Page 3 of 3 FirstFirst 123
Results 41 to 60 of 60

Thread: can't get rid of IPC and surge-mostly stock coyote-dashpot tune

  1. #41
    Senior Tuner
    Join Date
    Jan 2013
    Location
    Hawaii
    Posts
    2,101
    The + - and A buttons, there's a C button also, that shows cell count.

    Here's the minimum as well. It all indicates too much air or not enough fuel, I'm more leaning towards fuel.

    fuel trims DC filter minimum.PNG

  2. #42
    Advanced Tuner GapRider's Avatar
    Join Date
    Nov 2015
    Location
    South Carolina
    Posts
    246
    Quote Originally Posted by murfie View Post
    The + - and A buttons, there's a C button also, that shows cell count.

    Here's the minimum as well. It all indicates too much air or not enough fuel, I'm more leaning towards fuel.

    fuel trims DC filter minimum.PNG
    ah,thanks!
    Does that mean a vac leak looks less likely now because fuel trims are being made somewhat linearly from lo load to high load? I think I could hear a vac leak big enough to affect fueling at higher loads but I don't know.
    I'm out of town this weekend but I'll still do a check for vac leaks first and then bump down the inferred rail pressure and log it.
    I see two tables - a lo inferred and inferred, is the lo inferred the curve for pump at lo speed and the other one for pump at hi speed?
    2019 C7 Stingray M7 - long tube headers, 6.30/6.22 226/238 cam, supporting stuff, DOD and VVT delete.
    Stock everything else

  3. #43
    Advanced Tuner GapRider's Avatar
    Join Date
    Nov 2015
    Location
    South Carolina
    Posts
    246
    Quote Originally Posted by murfie View Post
    Torque ratio = "engine brake torque / scheduled torque"

    First you need to understand ICP error and torque ratio. Many IPC error happens when torque ratio is above 1. From the attached image you can see the worst of your IPC error is when torque ratio goes over 1, I lined up 0 IPC error with a reference line of 1 torque ratio. The ECU is thinking the engine is making too much/ more torque than it should be. You can see it tries to fix it with the throttle, when compared to your pedal. This situation happens when the throttle closes, so you see torque ratio spike but no IPC error. You can also get IPC error at less than 1 torque ratio, again the ECU is thinking the engine is making too much torque to satisfy the drivers demand.

    Attachment 87946

    If there is nothing wrong with fuel delivery, and because of your particular situation, you would possibly fix this in the driver demand. You basically have a stock engine so the stock calibration for engine parameters can be relied on as good.
    The Graph/histogram would be setup just as you see your driver demand table. engine speed for x axis, pedal position for y axis. For your pedal position being ADC, you can use voltage and convert it to ADC, but error in your math could be putting you in the wrong spot of the driver demand table. Just log trans pedal position and be comfortable in knowing you are in the same cell as the ECU. You would then filter the input by IPC error greater than 0. The input would be a user math of scheduled torque x .85 or what ever you think you torque ratio should be for the particular section of log you are looking at. This gives you the values to put into your driver demand table. A good part of what your torque ratio should be is dependent on how far your final spark advance is from MBT spark advance. I attached a graph file, you would just need to select trans pedal position, and create a user math for scheduled torque x desired torque ratio.

    The volt to ADC conversion instead of the actual logged value is throwing the cell off, but you can see from this pic what you should end up with.

    Attachment 87948

    You may notice theres some small IPC error in spots you are at constant RPM and pedal, This is when TR is low, and ECU thinks the engine isn't making enough torque for DD.

    I've gone through many tune variations trying to learn and the IPC is a common, to varying degrees, with them all. I looked back through this thread and want to log the DD graph you gave me.
    I tried to set it up but it only logs the top row for pedal position, can you look at what I've got and steer me?
    DD screens.JPG
    2019 C7 Stingray M7 - long tube headers, 6.30/6.22 226/238 cam, supporting stuff, DOD and VVT delete.
    Stock everything else

  4. #44
    Senior Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    1,089
    When the parameter used in a math is unavailable, it shows red like in your screen shot. That's why you are only getting the top row.

    Either derive pedal count from percentage (I made a thread about how to do that), or add the correct channel to your next log.

