trash site

[QTR FMS]

Have you tuned your spool ignition timing tables to help with the tip in knock?

You are hitting boost target in that log through the whole RPM range. Am I missing something?

i have retarded my global timing using temp correction didnt try changing the spool specifically, the reason im doing it this way is because i have lots of knock everywere. especially from 1000-3000 at every load even off-idle, maybe its phantom knock or my fuel is really bad.

my boost target is flat 19 but as you can see in the log its curved it start from 15 to 18 to 19 based on rpm.

Sponsored

 

[razorlab]

i have retarded my global timing using temp correction didnt try changing the spool specifically, the reason im doing it this way is because i have lots of knock everywere. especially from 1000-3000 at every load even off-idle, maybe its phantom knock or my fuel is really bad.

my boost target is flat 19 but as you can see in the log its curved it start from 15 to 18 to 19 based on rpm.

I would advise not to tune ignition timing that way. Tune it in the actual tables and spool tables for tip-in.

As far as boost target, did you set S_PLDSOLL_LLKNDK to 0?

 

[QTR FMS]

I would advise not to tune ignition timing that way. Tune it in the actual tables and spool tables for tip-in.

As far as boost target, did you set S_PLDSOLL_LLKNDK to 0?

no thats probably it, i will set it to 0 and try.

for the actual tables do we know how does the tables are interpolated? is there a channel to log them, and what is the difference between table 1 and 2, is it based on knock interpolation or vanos or sport normal mode? and when does the spool table start and ends.

 

[razorlab]

no thats probably it, i will set it to 0 and try.

for the actual tables do we know how does the tables are interpolated? is there a channel to log them, and what is the difference between table 1 and 2, is it based on knock interpolation or vanos or sport normal mode? and when does the spool table start and ends.

Path 1 = when everything is running normal
Path 2 = Persistent knock, fuel quality/octane index table gets tripped, high charge temp, etc

There are a LOT of ignition timing maps, most aren't defined in the public MHD XDF. I believe there are 70+ tables. Main variants, Warm, Cold, Best, Optimal, Minimum, Component Protection, Catalyst Heating, Vanos, Spool, Miller, Knock Protection, etc.

Here are some beyond what is in the public MHD XDF that I have defined, but still not all of them:

 

[QTR FMS]

I did set S_PLDSOLL_LLKNDK to 0, and it didn't fix the issue, so I scaled down my load to torque from 600nm to 500 nm at WOT, and that fixed the issue, so I guess there is an undefined torque table.

I pulled 10 degrees off the main and spool table to test, and I found that I still have a light knock at high load. I'm gonna try running Octane Booster or C16 and desensitizing the knock sensors.

does anyone know what is the row axis for this table? im thinking of reducing the values but keep it progressive

 

[razorlab]

I did set S_PLDSOLL_LLKNDK to 0, and it didn't fix the issue, so I scaled down my load to torque from 600nm to 500 nm at WOT, and that fixed the issue, so I guess there is an undefined torque table.

I pulled 10 degrees off the main and spool table to test, and I found that I still have a light knock at high load. I'm gonna try running Octane Booster or C16 and desensitizing the knock sensors.

does anyone know what is the row axis for this table? im thinking of reducing the values but keep it progressive

Oh I just realized you are a MT, yes, there are a bunch of torque limit tables for MT that aren't usually defined. Can you remind me of your BIN # again?

 

[QTR FMS]

Oh I just realized you are a MT, yes, there are a bunch of torque limit tables for MT that aren't usually defined. Can you remind me of your BIN # again?

00005D55466408

If you can also define I-gain and the minimum intervention table i would be grateful

 

[razorlab]

00005D55466408

If you can also define I-gain and the minimum intervention table i would be grateful

Here you go, they are all in the limits section.

I-gain I will find this week.

 

[QTR FMS]

Here you go, they are all in the limits section.

I-gain I will find this week.

thanks, i believe the address is wrong for minimum intervention. the limiters looks fine so far.

 

[razorlab]

thanks, i believe the address is wrong for minimum intervention. the limiters looks fine so far.

Yup, correct. I haven't found them in that BIN yet.

 

[ColonelAdama]

I found a table today that interests me a lot in fixing MT shift knock.

So I believe I have figured out what causes it. It is EZA (Einspritzausblendung) aka Fuel cut-off. It is a tq management path for big changes in torque when it cannot be handled by ign reduction alone. You can tell this is active when rail pressure spikes for a bit.

I noticed in all my logs that injector duty (IDC%) would drop sharply the moment that I released the clutch and touched the gas pedal. https://datazap.me/u/coloneltuned/l...s=50x50&sc=15-17-36-37-38-39-40-41-49&solo=15

I believe this to be part of the UESA (transition) path of EZA which stops injection on 3 cylinders Used to reduce torque and not full 6 cyl fuel cut off like decel. IDC would be close to 0 in full cut-off decel, whereas I am seeing IDC at 30% time of knock, about half of what it should be.

