Results 1 to 7 of 7

Thread: Scanner Logging Interval Explanation / Optimization

  1. #1
    Advanced Tuner
    Join Date
    May 2014
    Location
    USA
    Posts
    444

    Scanner Logging Interval Explanation / Optimization

    Hey, I was looking for a little background info on how the scanner works in order to help me get the most out of it.

    Car is a 2017 Audi S3, which currently only supports mode $22 scanning, in addition to the normal SAE stuff of course. I am also logging from both the ECM and TCM.

    First I exported a CSV with interpolation disabled so I could see exactly when each parameter was updated. I put this in a spreadsheet vs selected interval rate in the software.

    When I broke these out by SAE PIDs vs ECM $22 vs TCM $22 I started to see a pattern as to the ones that were hitting target logging rate vs slowing down quite a bit.

    I'll say up front, yes I am logging more stuff than I need to right now. Been logging extra things to see what exactly everything is, what might be useful etc. I know by simply logging less stuff the rate will go up.

    Ultimately, the ECM $22 stuff is mostly what I need, and the 10hz ones are bogging down pretty good, only hitting around 3.5hz on average.

    Is my best course of action to simply remove all the stuff I don't need and/or reduce less important things down to say 5 or 1hz?

    Will I see a speed boost in general if I don't log ANYTHING from the TCM? With bluetooth logging with Torque Pro or similar on a phone, by simply logging a combination of Mode $01 and Mode $22 or ECM and TCM I saw a big hit in overall performance. Increase in "overhead from changing modules/modes" as it was described to me.

    Would I see a performance boost if I tend to use more SAE pids and "even out" the Mode $01 vs Mode $22 ? For example, only use SAE ign timing, and SAE engine RPM (even though SAE is 2 bytes and Mode $22 is 1 byte), etc.

    Not complaining about anything, I understand this car still lacks a true high speed logger at the moment with no Mode $2C stuff-- I am simply just trying to make the most of what I have in front of me.

    In this log, I am getting an overall logging rate of about 150 PID/s.... I am just a little confused as to the under the hood operation. Why are SAE and TCM 10hz commanded items achieving target, at the expense of the Mode 22 stuff? Before digging deep into this data, I might have expected to the 10hz params to all be hitting around 5-6hz.... not some are chugging along at 10hz and others suffering even more at their expense.

    Thanks

    Last edited by aaronc7; 12-26-2020 at 06:15 PM.

  2. #2
    Advanced Tuner
    Join Date
    May 2014
    Location
    USA
    Posts
    444
    OK little update.

    I did a test where I removed some and all of the TCM PIDs to see its effect on logging speed. Plus, 4 of them were being logged at 10hz and a few at 5hz etc.

    Column 1 is "original"/no change since OP. Actual logged rate for each param and total "PID/s" at the bottom.

    Column 2 is all but 1 of the TCM PIDs removed.

    Column 3 is all of the TCM PIDs removed.

    No change to the logging of the SAE and ECM Mode $22 PIDs.... what gives? And overall PID/s rate went from 150/s to 100/s.

    I don't think this can be expected behavior... logging several less things and eliminating logging from one controller entirely doesn't result in faster logging for the ECM Mode $22 stuff?


  3. #3
    Advanced Tuner
    Join Date
    May 2014
    Location
    USA
    Posts
    444
    Well, I figured out why it was behaving the way it was. No idea if this applies to other platforms but....

    Let's say for example I am logging 4x Mode $22 params and 2x Mode $01/SAE params. I select the same logging rate for both, let's say 50hz or whatever.

    HPT Scanner will put the Mode $22 and Mode $01 params into separate 'piles'. And then it will request the params like this... param 1 from pile 1, param 1 from pile 2, param 2 from pile 1, param 2 from pile 2, etc.

    Mode$22 PID 1
    Mode$01 PID 1
    Mode$22 PID 2
    Mode$01 PID 2
    Mode$22 PID 3
    Mode$01 PID 1
    Mode$22 PID 4
    Mode$01 PID 2
    Mode$22 PID 1
    Mode$01 PID 1
    Mode$22 PID 2
    Mode$01 PID 2

    etc

    So, as a result, the Mode$01 PID will get polled twice as often as the Mode $22 stuff. And of course, Mode 22 is all the really useful stuff....but I still need a few SAE (fuel trims, e content, etc.), which really slows the mode 22 rate down.

