RomRaider

Hey mate thanks for the reply.. I think ill start my own thread here and put my head on the chopping block for all to see lol...
Just woken up so quick reply.. Yes ECU just been reset... the newer log im playing with a manual boost control .. i hit 39psi so minus 14 puts it in the low to mid 20s.. (it will be coming off the car)
I was logging all the sensors and volts so i could compare there readings to the scaling seeing how accurate they are..
the knock was from silly driving..into wot suddenly off throttle and back on...

Here is my stock rom.. AFR open loop gets down into the 8's.. i did think that was odd.. maybe this isnt a stock ecu to begin with.. later i might track down a stock one from the internet and compare..

I can resale the wideband.. low is set to 9.6 and high 22 AFR dont see why i need much above 18...

Ive been searching for a good what to log for what you wanna tune explanation... yet to find one..

Stock ecu read

take a look here for the stock roms :
https://www.romraider.com/forum/viewtopic.php?f=35&t=7591

I pretty much never get knock regardless of how I am driving it but I have all my closed loop to open loop transition counters zero'd out so it easily jumps into open loop.

There are a few main tables, fueling, turbo dynamics and wastegate duty along with commanded boost, and timing base/advance. Just look at the axis and cells on these tables to figure out what you should log in order to find the problem cells and adjust them.

Then there are tables that affect driveability in different environments like for coolant temp (warmup) and intake air temps (winter vs summer), so I look at what those tables are making compensations to and log those things as well as coolant temp or intake air temps as well as whatever is getting weird like fueling or boost mainly. These are more annoying to tune for a DIYer because you gotta get the car dialed in the compensations tables in a short period of time and then remember what the conditions (environments) were you got it dialed for, so next time it is different conditions and driving funny and you tune it and get it dialed, you don't mess up what you already did for other conditions. Once I have the compensations dialed for a couple different conditions, I plot it in excel to figure out what the in between conditions values should be. And then occasionally on a cold day I only interpolated the results for, maybe the car overboosts, and I have to go back and adjust the compensation tables, but yea, eventually you get there. This can take weeks though as the temperatures change.

I have found tuning to be a ton of trial and error making SMALL changes at a time to prevent catastrophe. Getting boost dialed in all conditions is tough, I have found for the 16 bit ECU that I like to control where it likes to overboost with Waste Gate Duty Max (2000-3000 rpm range) and then it can be affected by IATs with the IAT vs WGDC table, and this lets me not overboost when it is cold but still make full boost all summer long as I don't limit WGDC at all past 3000 RPMs, Turbo Dynamics controls the boost at that point. Just took a lot of trial and error stabbing the throttle at different starting RPMs in different temp days to figure out all conditions in which boost wants to run away on me and then get it tuned to control this without limiting boost the rest of the time. This is just my strategy and you may find you prefer another strategy, but it will take repeated failure to get you there. Just pay attention to knock, AFRs, and run reasonable boost based off efficiency maps for your turbo and you can prevent your motor from blowing up while you learn. I kept my car very rich for a while until boost was in check, and then slowly dialed out fuel, and dialed in some more timing, and then adjusted the boost to remain how it was behaving prior to leaning out and adding timing. IMO this is a good way to keep things safe while you experiment, and boost is the greatest variable and needs to be dealt with first.

Anyways if you make your own thread I will chime in if you have Q's I can answer, I am still a noob in the tuning world but I have spent hundreds of hours refining my tune at this point so I have enough experience to help with the basics.

I have been logging a bunch as I have been driving with the new EJ206 and one step colder plugs. I have a stock tune on there with changes for my fueling and MAF due to injectors and intake, closed - open loop transition counters are all set to 0, and WGDC max is all zeroed to run WG pressure. Also I added 5* to the timing advance table and pulled 5* out of the base timing table to give IAMs a little more impact. Interestingly I have hit 11.75psi with the WGDC table zeroed. I have been getting very little FLCK and FBKC during cruise which is mainly what I am focused on right now. Learning view at first had a few cells it was pulling about a degree of timing, and as I have kept driving the cells went away and it is all 0s now. My fueling in learning view has been slowly increasing how much it is pulling during closed loop, the values are now at -7ish across the board which is weird. My wideband has shown the car hanging out at 14.7 the vast majority of the time. It started around +-0.8 in most of the fueling ranges.

