AVCS drama/Standalone tuning disappointment

[rorygtb]

Hello all, thank you for taking the time to read this post. I was wondering if anyone has experienced anything similar or can shed some light on this issue...

Background context: 

The car is a 2006 Subaru Legacy GT which started out life as a 5EAT auto, subsequently swapped with a 6MT manual from a UK spec 3.0 legacy...

Engine (in short) is a forged EJ207 with a Spec C VF36 Hybrid turbo with forge actuator (0.9bar spring), 850cc injectors, parallel rails, FPR, 340lph pump etc... In theory, the engine which was freshly built by a very reputable engineer should be good for about 350-380bhp on the current setup. 

When the car had the 6MT swap completed, it was mapped on the road, and (despite no dyno graph), using MAF g/s as a rough guide on how much power it was making, it was doing (comfortably) 350+ bhp at 1.6bar boost at 70% WGDC.
However... it wasn't all good news... there was a persistent misfire at 4500rpm which was investigated very thoroughly by the garage who did the 6MT swap, to no avail.  

Fast forward 6 months, the car ended up at another garage who specialise in supplying/fitting/tuning standalone ECUs. The decision was made to just jump in at the deep end and do this in order to completely control all engine function with a separate ECU and remove any of the remaining 5EAT auto trickery hidden in the wiring (which we felt may be the root cause of the misfire issue). 

Ecumasters EMU Pro 8 fitted and mapped - absolutely no misfire. Car runs well, smooth, now MAF-less but very much drivable like OEM day-to-day. 

EXCEPT... it's not all good news. The car only made 300bhp and 295ft lbs at 1.35 bar!

When I spoke with the tuners, they said that the turbo was tailing off to 1.0 bar at redline and was evidently at the limit of its potential, therefore it stayed at 1.35 bar and the map was considered done...

When the car was mapped on the standalone ECU (by the garage who fitted it), they had no base-map for a 2006 JDM Legacy to build on, so it had to be written from scratch. This included the AVCS settings for the legacy quad (dual) AVCS. They told me that due to the engine being high compression (it is pretty high, higher than a standard twinscroll which is already on the high-side), they had very limited ignition timing to play with when trying to make extra power. They also appear to run AVCS straight through the rev range whereas the first tuner explained that it needs to turn off at some point the higher up the rev range you go.

Having taken this info back to the tuner who road-mapped the car initially, he was confident that there was something in the new map limiting the car being exploited to it's full potential. Both the first and second garage have spent considerable time with the car and any genuine mechanical faults would have been picked up by now, particularly when fitting and mapping the standalone ECU. He explained that that due to the fact the AVCS works slightly differently when tuning Subaru engines vs say a conventional ICE car, this could well be the problem. 

Now I'm at a bit of a crossroads. Does anyone have any similar experiences or knowledge which they'd be kind enough to impart? Both garages are very experienced in tuning various cars, the first tuner moreso in the Subaru world. What I don't want to do, is have to have about 4 maps done before it finally makes the power it should have in the first place! Do we think it's worth having the first garage look at it, again?! I've spent so much money I didn't intend on spending on this car, the fun levels are dropping at a rapid rate of knots haha... 
I'm also conscious that (despite me meaning absolutely nothing negative by it at all), I don't want to give the impression to either garage that I don't trust their work and am taking my business elsewhere (they're both very experienced and I've no real issue with either of them). In essence, If the second garage fitted and mapped the ECU at great effort (took an age to optimise all the base-map, weekend overtime etc), I don't want to cut my nose to spite my face and look like I'm untrusting of their work, taking it to the first tuner (who also happens to work closely with the second garage on occasion!)...

It's a bit of a tricky one. I've spent so long fighting with the car that I'm so relieved it works and drives eventually, but at the same time It feels even more frustrating that I've been disappointed by the final outcome, as it is way less than it should have made in theory!

Sorry for the long-winded post, really just looking for some second opinions and a bit of advice really. Bonus points if you're clued up on the technical tuning side of things and have any ideas why the car made so little power! 