I found a table in @razorlab 's S58 A2L that is the following:

BMWtqe_st_LimSrcDfcoTranProh_C

VALUE

Codewort, um einen Wechsel von TL nach UESA whrend eines noch aktiven Momenteneingriffs (BMWtqe_st_tq_LimSrc_ul) zu verbitten

Code word to prohibit a switch from TL to UESA during an active torque intervention (BMWtqe_st_tq_LimSrc_ul)

Basically a switch to disable this exact logic.

"Code word to prohibit a switch from TL (torque limiter) to UESA (3 cyl fuel cut off) during an active torque intervention"

Just intended to smooth out a gear change but ends up causing knock (or false knock). Likely some sort of anti-jerk or flywheel protection torque reduction that dme decides cannot be handled with ign pull alone, and needs partial fuel cut as well.

There should absolutely be no fuel-cut at all the moment I hit the gas pedal, so I would like to define this table and test. There should be no harm in unburnt fuel out the exhaust if catless.

@razorlab - these codewords are like impossible to match in hex to the S58 A2L, any way I could get your help? and confirm if my logic is sound

Thanks!


PS - I did set BMWtqe_st_ProhCylOffGc_C to 1 and it did not solve it Would love to just do away with all fuel-cutoff(EZA) on gear change if possible (should have been this table...)

 

[BumbleBoi]

Hey guys, I have been learning to read logs and figure out MHD tuning. So thank you all for the amazing write ups!

I'm using ColonelAdama data log in the last post to learn, and I have a few questions.

MHD+ WGDC Base (%) is the commanded duty cycle in the bin file right?
Then is WGDC 1(%) the actual duty cycle after accounting for PID gains and other tables that affect waste gate duty cycle?

If the above questions make sense, how come the WGDC 1(%) in this boxed region reading 97.5 when the MHD+WGDC Base(%) is only reading 22.2 and the P and D gains are not that high?

 

[razorlab]

Hey guys, I have been learning to read logs and figure out MHD tuning. So thank you all for the amazing write ups!

I'm using ColonelAdama data log in the last post to learn, and I have a few questions.

MHD+ WGDC Base (%) is the commanded duty cycle in the bin file right?
Then is WGDC 1(%) the actual duty cycle after accounting for PID gains and other tables that affect waste gate duty cycle?

If the above questions make sense, how come the WGDC 1(%) in this boxed region reading 97.5 when the MHD+WGDC Base(%) is only reading 22.2 and the P and D gains are not that high?

Yes, that is correct. There are also spool tables and base WGDC tables. Looks like you are off throttle so it's probably using these tables:

 

[ColonelAdama]

Hey guys, I have been learning to read logs and figure out MHD tuning. So thank you all for the amazing write ups!

I'm using ColonelAdama data log in the last post to learn, and I have a few questions.

MHD+ WGDC Base (%) is the commanded duty cycle in the bin file right?
Then is WGDC 1(%) the actual duty cycle after accounting for PID gains and other tables that affect waste gate duty cycle?

If the above questions make sense, how come the WGDC 1(%) in this boxed region reading 97.5 when the MHD+WGDC Base(%) is only reading 22.2 and the P and D gains are not that high?

Correct, partial load tables transition to MHD+ custom boost control around 80 load and up.

Here's my partial load and blend table:

I found the first table to be well worth adjusting, as the car would sometimes boost creep up to 10 psi at low load. Despite target at 0.

The fader is less well understood. I made it a bit more linear but hard to detect the change in a log.

 

[razorlab]

I've been working on building a Datalog and tune analyzation tool for myself based on what I find lacking from other tools.

Still in early stages but it's already proving to be super helpful.

Some features:

1. Visualize datalog with peak numbers, visual representation of WOT sections (great for track logs), visual representation of gear, etc
2. Trace feature on datalog. Load datalog, load XDF file, load BIN file and then you can scrub over a datalog and it will trace it in the selected DME table.
3. Tune analyzer which shows you any Torque or Fill Limits in the log and shows which DME tables effect that, and gives suggestions of changing those tables.
4. Fuel Scalar, WGDC Base, P-Gain, I-Gain, D-Gain tune analyzer that shows current table values, then offers suggestions based on the datalog loaded. For example, with Fuel Scalar, it analyzes STFT + LTFT and shows what to change on the Fuel Scalar table to get fuel trims closer to zero.

The peak numbers are handy to quickly see max values from the datalog. Example, in this track datalog, I went WOT 52 times in a 21 minute track session.

Quickly see WOT sections of the datalog with visual representation:

Visual color for each gear:

DME table cell tracing. Scrub over a datalog and it will trace it values in any DME table:

Fill and Torque limits found in a datalog, which limits, what they are, and what tables to modify, with suggestions:

DME table analyzation from datalog and suggestions:

Sponsored