I have noticed during my logs that even though my IAM is at 16, there are times that the timing advance is at 0. It is very low load <0.1g/rev, possibly this is decel (I don't think so though as RPMs were pretty constant)? Was not logging throttle position, I will on my way home today. But the IAM only advance is commanding +5 degrees, and the actual advance is at 0, why would this be? FLKC and FBKC are both 0.

I also had another event where the car very suddenly pulled some FLKC starting at -0.70 (so it had pulled more at some point and then added some back) and this number quickly increased incrementally up to -2.81 and then my IAM dropped to 8 and then 4 in a fraction of a second. This was at constant RPM (about 3000) and constant load (0.76-0.82 g/rev) and relative manifold pressure was between -2.6 and -3.2psi. It corresponded with a small EVAP leak code that popped on, I think there may be an issue with the purge line going to the tank or at the tank itself. IAM quickly (maybe within 20 seconds) rose back to 16 and all was fine after that. I will keep an eye on this and probably pull a half degree of timing in the -0.7g/rev range at about 3000 rpms, as learning view has reported this range a few times now.

I made some det cans too to actually hear what is going on and see if it is real or not. I want to change my oil again before doing any real pulls so I will keep logging cruise for now, but I am itching to add some boost back in soon. Car is hitting about 2.35g/rev at like 10psi currently which seems wrong, maybe it is related to my weird fueling numbers and something is up with my MAF, though it is a relatively new OEM part.

But if any of you have answers for me on why the car would be adding 0 degrees of timing when it should be adding 5, when IAMs are at 16 and there is 0 FBKC and 0 FLKC, that would be great. Is there an aspect of knock control I am not aware of?

In response to my question above, it was indeed adding 0 timing from the advance map during decel, when the MAP values would be at full negative. I have readjusted the lower end of the timing map back to stock advance levels (mostly 0s).

Alright I compiled all the knocking cells from hours of driving and graphed it in a scatter plot here. It is all mainly cruise in varying conditions and speeds, from backroads to highway driving. All of this fine learning knock would always return to 0 eventually, none of it was ever persistent, but it would always return at some point.



Then here is graphed the instance of enough fine learning knock to drop IAMs:



Interestingly it occurs in an area of the top graph where there was no recorded knock prior. So I will keep an eye on it as well as dig into my small evap leak issue as I think they are related.

And here is the recorded feedback knock so far:



It is mainly at about 3600rpm, so it could be a harmonics thing and the sensor is too sensitive there. I will keep an eye on it. I will probably adjust the timing at the higher RPM higher load instance, as I imagine that with the higher compression, higher RPMs and boost is where I will see the most need to drop timing.

I will probably adjust the timing in response to the top graph, seems like that whole area could drop a half a degree or so and then I will repeat this exercise all over.

New logs new scatter plots



Pulled timing in the areas asking for it. The car is running absolutely terribly. IAM was dropping, car was adding over 4 degrees of timing with IAM at 9 and then it drops to 8. I don't understand it, it drives fine at low loads and terrible at higher loads, feels exactly like it has nowhere near enough timing. The engine is 9:1 and I am running one step colder plugs. It breaks up and when it is not breaking up it feels and sounds like it has nowhere near enough timing.

Was wondering if I messed up copying over my MAF scale and I compared it to the OTS Cobb MAF scale for SF, and it is pretty damn close, definitely nowhere near off enough that my load would be off so far to the right that timing would be nowhere near what it needs. But it does seem odd to me that I am hitting 2.4g/rev at only 11psi, especially when I was maxing out at around 2.5g/rev at 18psi.

So I bought some stock heat range NGK plugs, I am going to do a compression test at the same time and see where this motor is at, my old motor tested at 150-160psi at all cylinders, maybe this new one is really tired and will be the same or lower and it actually needs more timing. I know the valve clearances are all in spec from measuring them.

The other unknown I want to deal with are the pink injectors in it, I think I am going to test their flow rates or just get some dark blue injectors from Deatschwerks so I can rule them out. My AFRs are dropping super low, learning view is saying I am pulling a bunch of fuel (5%+)across the board when with the old motor everything was pretty spot on with the same MAF scale and Injector scaling/latency.

