Page 1 of 2 12 LastLast
Results 1 to 15 of 24

Thread: SCT LiveLink - How to Log & Analyze Your Vehicle's Data

  1. #1
    Join Date
    Apr 2010
    Location
    Western Canada
    Posts
    1,088

    Lightbulb SCT LiveLink - How to Log & Analyze Your Vehicle's Data

    If you have an SCT XCalibrater 2 or X3 tuner you can use SCT LiveLink to datalog your vehicle while it is running. After, you can analyze the data to look for issues and/or to help with future tuning, whether you do it yourself or have a professional do it for you.

    These detailed data logs can help you identify fuel shortages, incorrect spark values, room for spark increases, airflow issues, and the list goes on and on. However, I have found it becomes difficult to analyze 100,000 lines of data effectively without some external help.

    In this thread I will provide a quick how-to that anyone with an SCT tuner can use to log their vehicle. If you've been curious about this but thought it was too complicated or hard, hopefully this will help you out. If you've already used LiveLink, you can skip to post 2 where I will discuss how to analyze the log files.

    So, let's hook a laptop to our Marauder using an X3 tuner:
    Step 1: Power up the laptop
    Step 2: Plug the USB cord from the handheld tuner into the laptop
    Step 3: Wait for the tuner to finish booting up
    Step 4: Plug the tuner OBD-II plug into the car's OBD-II port


    IMPORTANT: if you plug the tuner into the car first, it WILL NOT work with LiveLink. The tuner needs to be plugged first into the computer.


    To continue you need the ignition ON in your vehicle. You do not have to start it, but have the key in the ACC position.

    After starting up LiveLink, the first step is to select which data items we want to record. Under the COMMANDS menu, we select CONFIGURE DATALOG ITEMS.



    LiveLink will attempt to connect to the car ECU to identify which data items (hereafter referred to as PIDs) are available to choose.

    If you see an error message, it's because you plugged the tuner into the car first before plugging it into the computer USB, or because the vehicle ignition is not on:



    If you plugged it in correctly, you will see this:



    LiveLink will draw up a long list of PIDs that you can select from; these are the data items that will appear in your data logs:



    These are the PIDs I usually log:
    battery_voltage
    coolant_temp
    fp_duty_cycle
    FPUMP_ACT_VOLTS
    FPUMP_FLOW_RATE
    fuel_rail_pr_psi
    intake_air_tem
    KNK_SEN_RETARD
    load
    LTFT_BANK_1
    LTFT_BANK_2
    maf_counts
    PRES_DROP_F_INJ
    RPM
    spark
    STFT_BANK_1
    STFT_BANK_2
    TCC_ACT_SLIP
    TP_RELATIVE
    trans_temp


    You will find as you browse and select the PIDs, it can take some time. After we have made our selections, we will save our configuration so we can recall it again quickly. Under FILE we will select SAVE LIVELINK CONFIGURATION and save our PID list to a file.



    With the PIDs all selected, it's time to start data logging!


    To continue you will have to start your vehicle.
    You can log without starting the vehicle, but the data will be quite boring.

    To start datalogging, use CTRL+F1 on your keyboard, or select from the COMMANDS menu START DATA LOGGING.



    Here's what it looks like while logging (I did not start the car). You get a nice graphical representation of the data:


    The items that are check-marked will show up in the graph on the right. The currently highlighted one shows up yellow. The graph is useful for looking at items like fuel rail pressure that you want to normally stay constant.


    Once we're done our logging run, select from the COMMANDS menu STOP DATA LOGGING:



    Now we save the datalog to disk. From FILE, select SAVE DATALOG AS:



    LiveLink can open up data logs at any point and you can browse through the graphs in the program at any point in your datalog. This is a useful function and you can use it to check for many different situations in your vehicle. I recommend always having RPM and TP_RELATIVE or LOAD check-marked then you can identify many different scenarios (i.e., hard acceleration) and watch the data at those points.

    Continued in Post 2...
    Last edited by JoeBoomz; 06-25-2012 at 10:03 AM.