  5. #45
    Advanced Tuner
    Join Date
    Feb 2012
    Location
    TEXAS
    Posts
    631
    GAPRIDER,
    can you help me with the setup of the DDfix.xml thanks

  6. #46
    Advanced Tuner GapRider's Avatar
    Join Date
    Nov 2015
    Location
    South Carolina
    Posts
    246
    Quote Originally Posted by CCS86 View Post
    When the parameter used in a math is unavailable, it shows red like in your screen shot. That's why you are only getting the top row.

    Either derive pedal count from percentage (I made a thread about how to do that), or add the correct channel to your next log.
    thanks CCS, I repolled the car and trans pedal position didn't show up so I'll try the ADC conversion you mention and let y'all know.
    2019 C7 Stingray M7 - long tube headers, 6.30/6.22 226/238 cam, supporting stuff, DOD and VVT delete.
    Stock everything else

  7. #47
    Advanced Tuner GapRider's Avatar
    Join Date
    Nov 2015
    Location
    South Carolina
    Posts
    246
    Quote Originally Posted by mstang_man View Post
    GAPRIDER,
    can you help me with the setup of the DDfix.xml thanks
    mustang man, it was murfie the guru who came up with the xml, I still need to do the ADC conversion math like murfie and CCS talk about to get the input for my DD plot working so I'm no help on that.
    2019 C7 Stingray M7 - long tube headers, 6.30/6.22 226/238 cam, supporting stuff, DOD and VVT delete.
    Stock everything else

  8. #48

  9. #49
    Advanced Tuner
    Join Date
    Feb 2012
    Location
    TEXAS
    Posts
    631
    My other question, does anyone have good graph for using wb eq bank 1 + wb eq bank 2 / wb eq commanded i am not sure i got the math correct... i like using the hp scanner over livelink gen 2 seems faster less lag time.. Hope someone can guide me with the math/ parameter setup..

  10. #50
    Advanced Tuner GapRider's Avatar
    Join Date
    Nov 2015
    Location
    South Carolina
    Posts
    246
    Still no luck on this after many tunes. Dashpot, torque decel, much playing with torque and load tables (both ways + and -).
    Even got a new MAF.

    The problem persists so I've got a stock tune in it with adjustments for longtubes and off road x-pipe to rule out any shenanigans caused by dashpot tune.
    What is consistent is very much IPC around 3000RPM and 0.5-0.7 load and MUCH WORSE WHEN THE CAR IS COLD.

    I looked at the MAF pinout thinking something with temperature compensation wasn't happening but there's only four wires +,-,MAF sgn, and IAT,
    I logged IAF and it's steady and believable, all sensor values look smooth to me.

    Because the issue is persistent, and the response is much more temperature related and not nearly as responsive to torque table changes I'm hoping one of you guru's knows something I'm missing.
    I attached the current tune and log but I have many, many more which show about the same.

    Thanks for taking a look.
    Attached Images Attached Images
    Attached Files Attached Files
    2019 C7 Stingray M7 - long tube headers, 6.30/6.22 226/238 cam, supporting stuff, DOD and VVT delete.
    Stock everything else

  11. #51
    Senior Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    1,089
    Quote Originally Posted by GapRider View Post
    Still no luck on this after many tunes. Dashpot, torque decel, much playing with torque and load tables (both ways + and -).
    Even got a new MAF.

    The problem persists so I've got a stock tune in it with adjustments for longtubes and off road x-pipe to rule out any shenanigans caused by dashpot tune.
    What is consistent is very much IPC around 3000RPM and 0.5-0.7 load and MUCH WORSE WHEN THE CAR IS COLD.

    I looked at the MAF pinout thinking something with temperature compensation wasn't happening but there's only four wires +,-,MAF sgn, and IAT,
    I logged IAF and it's steady and believable, all sensor values look smooth to me.

    Because the issue is persistent, and the response is much more temperature related and not nearly as responsive to torque table changes I'm hoping one of you guru's knows something I'm missing.
    I attached the current tune and log but I have many, many more which show about the same.

    Thanks for taking a look.


    I feel like I have made big strides in my TQ/INV calculator recently. I went from being plagued with IPC errors on my PD blower Coyote, to having literally none.

