What a boring title… but I’ve got nothing catchy for the title as I gaze at my monitor through allergy induced tears and catching up on the “where’s my toon bro” emails after a crazy week that involved 3 days of parts testing & tuning at the shop that pulled me away from my normal routine at the desktop computer. Yes there was a joke in there, I know my humor doesn’t translate well on the interwebs at times so can I at least get a “Haha” before someone calls me an asshole?
But down to business! We have ECUTek as our tuning software for all Subaru platforms, and this week our victim was the VitTuned 2016 Subaru WRX. Love them or hate them — I don’t care, I enjoy working on a variety of platforms and Subaru is no different. I want to give PRL Motorsports a big shout out for supplying me with a full array of bolt ons to test on our car. This was also a great opportunity to break in the new AWD Dynapack setup at the shop.
The parts we’ll be using.
- PRL Motorsports TGV Deletes & EGR Delete
- PRL Motorsports Intake Kit & Charge Pipe Upgrade
- PRL Motorsports J Pipe
- PRL Motorsports Front Pipe
- PRL Motorsports Front Mount Intercooler
- STM Exhaust
I broke up the testing into 3 parts. First I did the car completely stock — just tuned it. Next I installed the intake upgrades (less the intercooler) and retuned. Finally I installed the full turbo-back exhaust setup and the front mount intercooler (you’ll see why…).
All these tests were performed on our Oregon 92 octane. No extra ethanol blending at all.
Part 1 – Stock Tuned
We’re using a Dynapack — so obviously it’s going to read super high and we’re going to be seeing rated crank numbers at the hubs… right? LOL, right… Not on this Dynapack. With an AM (Advance Multiplier) of .88 we had a baseline of about 210whp. After spending some time retuning the car I got it up to 240whp and 265wtq. Not a bad gain at all for a stock car. I spent time mapping the dual cam timing system and found that the stock settings were pretty much spot on with the stock car. Most of the extra power was found in cleaning up the boost curve and raising boost targets — a little bit in the timing map, but not a whole lot as the motor was definitely a bit touchy on the pump gas.
Part 2 – Intake Side
I was able to install all the intake parts without even removing the car off the dyno. On went the intake & charge pipe upgrade for the stock top mount. On went on the TGV deletes & EGR delete. The TGV’s were a very quick swap — each side came out in seconds (no trouble with the driver’s side getting stuck anywhere when removing it). The intake fit like a glove as well. I was able to hop back in the car and retune it again. It was a bit hotter this day and I was seeing 10-15* higher charge temps than when the car was tuned completely stock — however we saw a solid gain over our “stock tuned” baseline (to the right). It was pleasant to see that boost came in a considerable amount sooner, resulting in more torque a lot sooner in the curve. The gains over completely stock are on the chart to the left.
Part 3 – Exhaust (and FMIC)
The car came off the dyno and went on to the lift for some surgery. I started with the full exhaust setup. One look and I knew the stock J pipe was going to require some luck — those damn studs and nuts love to strip or come out as one piece. Luck was definitely on my side, two of them came out with no problem and the other two were saved by our tap kit and one Honda nut (haha!). Seems Subaru just loves their seized hardware — only other car this bad is the shop 370Z (good luck removing those cats!).
But once the stock exhaust components were off — all the PRL parts went on smoothly. The items were well crafted and up to the quality I’ve come to expect coming from PRL. The STM exhaust bolted without much fuss at all as well.
Finally I put took the bumper off and fitted the PRL front mount intercooler setup. Having done quite a few PRL turbo kit installs (we run two of their kits on our shop S2000 & FR-S even!) the intercooler for the WRX is just as beautiful as the ones they provide for all their other kits. The bypass valve is relocated to the passenger side of the bumper — which is a nice location as it makes servicing or replacing it easier in the future.
Now I had wanted to test the FMIC all by itself towards the end… but I’ll get to why I installed it while the car was already on the lift (other than it’s a royal pain to take AWD cars on and off the dyno, hah!) a bit later.
The car went back on the dyno, and now that I had all the exhaust components done I wanted to see what this little turbo could really do — and I found some annoying ECU related nuances along the way. No big deal, something for the engineers at ECUTek to dig into in the ECU code — have to make sure their day isn’t boring either.
Once I was comfortable with how the motor was behaving with the new mods (checking all the cam phasing as well), I wanted to see what kind of power I could get out of our car by going “all in” on the boost levels — let’s see what the turbo can do.
Given we have a roughly 2.7 bar manifold pressure sensor on the vehicle stock, I wanted to get up to those boost levels — and I did. The graph to the right demonstrates what happens when I target right up to the clipping limit of the map sensor and then taper boost down (as the turbo can’t hold this boost level anyway). The torque is fantastic — even with a conservative timing map in the peak torque area. 330whp and 365wtq on 92 octane — not bad. But you’re going to ask me about that torque dip at 4400 rpm — and you’d be right to! At first I thought it had something to do with the fuel system (pump not keeping up, DI pressures dropping) — but nope, everything is rock solid. After a few days of street testing since these dyno tests were done I can repeatedly duplicate this issue — it happens anytime boost pressure get up to the 2.6 bar absolute or higher area. In the datalogs you’ll see the AFR on the factory sensor read 12.4-12.6 (not that scary right? on the dyno tail sniffer it was 13.4-13.8, so a bit more concerning…), and it appears the ECU is applying some sort of torque limit or power reduction via fueling (seen this behavior on other ECU’s). I’ve been on the horn with ECUTek and we definitely have some digging to do.
So calling this our “all in” pull, let’s see what happens when we run a more conservative tune? Calling this our “safe” full bolt on run, you can see that dialing down the boost levels the torque level gets flatter and the ECU behavior going through that area isn’t pronounced (in fact power gets a bit better). One of those “tuning” battles… is fighting what the stock ECU wants you to do, even if that’s not what you want to do. How I would love me some MoTeC right now…
This is where the pretty graphs come in! After spending two days tuning against the climbing charge temps with the factory hot mount, I was ready for the FMIC upgrade. Having owned and tuned other platforms with top mount intercoolers and run them at the track, the heat soak is brutal (even at the drag strip — we’d see staging temps of 50-60 degrees Celsius on a good pass).
With the PRL FMIC and even more boost our charge temps actually continued to DROP after the pull started — and the temps started lower to begin with. With the factory hot mount temps would just climb every pull. Does this have an impact on power? Absolutely. There shouldn’t even be any argument here.
Now I’m ready for some rest and my weekend — and the car is begging for E85 (next week?).
I just wanted to briefly touch upon this point as a little birdie mentioned that some have claimed the PRL Intake has a “whack” or “terrible” MAF curve. I’ve found this to be absolutely false. I found a very clean MAF curve when tuning this intake, stock I/C or their FMIC. I’ve been tuning MAF for something like 14-15 years, it’s actually a break to tune a MAF vehicle — it’s quite easy compared to some of the other projects we tackle.
But what about your fuel trims you ask? Here we have a nifty graph that not only includes the fuel trims from a 45 minute drive, but a nice mean line to demonstrate the average of all the data sampled across the whole datalog. Note how the mean stays very close to zero — our long term has a 2% drift in a couple of areas and our short term is overall +/- 4% from the mean with one spot that drift ab it towards 6 with some blips in the 8% region. Not exactly bad for a MAF curve that literally came off the dyno and I drove the car home. One minor tweak and she’ll be tight around +/-5%. That’s pretty damn good for an aftermarket intake.