  2. #2
    Join Date
    Apr 2010
    Location
    Western Canada
    Posts
    1,088
    ...Continued from Post 1


    After sifting through thousands of lines of data, I have built an analytical tool in Microsoft Excel with which to summarize the data. All are welcome to download it here:
    http://download.ioncross.com/Datalog Analyze Spreadsheet.zip
    Here's what you get in the ZIP file:


    rawdata.csv is a sample datalog with 123,779 lines from two ~35 minute sessions combined of driving my Marauder. NOTE the charts shown below are not generated from the sample rawdata.csv file.
    Tuner v3.xlsx is the tool created to help analyze the datalog. You will need Excel 2007 or newer to use the sheet.

    IMPORTANT:
    Tuner v3.xlsx is hard-coded to read a secondary file called rawdata.csv. You will have to copy or rename your datalog from LiveLink to rawdata.csv to use it.

    To use the tool you must FIRST open rawdata.csv in Excel to get it into your computer's RAM, and then open Tuner V3.xlsx SECOND.

    If you don't open rawdata.csv first, Excel will proceed to make hundreds of thousands of hard drive reads, which will take a substantial amount of time (hours).

    At this point I will assume the viewer has a basic working knowledge of Microsoft Excel. I won't go into anything crazy, but you should be comfortable viewing spreadsheets, selecting workbook tabs, and recognize row/column references.

    Here's what the data looks like that we will get out of LiveLink:


    Column A in the datalog will ALWAYS be the time in seconds since the start of the datalog. The rest of the columns are the PIDs that you have selected to datalog in LiveLink.


    The Tuner sheet is set up to use Columns B and C in rawdata.csv as reference points for graphing.
    Specifically, in its current format, it is looking for RPM in Column B and LOAD in Column C.
    So if you want to use your own datalogs with Tuner V3, ensure you move RPM and LOAD into columns B and C as such:


    To do this, right-click on the column header you want to move and select CUT. Then right-click again to the column where you want to move it to, and select INSERT CUT CELLS. It will move the entire column to where you want it. Don't forget to SAVE the file when done.


    Now that we have rawdata.csv open in Excel, we are ready to open Tuner V3.xlsx.
    Warning! Opening the sheet might take up to a minute depending on your computer's speed:



    And once it's done calculating, here's a nice summary chart of some of our data:


    Here we have a chart with LOAD along the left (vertical) axis and RPM along the top (horizontal) access. The chart is populated by data from rawdata.csv that is averaged from all of the lines in the file that meet the values in these axis.

    Cell S3 allows you to select which data value you want to analyze. Currently we are looking at option H in the list at the right, which is FUEL_RAIL_PR_PSI.


    Let's take a look at the formula that drives this:


    In English, this is what the formula does to give you the value in cell G12:
    1. Take the LOAD value in A12, plus and minus the value B25 (so it will take LOAD values from 1.105-0.043 to 1.105+0.043)
    2. Take the RPM value in G2, plus and minus the value in P3 (so it will take RPM values from 2,550-212.5 to 2,550+212.5)
    3. Average all of the FUEL_RAIL_PR_PSI values in rawdata.csv whose LOAD and RPM values meet the criteria from above #1 and #2.

    The formula supports rawdata.csv files up to 900,000 in length.

    IMPORTANT: Tuner V3.xlsx displays AVERAGE values over the length of your datalog. This can smooth out (but will still be affected by) spikes and dips in the data.
    It is important to realize that, as you look at the chart, you could still miss important data due to the averaging.
    Consider a scenario where an alternator has an intermittent charging failure and the car voltage drops during short periods. This chart would smooth out these drops and your averages would be slightly lower, but it would be possible to miss the drops entirely without looking at the data firsthand.
    In Excel anything is possible and maybe one day I'll sit down and build some more formulas in that could look for something like this, but for now, this is a flaw in this averaging approach. The sheet does show the maximum and minimum of all of the PIDs at the right side, so you can eyeball any outliers there.


    So what does the chart say about FUEL_RAIL_PR_PSI?


