So 0 KFB is > -3Originally Posted by JustDSM
I have reached 28 degrees and still got 0 KFB. Does it mean that I can still add more advance timing?
I am surely missing something here.
One thing I noticed is that AFR went from mid 11s to high 10s at wot. Isn’t this an indication of Knock?
That's one way to work around the limitation of not being able to tune the torque management tables. You could also play with the torque based throttle limiting under "Airflow -> Electronic Throttle -> Torque Based" - you can grab the values out of one of the base calibrations as a starting point - that may also help with some of the harsh shifting under light acceleration.
Looked at your log, definitely different behavior than what I've seen, however the end effect is the same: as your Knock Feedback goes to zero, your knock learn decreases by the same amount, so Knock Learn + Knock Feedback is still the same total value. For example:
At 00:11.406, Knock Learn is 18.1, Knock Feedback is -3.0, Knock Learn + Knock Feedback is 15.1
At 00:11.626, Knock Learn is 15.0, Knock Feedback is 0.0, Knock Learn + Knock Feedback is 15.0
Knock Learn is still adapting at WOT (goes from 15.0 to 15.7 over time), and you have a few instances of Knock Feedback (ex., 00:15.786 has a Knock Feedback of -0.2), so everything is otherwise working as it should.
To answer your question on advancing the timing further: you'll need to do some more logging to see if Knock Feedback at WOT continues to learn up (setting your Learn Init higher will reduce the amount of driving you'll need to do for this) - if your Knock Learn reaches Learn Max you can bump up your timing tables. Keep in mind, even if it doesn't have knock, you don't want to target timing beyond MBT, without a dyno I'd work with the factory MBT tables ("Spark -> Advance -> MBT").
Typically, I log Knock Feedback vs Load/RPM (Knock Retard.Table.xml) to try to identify if there's a problem cell - based on your log, logging Knock Learn + Knock Feedback seems like it may be a wiser choice.
That's correct, any time KFB is < -3.0 timing is being pulled either in anticipation of, or in response to, knock
Commanded AFR dropping at WOT is typically a result of Component Protection (aka Cat Overtemp Protection or COT) - you can find the fueling tables for this under "Fuel -> Temperature Control -> EQ Ratio
Last edited by SlowNStock; 02-25-2020 at 01:19 AM. Reason: Addt'l Posts
Thanks a lot for your generous info, very much appreciated.
My Learn Init was at 19 degrees, I dropped it down to 15 degrees. I will put it back to 19 and work from there.
I will also disable the component protection to verify whether this has an effect on the AFR at wot after adding timing. will share some logging later.
Unfortunately no dynos for the moment. I am relaying on Dragy GPS to compare the timing intervals, my only reference at the moment.
Here's a stock read from a 2007 4.0 Single VVTi AT FJ Cruiser.
2007_FJ_my_stock.hpt
It should also be kept in mind that there are 3 "windows" for Knock Correct Learn. Each window is trimmed up/down independently of the others. Ie: you'll Likely see your 0-2900RPM KCL value higher than your 4500+ KCL value. You're probably spending a lot more time in the 0-2900 RPM window, under the Adaptive Learning conditions than you are above 4500RPM.
So don't be surprised to see your KCL value shift at 2900 and 4500. If there's a large delta between the windows, you'll probably see some KFB activity too if your Tundras are anything like the Tacoma. So keep that in mind.
Solid advice.To answer your question on advancing the timing further: you'll need to do some more logging to see if Knock Feedback at WOT continues to learn up (setting your Learn Init higher will reduce the amount of driving you'll need to do for this) - if your Knock Learn reaches Learn Max you can bump up your timing tables. Keep in mind, even if it doesn't have knock, you don't want to target timing beyond MBT, without a dyno I'd work with the factory MBT tables ("Spark -> Advance -> MBT").
Again, solid advice! I further that with filtering (in all cases ECT > 140deg), KCL > my Init. value and KFB > -3.0 to find where the engine is "happy". Filtering KCL < Init. value and KFB < -3.0 will show where you've got some areas that definitely need attention.Typically, I log Knock Feedback vs Load/RPM (Knock Retard.Table.xml) to try to identify if there's a problem cell - based on your log, logging Knock Learn + Knock Feedback seems like it may be a wiser choice.
