The Tuning Thread- EJ Motor

[Lance]

2. As mentioned before, decreasing IAM is a knock control strategy that is meant to make rough/general timing corrections. However, this might not always be needed. Sometimes, your engine may only be experiencing knock events in specific areas of your power band. In this case, the ECU will not make broad corrections to the advance timing table but will try to fix these areas directly.

If your IAM is at 16/1, you will only see negative figures in this matrix (never positive). This is because the ecu cannot add more timing beyond what has been dictated by the Knock correction advance table. Zeroes are fine but negative instances mean that your ECU has detected specific instances of knock and is attempting to pull ignition timing (from your advance table) to eliminate the knock. The ECU will pull timing in -.35 increments. The larger the number means the more instances of knock that has been detected. So if you see -0.70 it means 2 instances of knock, -1.4 (4 knock points etc.). The higher the number the bigger the knock problem!

If your IAM is below 16/1 then the ECU will attempt to add Advance timing back to the motor. In these instances you will see positive figures in the cells. You can also continue to see negative cells even after your IAM has dropped.

Generally, low load/rpm knock correction is not something you should be too concerned about. Your motor will be out of boost and the overall pressure on the motor will typically not be significant enough to cause any damage. I would typically be concerned with any knock above 4000rpm and above 1.25 q/rev of load. High Load / High RPM knock will damage your motor - if its significant enough - not even a forged motor will protect you.

[Lance]

(3) This aspect of learning view is probably the one that is most commonly overlooked by both owners and tuners. It can however be very useful in identifying tuning and mechanical issues with your motor.

The numbers on top are basically air flow ranges- representative of various closed loop driving situations (starting from idle to light cruising conditions). The percentages are learned TRIM values- essentially meaning the amount of fuel the ECU has had to add/subtract in order to hit your target open loop fueling table.

They should not be interpreted as running lean or rich (especially without knowing what your targets are). For instants, people tend to think that having a figure of -10% is a good thing because it means that your car is cutting fuel so its "rich" and "safe". Actually it means that your ecu has had to over time adjust your fueling to hit target afr's - indicating that your MAF scaling is probably INCORRECT. This usually leads to issues with fuel consumption, erratic idle conditions and poor drive-ability. Erratic idling is usually the most obvious manifestation- although tuners might try to mask this problem by raising your car's idle point in order to stabilize idle response. Persons frequently encounter this when they change MAF housing or injector size.

Generally, a correction of +/- 5% is seen as satisfactory. It is also important to note that the Subaru ECU is pegged at a correction percentage of +/- 14.99%. so above/below this it can no longer make learned corrections to facilitate the meeting of target fuel. These conditions pose major issues for safety.

Aside from improper MAF scaling, huge variations in these figures can also be caused by failing O2 sensors, boost leaks etc. Basically any mechanical issue that can cause or report variances in airflow after the MAF has accounted for it.

[Aaron 2NR]

i finishing up my install this weekend and would be tuning later in the week.

really anxious to see the difference the fmic make on my setup.

the only downside is I'm gonna be using a different dyno so i won't be able to pull my old charts

[SUBARU TT]

That's an impressive setup Aaron 2NR

[Lance]

Any more ecu/tuning experiences guys?

Learning view screenshots, dyno graphs, data logs etc.?

[16 cycles]

@ Aaron...you side mounting your #plate?

[castortroy]

New dyno opening up by my side will send u the info before and after meth goes in [GRINNING FACE WITH SMILING EYES]

[Aaron 2NR]

@ Aaron...you side mounting your #plate?

Nah car setup for time attack now so cage etc in, hardly gonna be on the road so I'd just remove it on the track. Can't cross 80 either way lol

[16 cycles]

how was/is the new dyno south side?

[Aaron 2NR]

new dyno is pretty decent. Newer technology and software so a lot can be done compared to the other dynos.

They have a eddy current attachment on the way to prevent traction control systems from going into safe mode etc.


Did some waste gate tuning on the car and thats actually the power I'm gonna leave it at for the street.

[Aaron 2NR]

Before tuning


After tuning and logging for a few days

[Lance]

A4TJ1X00 seems like the group N rom (or an iteration of it).

Are you still calculating fueling off the MAF?

Just a bit more scaling on the idle range it seems - with some table range redefining.

[Aaron 2NR]

Maf disabled since Im FMIC now. Everything settling in the car (piping etc) so in the event that a coupling blow out, we left it off.

Map was left a bit rich since I would be racing with this map this weekend. Didnt start tuning the 2nd map yet...