    The tune that was datalogged commands a fuel rail pressure of 40PSI, and as we see in this chart, the system averages close to 40PSI most of the time. There are some slight drops along the upper load values, especially as the RPM increases. This "could" indicate a fuel flow problem but the drops are very minimal.

    So what about the values on the right side of the chart that are higher than 40? These are at higher RPM's but lower load values, which is a scenario that happens when you have the pedal down, accelerate to an RPM, then take the foot off the gas. So the engine load drops dramatically, and the fuel pressure in the system spikes momentarily as the demand for fuel in the engine drops.

    Next, let's take a look at the results from the datalog for some different PIDs.
    Again, Cell S3 allows you to select which data value you want to analyze.

    Here's the Fuel Pump Duty Cycle, which we get by changing cell S3 to "E":


    Fuel pump duty cycle is a value from 0 to 1. Normal results are in the range of 0 to 0.5, and to read the values you multiply by two and consider it as a percentage. If you see values above 0.5, they are error codes being sent from the FPDM (this is something that the averaging process will smooth out and you won't see it in the chart, however you could spot it by looking at the max values shown to the right).

    So we can see here in this chart that at idle (around 850 RPM with relatively low load around 0.170), the duty cycle is down at 0.19 x 2 = 38%.
    The highest value we see in the chart is 0.33 x 2 = 66%.
    Conclusion: there is LOTS of room left to go with this BAP-driven GT40 fuel pump!


    Let's take a look at our MAF_COUNTS by changing cell S3 to "M":


    The car's mass airflow sensor (MAF) meters the airflow into the engine and we get this resulting count value.
    As expected, what the chart here shows us is that as LOAD and RPM increases, the air into the engine also increases.

    So how are we doing for MAF counts? Well,
    According to SCT, to get good MAF resolution, your maximum MAF counts should be in the neighborhood of 80%-90% (818-920 MAF counts) of the max 1023.
    Looks like we're doing just perfect, with our maximum recorded airflow in the 800 range.

    "But Wait!" says someone smart like RacerX, "we don't have anything showing up at 5500-6000 RPM!" Yes, we don't. There is no data in the file in those RPM/load ranges. This means I set the chart up wrong, or more likely, I didn't manage to log any events in those ranges while driving.
    Consider the amount of time your engine spends in the 5,500-6,000 RPM range - it's usually for very short periods of a second at a time, if even that. Logging in these areas will be difficult unless you are doing it on a dyno.


    Let's take a look at one more PID before moving on. We'll change cell S3 to "O" to look at SPARK:


    ...and there are our SPARK values at the different LOAD and RPM combinations! Note that your final SPARK values are affected by a few different variables including air temperature and coolant temperatures. So averaging your SPARK values out of the term of the datalog might not be very accurate approach because these variables can change. But it's still cool to see!


    Continued in Post 3...
    Last edited by JoeBoomz; 06-24-2012 at 11:28 PM.

  3. #3
    Join Date
    Apr 2010
    Location
    Western Canada
    Posts
    1,088
    ...Continued from Post 2

    Whew, another long-winded post. But we're not done yet!

    Next I want to cover the bottom portion of this sheet, that has a secondary chart titled DATCOUNT.


    This chart shows you exactly how many "events" match the axis criteria for averaging. So at 0.255 load and 850 RPM, the table averaged 6,561 values! And the file we are looking at had a total of 53,628 averaged events.


    Remember to take into consideration the +/- modifiers mentioned in the formula in the above post #2. The DATCOUNT table formula looks very similar:

    ...and if you follow it closely you will see again it is adding and subtracting cell B25 from the A-Column axis numbers to form a range for LOAD and RPM before averaging the data in rawdata.csv.


    So what can we do with DATCOUNT? The values in the DATCOUNT table give you an idea of the "accuracy" of the averaging the sheet does. So with 6,000 lines of data, you know your average is more relevant than say, another cell that only has 21 lines of data.

    Cell O8, "count exclusion" is a setting that you can use to exclude results with low DATCOUNT. So if you only want to review data that has a minimum of 50 points, set O8 to 50 and the values shown in the chart will become more accurate. Points with fewer lines than 50 will be excluded from the top chart.