The filtering can be tweaked to tailor your output for specific needs, but the above is a great snapshot.
Commanded AFR at WOT wont be affected by COT EQ Base if your Tundra is anything like the Tacoma. The COT EQ Base map is used when PE isn't enabled. Once you enter PE that map determines your WOT fueling.That's correct, any time KFB is < -3.0 timing is being pulled either in anticipation of, or in response to, knock
Commanded AFR dropping at WOT is typically a result of Component Protection (aka Cat Overtemp Protection or COT) - you can find the fueling tables for this under "Fuel -> Temperature Control -> EQ Ratio
That's correct: KFB will briefly offset the difference if you're transitioning to a zone with higher KCL from a zone with lower KCL, provided you meet Knock Adaptive Learning Enable thresholds.
Good suggestion on using KCL vs Init as a filterI further that with filtering (in all cases ECT > 140deg), KCL > my Init. value and KFB > -3.0 to find where the engine is "happy". Filtering KCL < Init. value and KFB < -3.0 will show where you've got some areas that definitely need attention.
On Tundra, fueling is determined by the richest commanded fueling - even in PE, if COT is active and commands richer fueling, it will take precedent.Commanded AFR at WOT wont be affected by COT EQ Base if your Tundra is anything like the Tacoma. The COT EQ Base map is used when PE isn't enabled. Once you enter PE that map determines your WOT fueling.
Take a look at this log - PE EQ Ratio is set to 1.20 across the board (0.83 Lambda). Even when in PE, once COT thresholds are reached, commanded AFR drops to the COT EQ Base fueling - the spark retard adder also kicks in as torque management pulls timing during shifts. Tundra_Tune_004.hpl
I can't thank you all for this post. I build Haltech kits for all of the much older Toyota's and have done some newer motors just with a stand alone not the factory ECU. I appreciate all of the info, tune files, and log setup you have all posted. I will take some time and look them all over and see the tuning strategies in your methods. It is nice to be able to tune the factory ECU's.
Thanks again guys
Are these files for stock 5.7 or for SC?
Appreciate all the great info you guys have been sharing! It's been really helpful as I try to figure out Toyota's tuning language...it's not exactly intuitive coming over from the Chevy LS world but thanks to this thread I've been able start piecing together what direction I need to be going in. I'll share my stock 2014 Tundra 5.7 non-flex fuel tune and as I learn more of what I'm doing I'll hopefully be able to contribute some tips as well...cheers!
My initial File is a TRD S/C Calibration for a 2015 5.7l
I am trying to figure out how size of injectors affect the injector constant value. I am told that stock tundras utilise 390cc injectors. So lets say that I decide to upgrade the injectors to 650cc, what is the ideal injector constant value?
In order to scale for a larger injector, you multiply the current injector constant by the ratio of the old to new injector as below:
New Injector Constant = Old Injector Constant * Old Injector Flow Rate / New Injector Flow Rate
You'll notice as you move to a larger injector, the injector constant decreases, which makes sense - a larger injector needs less time to deliver the same amount of fuel.
Keep in mind, a different injector may or may not have the same characterization data (offset, etc.); you want to make sure your seller can provide you with that data
Tundra 5,7 TRD Injector size 23250-YWF01=480cc Injector Constant=0,09838125
Lexus 5,7 Original 23250-38040=? Injector Constant=0,1316625
Counted by your example 480*0,09838125/0,1316625=358,6670464сс
Have someone precise information on the injector.
Toyota's manual for 3UR equipped vehicles spec a range of 83-99 cc per 15 seconds (332 cc/min to 396 cc/min) - the center of that range would be 364 cc/min, nearly equivalent to the ~360 cc/min per your above calculation.
Last edited by SlowNStock; 03-10-2020 at 04:24 PM. Reason: grammar
Because it's not totally clear what the Injector Constant is relative to, I went and did some logging to back out the logic; here's my current best model:
Injector Pulse Width = 10 * Cylinder Airmass * Stoich AFR * Injector Constant
Note that this calculated Pulse Width does not account for any corrections (Injector Offset, Load Mult, fuel trims, PE, etc.).
