RomRaider

I'm in the process of supercharging my Z and thought it would make for a nice example of tuning the car with romraider if and when I get to it.

I'm currently waiting for a warranty part from HKS and have been putting the car together enough that it can be driven as NA in the meantime. I've been looking into the ECU side trying to learn what is necessary for accomplishing this without blowing the engine to smithereens and there are still a host of things I'm not completely clear upon and if anyone is willing to help me out it would be much appreciated!

I intend to update this thread with any future issues/challenges/solutions so hopefully at the end it would serve as a useful reference to others.


Setup:
- DW300 340lph fuel pump
- DW600 injectors
-- Flow rates: 654, 654, 660, 660, 666, 672 @ 3 bar (Measured)
-- Flow rates: 708, 708, 714, 714, 721, 727 @ 3.5 bar
- HKS supercharger kit with GTS8555 compressor
-- 120 mm pulley
-- No intake restrictor, limit throttle valve maximum opening via ECU instead
- Stock engine internals
- No fuel return
- External flexfuel system, so ECU should still think it is running gasoline, except timing trims if I only use E85
- External wideband o2 sensors, Innovate DLG-1
- CD415 ECU
- Modest power target (<400hp at the crank)


Tuning notes:
Injectors
- New size via scaling K-value
-- OEM flow rate: 320 @ 3.5 bar
-- ( old injector size / new injector size ) * old K-value = new K-value
-- ( 320 / 708 ) * 27 592 = 12 471
-- ( 320 / 708 ) = 0.4519774
- Scale Cylinder Filling Up Efficiency Scaler (mPT100)
-- Why does this need to be scaled, what does it effect?
-- Scale by multiplying with ( old / new injector size )
- Modify minimum injector pulse width
- Modify injector latency at 14 V
- ITAC goes from 0-100 in NA roms and 0-200 in FI roms. Does it need to be scaled to match FI roms? If so, how to do it properly and what does it effect?

Fueling
- Adjust target AFRs !!! This only sets what the ecu's goal is, does not guarantee it is hit !!!
- Fuel compensation maps (steady state, decel, accel) for main changes
-- transient (decel, accel) are used to compensate for the distance between MAF and combustion chamber. How sensitive is this, and does it need to be adjusted due to the HKS kit? The MAF location is not completely identical to factory intake, but is the change large enough to warrant changes?

Timing
- Ignition timing trim maps are probably enough?

Air flow
- Use MAF Table to calibrate air flow with steady state driving in open loop and partial load <= 60% engine load
-- Is it enough to set the target AFR to something besides 14.7 to force open loop, say 14.0?
- MAF factor for scaling signal in case voltage bottoms out

Throttle
- Disable DBW torque flag so pedal <-> throttle movement is 1:1
-- Personal preference, only reason to not do this is if I get cruise control to work at some point as this disables it
- What should be changed if one wants to limit maximum throttle valve opening?
-- Target Throttle Valve Opening Table?

Transmission
- Increase engine torque maps based on dyno results to increase line pressure and prevent clutch slipping and burning out clutch packs

Datalogging:
- Use romraider with openport 2.0 adapter for logging
-- Expose vITAC, vQH0, vMBTCAL, vTQH0AP & vKMRM RAM parameters
-- romraider should support innovate Wideband AFR logging
-- First test connection to ECU was a failure. Not sure why.
- Megalog HD viewer for post processing/data analytics


I've added the following definitions to CD415.xml and modified them in order to take the larger injectors into account. The car started right up but it's idling really rich, around 10.2 AFR so either I've missed something or my memory locations are not all correct. If anyone more knowledgeable than me would be kind enough to confirm whether all these are correct, it would be highly appreciated!



