to -1 for all rpm ranges - this should afford the ECU the maximum ability to add spark.

    I set all enable flags to basically turn this on for all rpm / load ranges the car will see.

    Time, constant, and transient values left stock as I have zero clue how they work.

    because my original tables only allowed for .5 and -.5 respectfully I lowered by approx .5 the values in the octane adjust table under the advance tab. Again all values here are negative values. At maximum values for learned knock, or advance I am only giving the car the ability to add or remove 2 degrees in most areas and 3 in others.

    I believe my base tables are pretty good, and especially after dialing in SD to keep me in the correct cells I am going to try this. My hope is the learned modifier value is close to zero running 93 as that is my typical fuel, with bad gas the modifier should go positive, with additives like boostane it should go negative.

    The original calibration stayd locked at -.5 so I will probably need to add some spark slowly in some of the borderline tables to ensure clean 93 results in 0.

    After playing with it I might Increase the values in the Spark knock adjustment / load table.

    What I dont know is if or how any of this effects the maximum advance table.
    my understanding is it simple adding to boarder-line spark. I have found I do have to drive about 20 miles to get a full learn.

  5. #25
    Tuner
    Join Date
    Nov 2018
    Location
    Georgia
    Posts
    81
    Screenshot (11)_LI.jpg

    also looks to learn from +knock correction / -knock
    Last edited by 18f1505.0roush; 06-19-2020 at 08:54 AM.

  6. #26
    Senior Tuner eficalibrator's Avatar
    Join Date
    Mar 2006
    Location
    Detroit
    Posts
    1,023
    Quote Originally Posted by superman07 View Post
    SSetting it up wrong could add in place of removing timing, and get you a trophy (Greg Banish).
    At least someone pays attention when I crack a stupid joke in class.

  7. #27
    Quote Originally Posted by eficalibrator View Post
    At least someone pays attention when I crack a stupid joke in class.
    Its 2021 even losers get trophies............

    When is the new Ford stuff coming out?