The plan is to use the street map for circuit and the 2nd map would be used for drag or a race gas mix....

still shaking down the car and would be logging while on the track to fine tune since temps get ridic high which couldn't have been simulated on the dyno

[Lance]

I think I mentioned this in a thread sometime ago. For Learning View to offer useful snapshots, the RPM, load and a/f ranges need to be redefined.

For example, your first 4 load ranges (in the second screen shot) are capturing engine load conditions between 0.00 to 0.39 (g/rev). These probably all encompass idle conditions- situations where you would never experience knock and thus don't need deep monitoring.

Detrimental knock occurs on the upper end (higher load/higher RPM) and this is where LV would provide you with excellent track snapshots.

Redefining these boundaries would be incredibly beneficial.

When I could pull LVs, this is what my ranges looked like. I was on stock power so I was only concerned with engine loads up to around 2.6 g/rev (what I was experiencing). Your logs should be able to tell you what the upper end of your load table should look like.

[Lance]

It has always been my intention for this thread to be an open discussion without egos and ill feelings. Not many persons like to share details about their tune but I was looking through my files and I found the last log I pulled for my current tune - so I wanted to share it.

The tune is a remote tune. For those that might be unfamiliar with the format- the tuner is located in the US and map revisions are sent to me via email. I flash it onto the car, take logs and email them back for further revisions (a repeated process). As you can imagine, tunes in this format are usually more conservative than dyno based tunes - owing to the fact that the tuner is not physically with the car and cannot observe real time response. That being said, this is by far the most refined tune that I have had to date.

The log screenshot is below. But just a few notes. The tune is speed density based so no MAF parameters have been logged. They have been substituted with intake air temperature (IAT) and supplemented by coolant temp readings. The IAT readings are more accurate than stock as I have relocated the sensor to a post-intercooler location - where it is being measured by a AEM IAT sensor that has been wired into the stock ecu and re-calibrated for accuracy. Boost is being captured by an AEM 3.5 bar stainless steel MAP sensor - also wired into the stock ecu and re-calibrated.

I have omitted the AVCS, WGDC and timing values to respect what I think is the tuner's intellectual property. However, you can still see a good amount of data and be able to view the common parameters that you should consider logging. The boost pressure was originally captured in absolute terms but I have recalculated then relative to atmospheric pressure (so everyone can relate to them). The AFR values are captured by an AEM wideband setup and are not readings from the stock front 02.

To me, the most interesting part of the tune is the boost response itself and its "ramped up" strategy. I think this makes alot of sense but we can ofcourse discuss this. From 5100-6700 boost pressure is fairly stable with only an incremental increase (at this point it is being heavily restricted. From 6700-7500rpm the boost is ramped up for a short space of time to keep the power on hard and prevent it from dropping off - while ensuring that the motor is not overly burdened - at least this is the logic. No matter the abuse, the car refuses to knock (see log) so I think its working well for now.

The setup is by no means perfect. This is a fairly sizable journal bearing turbo and the initial lag is very noticeable (between 3000-4200rpm). I think a dyno tune and some adjustments to the timing map might have helped this a bit - but i'm not sure.

Hopefully this can get some open discussion going.

[16 cycles]

^you got any wtq curves off of virtualdyno to relate to the boost plot?

[Lance]

Yeah I do. Table top curve from just after 5100rpm. Looks like this:

[Aaron 2NR]

What solenoid ur using to hold the boost at the higher rpm? Or you using a controller as well

[Lance]

The boost control mechanism is actually quite simple. I've had a grimmspeed 3 port ebcs on the car for many years and I just carried that forward for this setup.

Works really well, at least up until this point.

I just added a Tial 38mm EWG on a 17psi spring to the setup.


What are you guys using?

[rylan1996]

I'm also using a speed density tune with a 3.5bar MAP (SS) and IAT sensor both from AEM so the use of my MAP sensor has also been eliminated. My AVCS has also been eliminated. The car has been road tuned as the base. I would possibly do some logging with AEMTuner along with the same parameters logged as Lance, and as it hasn't been Dyno tuned as yet I'll possibly export the relevant data from AEMData and post the Virtual Dyno once I do get some free time with the car (3rd gear).

[16 cycles]

Any advice , observations in going ewg ?

Supposed to help with better boost control and reduce chance of det by allowing air to pass through faster....

I'm on grimspeed as well...tial recirc...alot better drivability vs -Greddy vta ...esp when off throttle ...iwg still...