Changes to aforementioned tables
- K-value -> set to 12 471
- Cranking Base Fuel Schedule Table -> multiplied all values with k-value scaler (0.4519774)
- Injector Magnification Reciprocal Number -> unchanged
- Fuel Injector Latency Adder -> set to 0.15 ms/V
- Fuel Injector Latency at 14V -> Set to 1.3 ms
- Minimum Injector Pulse Width -> multiplied with k-value scaler => 0.44 ms
- Cylinder Filling Up Efficiency Scaler -> multiplied with k-value scaler => 8.76 ms

I wonder if your actual fuel pressure is higher then 51psi with no return line and a 340lph pump, could be contributing to the rich idle.

Does it get better as it warms up? There is also supposedly a post start enrichment table that I've seen in the turbo maxima thread that may need modified. I've only seen it defined in that ecu. Not sure if that would cross reference or help.

viewtopic.php?f=64&t=19248

Could also lower injector latency and see if that helps. To my understanding decreasing latency will remove fuel more at idle/cruise and nearly none at full throttle.

I am in a similar situation (hopefully first test start this week) and may just only be able to run e85 in mild weather without access to some of those tables.

Fuel pressure is close to correct value. Though you are right, I had to modify the in-tank pressure relief line on the fuel pump housing to bring the pressure to acceptable levels back when I installed the pump. It was idling around 70-80 psi before drilling the line bigger.

I did a short test drive and after warming up, the car does idle at 14.7 but AF trims dropped from 100% to around 80-85% so something's not quite right on the injector scaling. WOT afr is sitting at 11.2 which is also the ECU target so that's good.

Got first logs out of the car with romraider and openport adapter. Was missing the tactrix drivers on my first attempt.
ITAC, QH0 Flow Quantity & Potential, MBT Operation Value and Target AFR RAM parameters are giving sensible values, so that's good. I'm still missing a serial com to usb adapter so wasn't able to test wideband logging yet but it should arrive soon.

The default ignition timing, STFT and LTFT parameters are not logging values that make sense to me though. The ignition timing I can overlook as I have the MBT operation value working, but the fuel trims would be nice. I wonder if they need to have some sort of conversion done? Megalogviewer is showing them to be 0.29...0.43 [%] which.. doesn't make sense as I mentioned. Looking at the trims via consult 2 android apps shows values in the range of 80..100 [%] which is what I was expecting.

If anyone's curious to take a look, there are two logs in the attached zip file.

The fuel trims on romaider also don't make sense to me either. I've been using nissan datascan II to datalog and fuel trims look more normal and can see actual ignition timing. It seems I can't log cylinder filling efficiency anywhere so I've had to be a lot more conservative in the tune for now.

It turns out the g35x ecu was identical on the main engine pins everywhere except for the wide band oxygen sensors I had to pin in so took a little longer to get this car started than I expected.

I did a low boost pull and everything seems to be running good for the most part. Car starts up and idles at stoich. The main problems I had were not pulling enough ignition timing at first was running 27-30 degrees in boost despite running the high det ignition map for the high octane map. Also fueling was not really responding quick enough in boost so had to make most of the target afr map super rich and adjusted acceleration compensation. Still not sure which fixed it, I would like to make target afr leaner in certain areas but without logging cylinder filling efficiency it makes it kind of trial and error.

I still don't fully understand ignition timing map fully but seems like -14 in most of the map will run 15-20 degrees. Also when logging have to kind of convert oxygen sensor voltage to afr after logging but shows scaling in the map so not that big of a deal but looks weird at first.

The Main problem I still have is hot starts, will start perfectly cold everytime but have to hold pedal down to get it to start when hot sometimes. Not sure if I need to reduce cranking fueling on hot startup or give it more. I would think holding down the pedal would make it leaner? But not sure if these ecu's stop injector pulse width when holding down the throttle or not.

I attached rom and log- currently only running 3psi boost hope to run 6-8psi soon
Car is 2003 350z running a 2006g35x ecu
Gt35 turbo
580cc injectors (at 3 bar with return line)
stock maf converted to draw through
stock block vq35de

Hey, at least you got your car boosted and driving way earlier than me. The broken pulley bearing assembly is *still* stuck at US customs and it needs to end up in Japan before they even decide to send me a replacement.