Getting a bit frustrated and open to suggestions. If the car compression tests 160 or below then I am going to load my old "stage 2" tune on it with the boost turned down and see how it likes it, it is my own tune starting from the stock tune, except for the timing tables, I started with Testes timing tables and then adjusted them after logging. But that tune ran really good in all conditions with all the mods in the car currently.

Will see how it likes stock heat range spark plugs, maybe that will fix my issues, but it feels (judging from when the car had a strange issue and pulled 5* of feedback knock 100% of the time) like it is over 10* too little timing up top.

I really need an EGT gauge.

Spark plugs come in tomorrow. I think I am going to get this EGT gauge:

https://thesensorconnection.com/product ... hold?v=766

And this EGT Sensor to pair with it:

https://thesensorconnection.com/product ... eter?v=621

Exposed K type thermocouple for automotive turbo gasoline racing applications, +- 1*C, peak hold and warning light, can even wire in a buzzer or an additional warning light, 250ms response time, good price, I will give it a shot. Also, it will fit next to the cupholder in my Bugeye where that blank panel is now, it is 1" tall and less than 2" wide. Don't really want to clutter things up with additional 52mm gauges.

I compression checked the motor cold and it was about 120-125psi across the board, so not a motor in need of less timing IMO. I put the stock NGKs in it and well it still seems like it needs more timing up top, but it feels a lot better than with the colder plugs where it would break up and really fell on its face hard. I will log on my way to work tomorrow and see what I can see. I think I need to replace the MAF wiring as it was randomly throwing a MAF code until I jiggled the wires around. The wiring takes a hard 180 turn, so not surprising it is failing after 330,000 miles.

The new stock heat range plugs have stopped the breaking up at higher loads higher RPMs, but it still made no power up top and made a lot of noise like it had a timing issue.

In frustration I compared my old tune to the stock tune I was starting from, and the timing is almost 10* less in the old tune (Testes timing map based).

I dropped the boost to stock levels and loaded it up, and it is running great. I will post some data from my logs later once I go through it all.

I logged a bunch and took some learning view snaps at various points. Here they are in chronological order:







I think I need to get my MAF scaled better, I will spend some time running in forced open loop and I am going to simplify my Fueling Map to make it easier.

The car is running really well, it is flowing really well compared to the previous setup, flowing about 10% higher numbers at 10% less boost, I have to think it is those enormous intake ports. It was flowing just about 2.5g/rev with the previous setup at 18-19psi, and with an identical setup except for the EJ206 part, and I have it set to 14 psi but I have seen it peak at 16 in the logs, I did not have to change anything really, the unchanged Turbo Dynamics settings managed to control the boost without any fluctuations.

I logged all the recorded knock and scatter plotted them again. I need to learn to write a script to pull all the cells with knock into their own spreadsheet. Takes forever. Over an hour of logging. Mix of hard driving, some full throttle pulls, and cruising highway and backroads. I break them up into 3 minute sessions so I don't have one massive file.

Not too much feedback knock:



Here is all of the fine learning knock that occurred:



And here were the instances of fine learning knock I was most interested in, as these were all cells where it would continue to pull fine learning knock even after it pulled a degree already. So, -1.41 and above. There were 4 instances of it happening, I plotted them all in different colors. One event was very small, those are the gray dots:



I think I will get the fueling dialed in before dealing with this though, and I ordered the EGT sensor.

EGT Gauge and probe came in but they forgot the compression fitting, but when I emailed them they sent it right away. Will get it installed soon and post some pics.

if you have knock, better to pull a few degrees to get you knock free so you can work on the AFR target/maf scaling. once the fueling is spot on then bring the timing back in. Cab't really offer much more without seeing the rom/logs. You can turn off all the TGV codes as well so you're not looking at them assuming you have TGV's deleted.

Is it stock intake size?
Tell me what size you have i have a calculator that will get you very close to MAF rescale.
If its stock size and you re scaled it or re-scaled it to compensate for injector latency or pulse you've also adjusted your timing with out knowing.. but it wont show on the ignition tables... Its all to do with G/rev and G/sec..

Put stock turbo inlet back on.. reload Stock Maf scale, stock timing..

Very good and interesting post!!!
Thanks