Thank you,

​​​​​​​Legacy1997

[bigBADbenny]

Was the same dyno used between tunes?

What’s not to like about your figures?

221wkw on 19.5psi and 850cc injectors, a stock location turbo: seems pretty optimistic to me.

Personally I’d detune that to say 17psi max for some headroom and less peakiness.

An aftermarket ecu is not essential for a manual swap car.

AVCS is used primarily for cruise economy and can be used to help with spool, but most stock and tuned avcs maps drop away to zero in the upper load/rpm range.

The vf36 was designed for quick spool: twin scroll, titanium shaft, 350hp seems to be the top end on those over-optimistic American dynos.

Either way at the limit, it’s just making hot inefficient air, best keep it in it’s sweet spot, around 15psi?

The engine is just a pump, so it’s your bolt ons that will make the power your engine should support 300wkw with the right bolt ons…

Locally a similar build would use ID1000 injectors for 200 over or ID1300 if pushing to 300wkw…

[rorygtb]

Benny - thank you for getting back. Every time I come to these forums for advice you're here! Thanks for the continued support/advice.

To clarify, in the UK we use engine bhp vs you lads who tend to use wheel horsepower/wheel Kw.

This means that the car actually made 219kw at the ENGINE and (for arguments sake if we just say approx 25% loss via drivetrain etc) about 165wkw.

If you assumed the car made 221kw at the WHEELS, this would be about right as an optimistic power output as it's the equivalent of 370bhp at the engine using our same 25% drivetrain loss rule. This is at the upper limit of what I assumed the car would make and I'd have been over the moon if it made that much!

Re: the manual swap not requiring a standalone. I am aware of this, but have had a total nightmare with the initial swapping in of the MT and a persistent misfire which made life hell for the best part of a year (two motorsport garages were completely stumped and had no access to any JDM pinout diagrams to check wiring etc). Was a standalone excessive? Proabably, but the car drives and runs now. It wasn't worth much to me or anyone else as a half finished paperweight with a phantom misfire!

Interesting about the AVCS being used for cruise economy. I didn't know this!

Worth noting about my VF36, too, that it's a hybrid which has been both ported and has a billet inlet wheel. If we think about the cars these turbos were fitted to from factory (say a 2003 Spec C STi), they will make 350bhp/260kw (at the engine) all day long with basic mods (downpipe/dump pipe. fuel pump and uprated air filter). Therefore, theres no chance the setup on my car is at it's limit at 294bhp with what is effectively a Spec C setup (but better)... At least this was my logic, I may about to be humbled !
The only other thing I can think of is that the standard twinscroll heads are on the EJ207, and someone suggested these wouldn't breathe anywhere near as well as JDM Spec C heads, however I can't see this being the limiting factor when making power at a measly 294bhp!!!

Thank you again for the reply!

[bigBADbenny]

Ok plot thickens!

Hopefully the aftermarket ecu has a bunch of extra functionality over a stock ecu: eg integral oscilloscope, external sensor & failsafe channels etc.

It’s interesting that the vf36 is on stock ej20x/y heads and headers, did they mod your up pipe and dump to suit jdm STi twinscroll turbine housing? Sorry just ej20x/y heads…

Were the before after dynos on the same dyno?

Do you have datalogs from the whole process? I’d love to see any and all available, share a link to your Dropbox, google drive or datazap.

I’m absolutely not qualified to even comment here, I just aggregated some learned knowledge from my time reading posts here over the years: rule of thumb stuff.

The 350hp I saw from searching were US cars on vf36 and would be wheel HP figures generally around 20psi, generally happy US dynos.

I’ve always had my car tuned on the safe side of the power equation: less is more for me as regards a smoother less peaky power delivery (essential for a vf46 lol).

I’ll invoke an old member Dr20t who can give you tons of expert advice, assuming he’s interested and available.

Three questions above, love to see the answers.