I was curious to see how your fuel trims were but the log you attached is missing them, though since it's a pull it wouldn't have mattered anyways I guess. It'd be interesting to compare idle and steady state fuel trim behavior as I feel mine behaves weird after looking at fuel trim logs I did with NDS2.


Yeah, I'd try reducing hot startup fueling and see if it helps. Cranking the car is the one thing I've not had any issues with so far at least, turn the key and it fires right up. Wonder if me using E85 helps that any.

Curious how "- Disable DBW torque flag so pedal <-> throttle movement is 1:1" that little portion is done there. QH0 adjustments are as Pyrtex puts it "finicky" lol

Disable mfTTE_ON bitwise switch which is in the Electronic Throttle Control flags.

At least, that's what I did and when logging pedal/throttle valve positions, they track each other really closely.

Well.. Very, very late update due to USPS losing the shipment and ~1 year delivery for a new spare but first test drive was done today. Car was working more or less fine until the first test acceleration, where at 4800 rpm the ECU cut fuel and switched to maybe limp mode or high det maps, I'm not sure. It was running really sluggish until I stopped and restarted the car and was then back to normal. Did not continue driving as I need to figure out what happened and how to fix it.

I've attached the log in case anyone is curious to have a look. The weirdness happens at 455914 ms timestamp, right before the MAF voltage hits 5.055. Does the z33 ecu have some sort of failsafe for MAF maxing out?


I suppose next step would be to rescale the MAF signal akin to this my350z thread instructions.
https://my350z.com/forum/tuning/592155- ... eloading=1

Added the 40 kOhm resistor to MAF signal wire, rescaled MAF table using the provided excel sheet and multiplied K-Value by 1.4 to 17459. Main Fuel compensation map was reset to 100 except for the four cells in the top left corner.

No fuel cut this time around, though I was still obviously taking it pretty carefully. I've attached the logs from the drive again. To get sensible value for fuel trims and final ignition time, I had to modify the logger definition conversion factors for both. Fuel trims had a conversion "x/256" but changing it to "x" ie. no conversion gives me comparable and sensible values compared to several OBD diagnostics apps that use Consult 2 protocol. For ignition timing the conversion had to be changed from "60-x" to "110-x", which I found out later was already present as a fail safe ignition timing in the logging definition for nissan ecus. I just didn't have it listed in the profile I have.

So now comes the more difficult part of trying to figure out what to change and how.

Car is now successfully dynoed and MOTed. Emissions tests pass just like a stock car so that's nice. I've attached a download link for all the different versions of my ECU changes, and logs for the different versions, along with a text file detailing the changes done to specific versions.

Instead of adjusting the K-value at first after the resistor mod, I used the MAF scaling factor first. While correcting the MAF Table, I ran into uint16 max value limit on higher air velocities and ended up having to adjust the K-value larger because of it. That ended up being a good thing as now the K-value is closer to stock than it had been and the annoying LTFT reset happening at closed loop is almost completely gone. Ignition timing has been left to pretty safe levels as I'd need proper dyno time to finetune it. The car dynoed 340hp / 400 Nm with rom version #16, which was lower than I expected from how quick the car feels. There was also an odd drop in power happening around 6k rpm but the log looks fine to my eye, and I have not noticed something like that during on road tuning and driving. I plan on bringing the car to another dyno measurement at some point to see if it was just a fluke/artifact of that specific place. For now I just needed some sort of dyno sheet that is accepted for MOT.


Another oddity is that QH0 Flow Quantity will drop from 100% to around 80% during WOT acceleration as the rpms rise. It hasn't seemed to effect anything adversely yet.


Hopefully looking at these roms, logs and version history will be of some help to others. I will also be really appreciative if anyone notices a mistake in my roms or has some advice based on the attached logs. I've added a bunch of map definitions for CD415 that I'll eventually add to Pytrex' nissan definitions but there are still a few that I want to double check before sending the changes. The added definitions along with the CD002 defs should help with defining the other 7055 ECUs.

https://www.mediafire.com/file/lalh20lb ... ms.7z/file