    Let's switch to a larger rawdata.csv file; check out the DATCOUNT chart on this one!

    You can see in this chart where the majority of the time driving is spent, where the DATCOUNT values are the highest. We also did manage some high-RPM data points in this run.


    Next, I want to point out there are four different worksheet tabs in the Tuner V3.xlsx file:



    Here's a shot of what the SUMMARY tab shows:


    SUMMARY shows a list of averages taken from the rawdata.csv in a non-chart format. I actually built SUMMARY first when looking at the files, but found it was less useful than the charts and now don't really use it. But it is a different way at looking at the data and I left it in for anyone who might like to use it.


    SmallTable is a smaller version of the BigTable tab that we have been looking at. Here's the same DATCOUNT as above, but in the SmallTable tab:


    Above are the Short Term Fuel Trims from the car's O2 sensor. You can see at the higher loads and RPM's, the air/fuel ratio is less than 1, so we have more fuel than air. This is a great way to check and make sure the car is not running too rich or lean.


    I actually built SmallTable to emulate the tables in SCT's Advantage III software, which is used for making tunes. One other side effect to using SmallTable is that each cell evaluates a larger range of the axis values (RPM and LOAD), because there are fewer cells in the axis, so you will have a different averaging than you will find in BigTable.


    You can copy/paste directly from Advantage III into this sheet, or from this sheet into Advantage III.
    With some changes to the values in the table axis, notice how similar these values start looking between our SPARK datalog and the Advantage III Borderline Knock Table:




    Notice at 0.15 load and 750 RPM, the actual SPARK values are much different than specified in the borderline knock table. This is because of the other variables that are used to calculate final SPARK, including intake air temperature and coolant temperature.


    Here's an example of changing some axis values to work with Advantage III.
    After you install a BAP, the ECU needs to re-learn what voltages to pass to the fuel pump in order to reach the commanded fuel pressures in the system. The problem is, every time you update the tune in the car, this re-learning has to happen again, unless you alter the tune for the new voltage values. The ECU starts off with a base voltage table and learns from there.
    SO, with Tuner V3.xlsx, I am able to look at the datalog, and see what voltages the ECU is using to achieve the 40PSI pressure at the fuel rails, and then paste that into the voltage table in Advantage III:


    Notice the axis PIDs are CHANGED to FUEL_RAIL_PR_PSI and FPUMP_FLOW_RATE



    With the axis values matching those of the table in Advantage III, I can copy/paste the new voltage values directly in and the car is properly tuned for the new fuel hardware.


    So I'm at the end of what I wanted to post about this. Feel free to download the Tuner V3.xlsx file and play with it. As demonstrated above, you can change axis values or even change the axis data to PIDs other than I've done here. Don't be afraid to experiment; if you screw up the sheet or formulas, you can easily re-download it.

    Hopefully some of the experts here have some datalogs they can analyze with the sheet to reveal new information about their cars that was previously unknown. If you have success with this or figure out anything else cool with it, I'd love to hear about it!
    Last edited by JoeBoomz; 06-25-2012 at 10:18 AM.

  4. #4
    Join Date
    Jun 2009
    Location
    Triangle, NC
    Posts
    439
    This needs to be a sticky.

    Thanks for the info it help me alot.

  5. #5
    Join Date
    Jun 2010
    Location
    Plano
    Age
    50
    Posts
    200
    Thanks for putting this together. I have been trying to learn more and have been playing around with LiveLink and datalogging, but there is a suprising lack of information available on the interweb (at least that i have been able to find).

    I'd buy a book if I could find one that went into detail.
    2003 MM
    1985 CJ7

  6. #6
    Join Date
    Apr 2010
    Location
    Western Canada
    Posts
    1,088
    Thanks guys, always good to get feedback.

    I wanted to add this in:

    I highly recommend using LiveLink 6.5, you can download it from http://sctflash.com/LiveLink.php

    I did some digging and educated myself a bit more about LiveLink Gen2. This was the first time I've heard of it. Good luck finding it on their site today, they seem to have pulled it down.


