Unrealistic VE numbers in modelled fuel table

[TENSIX]

I have a GC8 with a EJ207 Ver7 STi engine fitted.

car has been sitting for awhille, though did previously run good with a different ecu, but the wiring harness was failing.

Since rewiring and installing a FuryX to the car, every thing tests fine and works normal.

but the fuel table numbers are just wack. And theres a part were at 3000rpm the car seems to want all the fuel available then dips.

also the cam timing is going to ~5 degrees advanced when it is in active, which didn't happen before but other than that the cam timing tracks well. I haven't pulled the covers off to check the engine timing. but I haven't had to change the offset from zero and as recomended the 100ms dwell works a treat and the timing locks on at the calibration and doesn't move up the rev range. 

Im out of ideas with out putting the parts shot gun on it but i don't like doing that, the injectors were new out of the box but probally the best part of 5-7 years old so they would probally be the first thing I'd replace if i can't identify the issue. also to be noted that it is still top mount intercooler and a long probe hella IAT sensor is fitted in the factory location under the throttle body

right before i drove the car for the provided log file and tune i zero'd out the IAT correction map so its more rich than normal. please note this is not my atempt of tuning the car im just trying to get the fuel map to work correctly before i start the calibration process.

after writing this im leaning towards to replacing the injectors, but willing to see if i've stuffed something else up in the process.

Wrx10082023.pclx ECU Log 2023-08-10 8;31;09 pm.llgx

[DerekAE86]

You probably want to do a PC log rather than the ECU one. There are heaps of runtimes missing and it's hard to get a complete picture of whats going on.

Go to Logging -> Setup Logging -> PC Logging and then click "Add All >>"

Then you can press F8 to start the logging process. Go for a quick drive around the block and then F8 again to stop logging.

But such high numbers in the VE indicate something is wrong in the setup. Do you actually have 1000cc injectors?
The log doesn't show your Fuel Pressure readings - do you know how much pressure is actually in the system?

Your accel enrich settings seem high to me, but I don't really know enough about Subaru engines to really say that's an issue.

[dx4picco]

New age engines with stock fuel rails will have a bump in the fuel map due to funny resonances going on between 2500 and 3000 rpm. 

At low intake pressure. 

That may be what you experience

[TENSIX]

19 hours ago, DerekAE86 said:

You probably want to do a PC log rather than the ECU one. There are heaps of runtimes missing and it's hard to get a complete picture of whats going on.

Go to Logging -> Setup Logging -> PC Logging and then click "Add All >>"

Then you can press F8 to start the logging process. Go for a quick drive around the block and then F8 again to stop logging.

-Can do a PC log soon.

But such high numbers in the VE indicate something is wrong in the setup. Do you actually have 1000cc injectors?
-Yes they are injector dynamics ID1000 injectors
The log doesn't show your Fuel Pressure readings - do you know how much pressure is actually in the system?
-the log does show differential fuel pressure which is stable

Your accel enrich settings seem high to me, but I don't really know enough about Subaru engines to really say that's an issue.
-it would be just the defualt setting in the file. but i doubt it would have any relation to the issue

 

17 hours ago, dx4picco said:

New age engines with stock fuel rails will have a bump in the fuel map due to funny resonances going on between 2500 and 3000 rpm. 

At low intake pressure. 

That may be what you experience

-this engine has always had the same aftermarket fuel rails fitted.

 

 

I got this car 2015, apart from using it as a test bed for multiple ecus it has mainly stayed in the same config. 

The last 4 years the car had no use and been sitting. I've since given it a big birthday, cleaned the fuel tank, new fuel pump, new ecu as I sold the last one that was in it. when I got the car it came with a brand new set of ID1000 which I never used and for up untill now it was using the OEM pink 550cc top feeds.

so with the previous ecu with no issuesthe car was using:
-the on board 3bar map sensor 
-OEM 550cc injectors
-1/8 npt IAT sensor between the throttle and the top mount intercooler
-aftermarket pod filter in the engine bay.

Now with the G4X FuryX it now is using:
- Factory Ver.7 MAP sensor
- Injector Dynamics ID1000 Injectors
- Hella 6PT 009 109-041 IAT sensor located in the factory position in the manifold
- Original air box has been installed

the rest of the car is all the same from the tomei arms turbo and turbosmart fuel pressure regualtor.

[TENSIX]

21 hours ago, DerekAE86 said:

You probably want to do a PC log rather than the ECU one. There are heaps of runtimes missing and it's hard to get a complete picture of whats going on.

https://drive.google.com/file/d/1h7AkwXhXoczNkqHQNZZNyONqxOwpaTio/view?usp=drive_link

the link above is for the PC log

[DerekAE86]

I would probably turn off over-run fuel cut while you're chasing issues because it might confuse the situation when it keeps enabling and making the lambda go lean.

But I don't see anything crazy out of place apart from maybe the "Fuel Charge Cooling Coeff" being set to 0.
The helpfile says a typical setting is 10. But I don't think that will cause such large numbers in the VE table.

I would consider getting those injectors flow tested. 

[essb00]

Your dead time data and SPWA seem too far different from the ones provided by ID.
https://help.injectordynamics.com/support/solutions/articles/4000074340-link-engine-management

Try to get the Excel sheet there for ID1000.

[TENSIX]

16 hours ago, essb00 said:

Your dead time data and SPWA seem too far different from the ones provided by ID.
https://help.injectordynamics.com/support/solutions/articles/4000074340-link-engine-management

Try to get the Excel sheet there for ID1000.

Thanks. though those main deadtime numbers aren't too far off the universal ones i used off ID's website. I did try it out with little to no effect.

I'm still basically lost. seeing as theres no one here anymore (without being a tool) where I am that will that will flow test/clean injectors. im going to use a good known set of injectors before i get a new set.

have included the data log of the deadtime changes.

PC Datalog - 2023-08-13 11;12;53 am.llgx

[Adamw]

I dont see anything obviously out of place in your tune or log.  Assuming the fuel density and stoichiometric ratio setting match your fuel, then the most likely reason for a higher than expected VE would be less fuel flow than the ecu believes there to be.  

It sure seems excessive that you need 65% inj DC with a 1000cc injector at only 12psi boost on pump gas.  

[remski2]

Have you checked for exhaust leaks ?

There might a leak somewhere close to your lambda sensor.. sucking air in and skewing the reading.

[TENSIX]

10 hours ago, Adamw said:

I dont see anything obviously out of place in your tune or log.  Assuming the fuel density and stoichiometric ratio setting match your fuel, then the most likely reason for a higher than expected VE would be less fuel flow than the ecu believes there to be.  

It sure seems excessive that you need 65% inj DC with a 1000cc injector at only 12psi boost on pump gas.  

The fuel used is Ampol 98 unleaded.

I played around with it abit yesterday, the small fuel pressure pulses seam to coresponed with the werid high ve around 3000rpm. so I removed the reference line off the FPR as it was mainly getting its source from number 1 runner. Reset the base pressure to 60psi, it held 60psi unreferenced pretty well still slight fluctuations but nowhere like it was before. it took about 8% off the VE needed but the 3000 rpm area is stil prevalent.

I think its a combination of things going on. I'm assuming the injectors are the root cause, the FPR is audibly a little more chattery than usual and the last bit of my assumptions are the cosworth TGV deletes. 

reasoning for the TGV deletes are that it still has stock Ver 7 sti manfold and heads, and the cosworth items are manufactured to match thier ported heads and lager manifold. so the edge of the stock head ports are visually seen ~4mm around from the inside of the cosworth TGV deletes. so that might be part of the weirdness. thats how the car was when I got it but i don't think it was much of an issue with the STI 550cc injectors as the spray patten was probably different.

Im just going to hit it was the parts shot gun now. new AVO TGV Deletes, new injectors (ID1050X unless you have a better option im all ears), I'll get a new FPR and maybe some fuel pulse dampeners.

4 hours ago, remski2 said:

Have you checked for exhaust leaks ?

There might a leak somewhere close to your lambda sensor.. sucking air in and skewing the reading.

I'm fairly confident that is not the issue, as there is a direct correlation between the engine running like crap when its lean. 

[Adamw]

On 8/14/2023 at 11:34 AM, TENSIX said:

new injectors (ID1050X unless you have a better option im all ears)

I have good data for the Bosch 0280158040 which are a good injector, about 890cc @ 3bar.  I haven't yet tested them at other pressures but will do in the near future.  They are have a long pintle on them but many suppliers sell them with adapters.  

[TENSIX]

On 8/15/2023 at 8:58 PM, Adamw said:

I have good data for the Bosch 0280158040 which are a good injector, about 890cc @ 3bar.  I haven't yet tested them at other pressures but will do in the near future.  They are have a long pintle on them but many suppliers sell them with adapters.  

I've used them before (would be nice you see your data as I struggled to find good data last time) but in this case I ended up using a set of Xspurt XS 1000c injectors.  The other injectors I had were toast, it fixed most of the fuelling issues but not the spike of fuel it wants at 2500-3000 at around 0 MGP. I've fluffed around checking maps sensors are among other things but can't find anything out of wack, only thing I haven't checked is the engine timing but knowing my current luck I can't get the bolts out of the plastic timing covers to check. though the sync timing is normal for the ecu.

To my self it doesn't look normal, but I'm bit at my wits end with it. do i just send it and give it the VE it wants...?

Wrx03092023.pclx

[Adamw]

The spike in VE @ 2500-3000 is common in these, it is resonance in the fuel rails and sometimes the reference to the FPR as well.  

Are your rails plumbed parallel or series currently?  Does the FPR refernece connect to a reasonably central point?

[TENSIX]

9 hours ago, Adamw said:

The spike in VE @ 2500-3000 is common in these, it is resonance in the fuel rails and sometimes the reference to the FPR as well.  

Are your rails plumbed parallel or series currently?  Does the FPR refernece connect to a reasonably central point?

The rails are in parallel with -6 AN hoses. Equal lengths from the Y blocks.

I've played with the FPR reference a bit. Moved it from the centralized location from #1 runner to top section of the plenum where the idle solenoid ports into (see picture). As I found that to be the reference that gives the least fluctuations

Physically the main differences with the fuel system from previous is that it went from 340lph fuel pump and 550cc injectors to 255lph and 1000cc. 

I did find the pumps limit when installed the new set of 1000cc injectors. close to one bar of boost and 70% duty on injectors at 5500rpm it'll loose fuel pressure quite exponentially. Obviously it was extremely rich and the condition wasn't repeated again. 

But if your suggesting it's a resonance issue. I'm pretty much limited to reinstalling a 340lph pump to increase the flow in the system and maybe fuel pulse dampers?

[TENSIX]

found time yesterday to install the radium fuel rails and fuel pulse dampers. have only taken it for 20 min return trip, but it now with out changing the map, idles at 0.85 from 0.99, and durring driving in the spike area, the CLL is now Pulling 10% and its still around 8% too rich. so looks promising, the only other thing noticed is the fuel pressure now pulses 10-15kPa which dies down to 5kPa when map is 0 and thereafter. dunno if this is an issue or if there is a way to filter it out.