    Here is a modification to only MP6 and MP7 TQ tables. See if it helps you.
    Attached Files Attached Files

  12. #52
    Advanced Tuner GapRider's Avatar
    Join Date
    Nov 2015
    Location
    South Carolina
    Posts
    246
    Thanks CCS! I?ll try it out in the morning

  13. #53
    Advanced Tuner GapRider's Avatar
    Join Date
    Nov 2015
    Location
    South Carolina
    Posts
    246
    I was glad to try someone else's tune, thanks so much CCS, I didn't notice a lot when comparing to all the other logs I have but wanted to show this against stock log
    stockTQ.JPGCCS-TQ-MP6MP7.JPG
    Still, I don't have control of this which is:

    • IPC errors which cause throttle corrections (closing surges)
    • centered around 3000RPM, 0.5-0.8 load
    • I always get more MAF than desired
    • worse when ECT less than 185
    • worse in 1st gear, nearly always non existent in 4th


    I've tried:
    • TQ tables adj plus and minus in parts and the whole tables, a few mapped points and all mapped points
    • I tried adjusting TQ tables alone and with complimentary InvTQ tables calculated and calculated various ways
    • InvTq, same thing as torque tables.
    • new MAF, KAM reset, short term fuel trims don't look bad
    • throttle tables
    • reduced engine inertia torque (because the IPC is worse in lower gears, gains RPMs quicker, trying to reduce any additional throttle request to account for inertia)
    • Speed density, plus and minus on engine displacement (I believe this is the base for all speed density calculations), adjusted slope parameters for the 3250RPM row because it sticks out in relation to the RPM rows above and below it. Did this for MP3-8 I believe.
    • Toggled different combos of Oscillation switch, Torque Intervention Switch, Tip management.


    I don't know what other area of the tune I can try, I can make the problem a little worse maybe but I haven't been able to make it better
    Somehow, I need the car to expect the amount of MAF it is getting (DesMAF close to ActMAF) . It doesn't even have a power adder, just longtubes.
    2019 C7 Stingray M7 - long tube headers, 6.30/6.22 226/238 cam, supporting stuff, DOD and VVT delete.
    Stock everything else

  14. #54
    Senior Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    1,089
    Yeah, you have something funny going on, to be running that close to a stock tune and stock engine, but have such serious TQ errors.

    I wonder if something is getting flaky with your TPS or TB drive.

    You could pop off the intake tube and with the engine off / key on, video the throttle blades while you linearly roll onto and off of the pedal. Make sure the motion is smooth and matches your inputs and logged angle values.

  15. #55
    Advanced Tuner GapRider's Avatar
    Join Date
    Nov 2015
    Location
    South Carolina
    Posts
    246
    Thanks for the sanity check. I did watch the throttle the other day while my wife cycled the pedal and it looked smooth and ?normal?. I will try TPS next even though it seems smooth but the car only has so many sensors

  16. #56
    Advanced Tuner GapRider's Avatar
    Join Date
    Nov 2015
    Location
    South Carolina
    Posts
    246
    SOLVED - thank God, all who helped and HP Tuners.
    I will post an explanation of what I've learned later but for now just to let everyone who helped on this know that it's fixed.
    So grateful for HP Tuners speed density calculator, what a great tool.

    Recapping, I was plagued with bad IPC errors centered around 0.5 - 0.8 load and 3000RPM, worse while car was still not up to temp. MAF > DesMAF
    I adjusted tq / invtq tables all sorts of ways to no avail.

    The fix - I started adjusting speed density map and the car responded in a way that made sense.
    Simplified, I increased the map in the troublesome area and the IPC went away.
    I'll show more detail later.
    2019 C7 Stingray M7 - long tube headers, 6.30/6.22 226/238 cam, supporting stuff, DOD and VVT delete.
    Stock everything else

  17. #57
    Senior Tuner CCS86's Avatar
    Join Date
    Nov 2017
    Location
    Austin
    Posts
    1,089
    That's awesome man.

    I think that there is no one answer for why IPC errors occur. Possibilities include (but not limited to):

    Inverse values being off
    Relationship between TQ and inverse wrong
    ETC table errors
    (as you found) Bad SD data
    Mechanical issues, especially with TB or MAF

  18. #58
    Advanced Tuner
    Join Date
    Jul 2017
    Posts
    529
    Where exactly is this speed density map?
    I have a speed density calculator, but no actual map.