To estimate a totally unknown injector constant, you can use a formula for fuel mass:
Cylinder Airmass / Stoich AFR = Injector Pulse Width * Injector Flow Rate
Rearranging terms:
Injector Pulse Width = Cylinder Airmass / (Injector Flow Rate * Stoich AFR)
Substituting into our earlier equation:
Cylinder Airmass / (Injector Flow Rate * Stoich AFR) = 10 * Cylinder Airmass * Stoich AFR * Injector Constant
A little more algebra with the above terms, we finally arrive at:
Injector Constant = 1 / (10 * Injector Flow Rate * (Stoich AFR)^2)
Using our earlier figures for injector sizes, the injector constants calculated are ~17% too low relative to the stock values, so multiplying the calculated Injector Constant by 1.20 would arrive at a reasonable baseline (fathoming a guess, the error may be due to other corrections baked into the Injector Constant).
Whenever possible, I would scale injector constants rather than try to calculate new ones.
If anyone has better info or corrections, please share.
i have upgraded the injectors to 650cc and reduced the injector constant value to 0.063.
Fuel pump pressure is set at 60psi.
The truck idle at 20% lean. I have tried scaling the MAF without luck. This is frustrating
Could someone have a look at the attached tune file and let me know where I have gone wrong? Thanks in advance
Last edited by korbiams; 03-12-2020 at 12:49 PM.
Your injector constant looks about right for injector size and pressure, may need a little more fine tuning.i have upgraded the injectors to 650cc and reduced the injector constant value to 0.063.
Fuel pump pressure is set at 60psi.
The truck idle at 20% lean. I have tried scaling the MAF without luck. This is frustrating
Could someone have a look at the attached tune file and let me know where I have gone wrong? Thanks in advance
Injector offset ("Engine -> Fuel -> General -> Offset vs. Volts") and load multiplier ("Engine -> Fuel -> General -> Load Mult") need to be addressed - injector offset is the big factor in play at idle.
As a starting point, in the attached file I've dropped in the high pressure offset and load mult from a FFV Tundra (and set your MAF correction back to 1.00) - your seller should be able to provide you with offset data for your exact injector, the attached is just a holdover to help move things in the right direction.
2012 Toyota tundra Supercharged TRD 5.7L - FFV High Pressure Injector Offset.hpt
Found the offset data here:
https://www.deatschwerks.com/sites/w...%20Summary.pdf
The truck runs great at 9.5 Psi of boost targeting mid 11s AFR. Got to thank my mate who took me through all of the steps indeed. We kept the spark advance at 12*. I am sure that we still have 30-50 HP give or take hidden ��. Thanks once again @SlowNStock
@SlowNStock and rest,
Been trying to tune a 2016 FFV supercharged Tundra. We managed to get it almost tuned, but the truck shifts like crap specifically 2-3 shift. I am trying to force shifting at 5800 RPM, now it shifts at 6100 RPM. Also is there a way to remove rev limit or allow fuel cutout?
Attached is a log with various loads. Could you please let me know what to try in order to change shifting points?
Thanks.
Increasing the rev limit is straightforward, it's under "Engine -> Fuel -> Cutoff, DFCO", just increase to the value you'd like. Don't go crazy on the stock valvetrain...
As far as the 2-3 shift - couple things to be done.
Per your attached file, I see you've dropped your 2-3 shift in Normal and Tow/Haul, you should also drop those corresponding values in the "Cold" and "Fault" tables as well.
If the truck still shifts too high, multiply by the ratio of target/actual. In this case, with a target shift of 5800 and an actual shift of 6100, you have a ratio of 5800/6100 or ~0.95, multiply your WOT shifts by this figure, and smooth into the remaining as you see fit.
I've made those hypothetical changes in the attached as an example - rev limit is raised to 6400, WOT shift points dropped by 5%, and the resulting shift scheduling copied over to the "Cold" and "Fault" tables.
Example file: oem tundra v5.9 (2) _ EDITS.hpt
A word on your log - you'll want to pare down the channels for the specific task at hand, having the whole kit and caboodle is fine to take a high-level look, but for something like this the net polling interval is a bit slow to get an accurate picture (I realize that's probably my fault for not putting more explanation in my earlier posts). Here's an example of a smaller set of channels with better polling intervals, if you do a log with both you should notice the difference right away (you may need to fine tune polling intervals further): Tundra Trans Logging Simple.Channels.xml
Reply With Quote