[Aaron 2NR]

I'm on the grimmspeed ( thinking to change out to the perrin) and internal gate with a 22psi spring. i think i need to preload it a bit more....

greedy bov....

[rylan1996]

I'm running a 38mm tial EWG with grimspeed 3 port EBCS. Additionally, it does in fact help with better boost control.

[Lance]

I'm also using a speed density tune with a 3.5bar MAP (SS) and IAT sensor both from AEM so the use of my MAP sensor has also been eliminated. My AVCS has also been eliminated. The car has been road tuned as the base. I would possibly do some logging with AEMTuner along with the same parameters logged as Lance, and as it hasn't been Dyno tuned as yet I'll possibly export the relevant data from AEMData and post the Virtual Dyno once I do get some free time with the car (3rd gear).

Excellent man. Car has come a long way. Keep us updated and be sure to post the logs and dyno sheets!

[Lance]

Any advice , observations in going ewg ?

Supposed to help with better boost control and reduce chance of det by allowing air to pass through faster....

I'm on grimspeed as well...tial recirc...alot better drivability vs -Greddy vta ...esp when off throttle ...iwg still...

Going EWG is probably the best boost control upgrade I've made.

It has really helped to dial in the boost response perfectly.

If I had two recommendations for those going EWG it would be regarding waste gate placement and venting.

I chose to mount my wastegate on the up pipe. As oppose to welding it onto the exhaust housing like alot of other local guys.

Up pipe mounting is the harder route for us RHD owners as there are a lot of mechanical parts in the way. But trust me it's worth it.

I think that venting the exhaust gasses pre-turbo rather than inside the turbine housing has really had an impact on reducing turbulence and promoting a much more efficient flow of gasses in the setup.

Secondly, and noise aside, consider VTA rather than plumbing back into the downpipe. I think that this also has efficiency and restriction benefits. There is a post on NASIOC that seems to support this VTA theory (I believe it was you that posted it 16 cycles).

Of course there are certain benefits to be had from plumbing back the exhaust gases e.g. noise reduction, controlling exhaust fumes etc. So it's something to trade-off.

[16 cycles]

I remember that thread on n a s I o c.


Much thanks for the real world local experience...always thought the ewg /vta option was loud/scary...

Drove next to an e8 with that set up and I could not hear my own car...was a wtf moment...so I could only imagine that set up on an ej207..

[Lance]

I remember that thread on n a s I o c.


Much thanks for the real world local experience...always thought the ewg /vta option was loud/scary...

Drove next to an e8 with that set up and I could not hear my own car...was a wtf moment...so I could only imagine that set up on an ej207..

Mine is loud but not as loud as some of the others I've heard.

I think its because I don't have a screamer pipe attached to the wastegate outlet.

After I routed the flange away from the steering mechanism I realised that the wastegate was already venting away from any of the engine parts and acessories. So I just left it as is lol

[castortroy]

Any advice , observations in going ewg ?

Supposed to help with better boost control and reduce chance of det by allowing air to pass through faster....

I'm on grimspeed as well...tial recirc...alot better drivability vs -Greddy vta ...esp when off throttle ...iwg still...

Going EWG is probably the best boost control upgrade I've made.

It has really helped to dial in the boost response perfectly.

If I had two recommendations for those going EWG it would be regarding waste gate placement and venting.

I chose to mount my wastegate on the up pipe. As oppose to welding it onto the exhaust housing like alot of other local guys.

Up pipe mounting is the harder route for us RHD owners as there are a lot of mechanical parts in the way. But trust me it's worth it.

I think that venting the exhaust gasses pre-turbo rather than inside the turbine housing has really had an impact on reducing turbulence and promoting a much more efficient flow of gasses in the setup.

Secondly, and noise aside, consider VTA rather than plumbing back into the downpipe. I think that this also has efficiency and restriction benefits. There is a post on NASIOC that seems to support this VTA theory (I believe it was you that posted it 16 cycles).

Of course there are certain benefits to be had from plumbing back the exhaust gases e.g. noise reduction, controlling exhaust fumes etc. So it's something to trade-off.

I'm not sure if i saw it anywhere, but is there a comparison of the EWG placements? The ATP i have is built for the EWG plumbed into the back end.

[Lance]

I have not seen any experiment done to test the various positions.

My personal preference is for an up pipe mounted setup because I believe it preemts turbulence issues within the exhaust housing. But this is just a theory.

ATP is a standup company and im sure they would have done considerable testing on their exhaust housing design before release.

Either option should be more than sufficient.

[castortroy]

True, may boil down to user preference, GS makes several variations of their gated uppipe for LHD