RomRaider

Hi Guys (I'll try to keep this short),

First of all - Apologies if someone else has already covered this / found what I've found etc, but wanted to post this because this would have saved me a lot of time if I had found it and read it before changing the injectors - basically make sure you at minimum replace the fuel rails at the same time. (or stock injectors and fighting this problem - you need aftermarket fuel rails).

Just thought I'd write up my findings for the stumble issue on the 2008+ WRX STI (in my case 2010 JDM EJ207 Impreza WRX STI), and what I did to completely cure it.
When completely stock - it was not too bad (load compensations tables doing their job in stock form and factory tune)
But of course I started modding, and once I did the injectors ID1050x, I started to get the dreaded stumble 2600-3400 rpm, reading online this is very common with lots of black box fixes with no real answer to what was the cause.

I've never struck this before in tuning cars (well nothing as bad as this for a MAF based system), and at points I was getting pretty frustrated with it - I wanted it gone, but also wanted to understand what was happening ( I'm also pro fuel dampers - they're meant to help things not make then worse!).

I started trying all sorts of things and combinations (back to stock between each of these), changing the damper, removing the damper, blocking the bypass, removing the damper and blocking bypass(aka Cobb fix - the helped them most but not gone completely), blocking the fuel lines between the fuel rails in the middle - making them dual dead ended, added the hose extension, none of these completely eliminated the issue or negated it enough that the perfectionist in me was happy lol - I love to have things well tuned!!

Also note with all the above, they didn't seem completely stable, like the temperature of the motor or a hot day would shift things a little as well ( resonance speed does change with temperature so that lines up.)

All of this testing was done with a fuel pressure sensor (with logging) so I could monitor all changes etc.
What I found here was, anything close to the fuel rail on the (left side when sitting in the car looking forwards - the drivers side in the US), would make things worse - the damper worse, the bypass to the FPR worse (which is why the Cobb solution removes both of these and helps a lot), but I even found that putting my pressure sensor in line here would change things as well - So started thinking it must be the fuel rails!, anything close to that fuel rail would start causing the resonance to become worse or change it.

I found the cheapest aftermarket fuel rails with AN fittings (I wanted high flow ends on them, not capped off ends like the factory ones -resonance bouncing...), left it in series (not parallel), stock FPR and Damper with the bypass still there (so all the new FPR/Damper/Bypass setup from 2008+ in place as stock), and the issue is now completely eliminated - the fuel rails cured it completely and the car runs beautifully - it now revs freely and quickly through the problem rev range like it should! - more power there now too, can feel the boost/torque increasing - its quite different now.

So quick findings, all these logs and compensations were done with the load comp all zero'd out (and all with the ID1050x injectors)


Below are the logging and compensations calculations (+ values = adding % more fuel, - value, removing % fuel)

Here with stock FPR/Damper/Bypass and stock fuel rails, didn't like this at all when driving (peak adding 16.3% more fuel - but wider range)


Here with the hose extension in place (it seemed to compress the resonance range, and was easier to tune, but higher compensation values peaking at 19.5% more fuel) - changed its shape, nicer to drive but not right.


Here with all stock fueling FPR/Damper/Bypass and with the ID1050x injectors in the new Rails - a much more normal looking variance - not adding fuel at all - now normal trim, and like normal car to drive lol - I could hear the difference - its now purring.


So it looks like Subaru somehow created the perfect dimensions for the factory fuel fuels ( the barrels the injectors sit in) to cause a resonance frequency around the 2400 - 3600 rev range, they knew there was something wrong as they added the load compensation tables in the ECU just to dump more fuel in at this range to make it derivable.

To completely eliminate it, you need to replace the fuel rails as it seems most other solutions just move it or change the resonance especially if you car is sensitive to this issue (by sensitive to it I mean - some people claim there's is fine - no issues).

exactly same issue i ran into but reverse i did 6an lines new rails fpr aeromotive 1050x injectors parallel flow and was tuning wondering y im goin from 2% trims to 20% trims in load comp range to lean as crap OL high end then i realized those damn tables were the cause after scaling everything for days. Also ran into a fueling problem up top do to 340lph pump i installed in stock sending unit pressure would climb to 80 psi as days got hotter and pumps duty cycle went up wierd thing is it would lean everything out especially under boost fix on that was the venturi valve in oe sender the return holes are like a 1/32 of an inch to create vacuum to pull fuel from other half of tank do to it being a saddle tank design. I ended up drilling the return hole out on the venturi to an 1/8th aand viola full pressure control back 43.5 psi all the time with 1 to 1 pressure increase under boost. My fuel system was to much for the oe sender and venturi because we solved there horrible engineering. now i use load comp for boost transition to smooth accel and fine knock when giving her the beans. rest of table is zeroes

Can you post photos of your new setup and links to what you purchased specifically? I am on a stage 1+ OTS map from 2014 (came with the car) and have started working on my own map for the first time and this was one of the bucket list items I wanted to fix on the car asap. So far the map changes I've started making are only

MAF Sensor Scaling 1.614 Times stock for the larger diameter
Engine Load Compensation Cruise (MP) Zeroed
Engine Load Compensation Non-Cruise (MP) Zeroed

Haven't touched any timing or other fuel trim tables yet, but I was hoping once I flash the completed map I'd see improvement in the stumble with the above.... lol guess not

Note: I used the cheapest Fuel Rails I could find, as it was meant to be more of a test at that point in time, these are extremely cheap, but have been working fine for me, Epman also do a lot of AN fittings (extremely cheap as well) but don't get those unless you have an AN fix/resurface tool, as I did get some of these as well, but all of them leaked! (at the cones/AN part, so would need resurfacing/fixing even though they are new).
the Rails:
https://www.epmanracing.com.hk/EPSB0214 ... ption=true
I ended up getting Orb (O-ring sealing for the rails) fittings and AN from another supplier (Aeroflow) instead, and just ditched those cheap leaky EPMAN AN6 fittings.
I used "Gates 5/16in x 25 feet Reel 225 PSI Barricade Fuel Injection Hose" for the hose (way more than I needed of course).