    Here is a download link for those who want to try it:
    http://sctflash.com/software/LiveLink_GenII.exe

    Gen2 has the addition of Histograms which is the same thing I accomplish in my LiveLink post using Excel (http://www.mercurymarauder.net/forum...ewpost&t=80544)

    Once SCT gets Gen2 compatible with all vehicles and tuners and makes it consistently available to the public, I look forward to working with it, then I can stop using my klunky Excel solution!
    Last edited by JoeBoomz; 06-26-2012 at 11:56 AM.

  7. #7
    Join Date
    Mar 2009
    Location
    North Aurora, IL
    Age
    49
    Posts
    11,836
    Darn good info as I'm not familiar with data logging.

    Thanks for posting.


    Sent from my iPhone
    Eaton Swap + 4.10's = Wreeeeeeeeeedom!!





    Great quotes!!
    Quote Originally Posted by CBT View Post
    you were too busy getting your ass kicked by cool people because you suck and are a worthless excuse for a human being.
    Quote Originally Posted by CBT View Post
    I'm going to punch you so hard between the legs your going to need a GYNA BO to fix it!!
    Quote Originally Posted by Haggis View Post
    Eat poo poo and die heathen.

  8. #8
    Join Date
    Jul 2009
    Location
    NH
    Age
    55
    Posts
    7,606
    Very nice Joe! You've been busy! How's the car running?

    P.S. Try and use PIDs in Caps whenever possible.
    Other nice PIDs depending on what yer watchin:
    Commanded Gear
    open loop flag
    Spark Advance
    Spark Source
    ISC Duty Cycle
    Analog (if you have a wideband you want to log, add equation from drop down menus)
    Last edited by RacerX; 06-26-2012 at 02:55 PM.
    11.931 @ 116.33 (MT ET Streets)
    12.073 @ 118.43 (Nitto 555)

    I love the smell of Nitromethane and rubber in the morning!
    Drag racers prayer:
    As I lay rubber down the street, I pray for traction I can keep, but if I spin and begin to slide, please dear God protect my ride." -Amen

  9. #9
    Join Date
    Feb 2011
    Location
    Asheville, NC
    Age
    53
    Posts
    2,724
    Had no idea my tuner could do this....thanks JoeBoomz!
    In Memoriam....11-14-13

    Rex Weinbender

    The best friend a person could have


    In Memoriam...05-26-14
    Lillian Kostac

    I love you grandma!!!


  10. #10
    Join Date
    Apr 2010
    Location
    Western Canada
    Posts
    1,088

    Wink

    Quote Originally Posted by RacerX View Post
    Very nice Joe! You've been busy! How's the car running?

    P.S. Try and use PIDs in Caps whenever possible.
    Other nice PIDs depending on what yer watchin:
    Commanded Gear
    open loop flag
    Spark Advance
    Spark Source
    ISC Duty Cycle
    Analog (if you have a wideband you want to log, add equation from drop down menus)
    Bill! Always great to hear from you. The car is running perfect, like a top, and I have all the kinks worked out of it now. Now it's time to set the sights on Dad's build.

    As a quick question for you - what's wrong with the lowercase PIDs?


    Anyway, it might be really "tin foil hat" of me, but whenever logging I purposefully avoid any PIDs that are speed-related or that can be used to calculate vehicle speed. If I were to end up in court for some speed-related reason, such logs would provide irrefutable evidence.

    So this means I avoid the PID for speed itself, and also any combination of RPM and current transmission gear at the same time.

  11. #11
    Join Date
    Mar 2007
    Location
    Houston, TX
    Age
    53
    Posts
    217

    Exellent work!!!!

    This is really great work. It took me almost a year to figure out what you wrote in one single post. You have given people a real jump start on the data logging and potential tuning.