    I had an airbag replacement today, had to drive my 2010 GT500 for over an hour, all stock, so I decided to log it to see errors & such, a good bit of ipc errors on this one, so they may be normal to some extent.

  19. #59
    Advanced Tuner GapRider's Avatar
    Join Date
    Nov 2015
    Location
    South Carolina
    Posts
    246
    I learned a lot through this and wanted to share. Sorry if this explanation has been posted elsewhere.
    Speed density is a numerical model that depicts the cylinder filling characteristics for the engine through the range of RPMs and intake pressures (MAP).
    Since these engines have 14 cam positions, it's like 14 different engines as far as that model goes - an SD map for each cam position.
    All based on standard temperature, variations in temp are compensated for.

    This much MAP relates to that much cyl mass for the RPM you're at. Or, this RPM, MAF and mapped point mean this much inferred MAP

    Speed density used to be expressed as a linear equation and gave you a straight line, this is for each RPM.
    SD for a given RPM - linear calculation.JPG
    These cars express speed density using a quadratic equation. It's better because those equations define a curve which closer matches what's real in the engine. The polarity of the quadratic term defines the up or down shape of the curve, you see both in the tune. I've read somewhere that the computer can ignore the quadratic term at times and rely on the linear equation.
    SD for a given RPM - quadratic calculation.JPG
    The offset, slope and quadratic parameters are this map for each mapped point.
    When you set all those curved lines for each RPM in a table you get a 3D map.
    This for one mapped point. Notice the ridges and valleys, that shows that the cylinders fill differently depending on the RPM. If this were a fixed cam engine it would be that one map.
    Since these cars have 14 cam positions, you will have 14 maps and they look different for each mapped point.
    SD map for one mapped point lines for each RPM.JPG
    In HP Tuners software, you can see this for every mapped point.
    SD-MP6-calculator numbers view.JPG
    Click on the 3D map view and you see the map
    SD-MP6-calculator map view.JPG
    I maximize the view to see it better.
    Select the mapped point you want to adjust, use the numbers view and change your numbers.
    Go back to the 3D map view, smooth, extrapolate etc how you want.
    Click the airflow coefficients button and it calculates all the new parameters for that mapped point.

    In my situation, I worked on MP5-8 first. I noticed there was a trough at the 3250 RPM range in the maps.
    First I smoothed the trough between about 2250 and 4500 RPM and logged again, IPC went way down. Yay
    Then I increased the 3000RPM row by 5% and smoothed again. IPC great but getting these IPC spikes in the lower RPMs now.
    I made similar but smaller changes to all mapped points and IPC is great, no more surging, just great to drive.
    I logged IPC in a histogram set on max, not average which always was near zero.
    SD-MP6-stock.JPGSD-MP6-mod.JPG
    These are MP6 - left is stock, right is modded. If you can see the ridges in the stock map and smoothed in the modded map.
    My thought is that especially in the 3000RPM midload situation, the stock exhaust didn't do as good a job at evacuating the cylinders as neighboring RPMs, hince less fresh air charge entering the cylinders.
    I have headers, no cats and 3" exhaust on the car, I was getting more MAF than the ECU expected, the SD model changes showed the ECU the engine would flow more now at that target range.
    Errbody happy.

    I think the whole torque control thing goes like this: (simplified, no nannies that I don't understand involved)
    pedal -> torque(DD table) -> load(torque to load table) -> load & RPM [speed density model] -> requested airflow -> [throttle tables] -> X amount of throttle opening.
    If actual MAF is greater than desired MAF correction takes place.

    I believe the reason I had to change the speed density model to fix mine is that the model was now wrong because of better exhaust evacuation of the cyls (more fresh air)
    If I'd added a supercharger instead, I believe the speed density model would not have needed changing because a blower doesn't affect how much air enters the engine for a given MAP [speed density model]
    Different cams, heads would.
    I'm just a guy with this one Mustang, not a tuner, so take it for that.
    2019 C7 Stingray M7 - long tube headers, 6.30/6.22 226/238 cam, supporting stuff, DOD and VVT delete.
    Stock everything else

  20. #60
    do we have to chane the SD map AND change torque inverse too ?

    or only SD map ?