    Possible solution is setting Mode 22 items to a polling rate 2x as fast the SAE ones.

    But anyways, the way it is now is not really intuitive and what logging rate you select may not actually represent how fast the logging rate will be (relative to other stuff).

    Attached is an HPT log, exported csv (no interpolation so you can see when things are actually getting updated), and a CAN dump log (which is what I used to really figure out what was going on).

    https://drive.google.com/drive/folde...dq6f16zX4S030k
    Last edited by aaronc7; 01-30-2021 at 11:05 PM.

  4. #4
    Senior Tuner 10_SS's Avatar
    Join Date
    Mar 2012
    Location
    Detroit, MI
    Posts
    1,320
    Look at your "arrows" in channel list, One arrow to the right means those are broadcast, meaning HPT is just listening to them, whereas the others (mode22) are polled. Your $01 channels almost appear to be the common broadcast PIDs. Also sometimes you can get a broadcast channel instead of the two arrow polled channel for the same item, which can help speed up the polled PIDs since there will be less polling going on.
    2010 Camaro LS3 (E38 ECU - Spark only). MS3X running complete RTT fuel control (wideband).
    Whipple 2.9L, 3.875" Pulley, kit injectors, supplied MSD Boost-A-Pump, stock pump
    LG Motorsports 1 7/8" Headers - No Cats, stock mid pipe with JBA Axle Back
    ZL1 Wheels/Tires

  5. #5
    Advanced Tuner
    Join Date
    May 2014
    Location
    USA
    Posts
    444
    Quote Originally Posted by 10_SS View Post
    Look at your "arrows" in channel list, One arrow to the right means those are broadcast, meaning HPT is just listening to them, whereas the others (mode22) are polled. Your $01 channels almost appear to be the common broadcast PIDs. Also sometimes you can get a broadcast channel instead of the two arrow polled channel for the same item, which can help speed up the polled PIDs since there will be less polling going on.
    No broadcast params are available on my ECU, sadly, I even double checked in an effort to speed things up.

    What's funny is on this ECU, all SAE / $01 PIDs are available on $22 as well. Just F4 in place of 01. For example 01 0C = 22 F4 0C.

    Maybe in the future they can make this change for this particular ECU so all the params get put into the same "pile" or whatever and you get the expected polling results.

    For now, I am generally able to select about 10 SAE PIDs and 10 Mode 22...so the polling rate is 'even' / 'about right'.

  6. #6
    Senior Tuner 10_SS's Avatar
    Join Date
    Mar 2012
    Location
    Detroit, MI
    Posts
    1,320
    Quote Originally Posted by aaronc7 View Post
    No broadcast params are available on my ECU, sadly, I even double checked in an effort to speed things up.

    What's funny is on this ECU, all SAE / $01 PIDs are available on $22 as well. Just F4 in place of 01. For example 01 0C = 22 F4 0C.

    Maybe in the future they can make this change for this particular ECU so all the params get put into the same "pile" or whatever and you get the expected polling results.

    For now, I am generally able to select about 10 SAE PIDs and 10 Mode 22...so the polling rate is 'even' / 'about right'.
    You can usually request the same 01 PIDs as Mode 22, that part is consistent with GM ecu's anyway. For 1byte PIDs in mode 22 I would enter 22 00 0C, but the 2 byte PIDs fill in the 00 spot, that's the only difference.
    2010 Camaro LS3 (E38 ECU - Spark only). MS3X running complete RTT fuel control (wideband).
    Whipple 2.9L, 3.875" Pulley, kit injectors, supplied MSD Boost-A-Pump, stock pump
    LG Motorsports 1 7/8" Headers - No Cats, stock mid pipe with JBA Axle Back
    ZL1 Wheels/Tires

  7. #7
    Advanced Tuner
    Join Date
    May 2014
    Location
    USA
    Posts
    444
    Ah, good to know. I think this particular car/ECU is just new/not as mainstream as the GM stuff for HPT, so the first implementation was pretty basic is what it all comes back to.

    I sent them a huge mode 22 PID list, just hoping they can add them soon .