    Here are a couple of things I discovered on the subject.
    1. There are PIDs that you can use to data log the fuel pump voltage and then correct them in the Advantage III table. There is a needed step to disconnect the battery to get the ECU to "accept" them better.
    2. A good way to validate the confidence of your average (Mean) is to compare it to the calculated Median value. The furhter appart the numbers are then your data may not be "normal". Statistically if the "P" value is low then your null hypothesis of the problem must go.
    Last edited by courtmarshall; 07-11-2012 at 07:04 AM.
    03 Marauder
    Silver Birch-300B
    Born 06/03
    Heated seats- Light Flint
    6 disc changer
    Silverstars
    Trilogy Kit #22
    Ford SVT Cobra Engine
    Meth Injection
    Stiegemeier S/C Porting 19.0 PSI
    Alternative Auto tuning
    K&N Cold Air Intake
    SCT 90MM MAF
    Addco front and rear sway bars
    Rossler Transmission
    Stallion Torque Converter
    24000 GVWR tranny cooler
    Baer brakes front and rear
    220 Amp police alternator (130 amp @ idle)
    "Stainless Works" Long tube Headers and complete exhaust.
    Scan Gauge II
    Proguard

  12. #12
    Join Date
    Jul 2009
    Location
    NH
    Age
    55
    Posts
    7,606
    "PIDs are lowercase. DMRs are uppercase. These log much faster, but you are limited to logging a certain amount." Eric Brooks

    "DMRs are the ones in all caps, they are more accurate and use more memory and processor than the pids which are the lower case ones." Kris@SCT

    If you have a decent laptop, you get more counts/second and more accurate data with the uppercase DMRs than with the lowercase PIDs. Me. I'm using a dedicated laptop with 8Gb DDR3, I5, 128Gb SSD drive, W7U. Your spreadsheet analyzer is nice! I use EEC Analyzer also. One key ingredient is to have good, working O2 sensors, wideband too. Yeah, I don't usually like having incriminating data running on my laptop while I log.
    11.931 @ 116.33 (MT ET Streets)
    12.073 @ 118.43 (Nitto 555)

    I love the smell of Nitromethane and rubber in the morning!
    Drag racers prayer:
    As I lay rubber down the street, I pray for traction I can keep, but if I spin and begin to slide, please dear God protect my ride." -Amen

  13. #13
    Join Date
    May 2003
    Location
    Waterloo, IA
    Age
    64
    Posts
    4,464
    This is a really good write-up.

  14. #14
    Join Date
    Jul 2010
    Location
    Portland, or
    Age
    53
    Posts
    526
    Sticky!!!!!!!! Please !!!!!! Thanks for posting!'!
    03 SB 300B

    Eaton Swap Tuned by Fast Specialities

    Trans rebuilt by Darin at BC Automotive with a Triple Disc 11in Circle D TC. Tuned by Zack. The man that made my Marauder fun again.

    4.10 gears with Cobra clutches. No more one wheel peel

    Metco Drive Shaft Loop.

    Next bigger brakes!!!!

  15. #15
    Join Date
    Jul 2009
    Location
    NH
    Age
    55
    Posts
    7,606
    Joe! You gotta try the histogram feature in GenII! I was messing around with it to get my LWFM table correct so I don't have to mess with Manifold volume for my slight tip-in issues.

    11.931 @ 116.33 (MT ET Streets)
    12.073 @ 118.43 (Nitto 555)

    I love the smell of Nitromethane and rubber in the morning!
    Drag racers prayer:
    As I lay rubber down the street, I pray for traction I can keep, but if I spin and begin to slide, please dear God protect my ride." -Amen

Page 1 of 2 12 LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. SCT iTSX Wireless Vehicle Programmer
    By NATEHAWK in forum Community Discussion
    Replies: 0
    Last Post: 04-02-2012, 02:36 PM
  2. How to data log
    By burt ragio in forum Community Discussion
    Replies: 2
    Last Post: 01-10-2011, 04:57 PM
  3. Livelink log - am I knocking?
    By Paul in forum Tuning
    Replies: 18
    Last Post: 08-22-2008, 11:28 AM
  4. Data Logging with the SCT XCal II
    By TooManyFords in forum Shop Talk
    Replies: 1
    Last Post: 03-19-2006, 05:19 PM
  5. Data Log What?
    By DEFYANT in forum Shop Talk
    Replies: 3
    Last Post: 09-04-2005, 12:48 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •