ECU Tuning

« August 2005 | Main | October 2005 »

September 29, 2005

What makes a 330 hp smog-legal Acura RSX?

Every now and then I marvel at the fact that my rather stock-looking Acura RSX Type-S has a 330 hp engine under its hood, can do 0-60 miles per hour in 4.8 seconds and run the 1/4-mile in the mid 13s. What did it take to do all that?

Truth be told, I spent a lot more on the car that I needed to get all this power. A completely stock RSX with a suitably configured Comptech supercharger can get very close to the crank horsepower of my car. But that is not how I got started. Like most other modders, I added piece after piece, month after month, and each made me feel good.

First I got a set of nice 17-inch Kazera wheels from tirerack.com, with Kumho 712 tires. Then I installed a Comptech Icebox intake and a Comptech street header. Then a Fujitsubo PowerGetter exhaust. Then the Hondata ECU flash. For handling I put in the Mugen Sport Suspension. Then I upgraded the ECU to the incomparable Hondata K-Pro. The Hondata Intake Manifold Gasket went in. A tuning session led to switching to an AEM Cold Air intake. A B&M short shifter went in. For cosmetics I got a JDM-style wing, had the headlights blacked out, and redded the taillights. Then came the Comptech supercharger for which I played guinea pig for Comptech. In the process, they also installed their titanium front strut brace and rear sway and tie bar. Now that the supercharger was in, I got an Autometer boost gauge, and ACT performance clutch and heavy-duty pressure plate and a Comptech lightened flywheel. While the transmission was off, we also installed a Quaife limited slip differential.

Has it been fun? Absolutely. Tons of it. Would I do it again? Yes. Sure, for the extra money I spent I could easily have bought a car that already had that sort of performance, but that's no fun.

Posted by conradb212 at 2:04 PM

September 22, 2005

Live ECU tuning at the track, and a trophy

Last night I got my first drag racing trophy. It wasn't much of a trophy, just sort of a big mug with the Sacramento Raceway logo and graphic. And I was only runner-up, losing the final bracket race by 8/1000th of a second to a much slower but extremely consistent automatic. But it felt sooo good to get that silly trophy.

This was one of the last Wednesday drags of the season, so about 250 cars showed up. We only got two time trials, which was bad news for me as I had just installed a new pulley on my supercharger and needed to adjust the Hondata K-Pro calibration for it. I had done some street tuning, but the track is different. I always datalog all of my runs.

On my first attempt I did a 13.9 at 104mph, but my air-fuel ratio went into the low 13s above 7700 rpm. Boost was lower than expected, mostly in the high 7s with an occasional rise into the 8s. Anyone who has seen a Comptech boost curve knows that it is difficult to read because of some internal buffeting that makes the MAP sensor show peaks and valleys with amplitudes of up to 0.8 psi. I also noticed fairly substantial knock. None in second, 6 at high rpm in third, and another 12 all through 4th. Not good. I wasn't going to remap my entire ignition tables sitting in a line of cars, not knowing when my next time trial run would be, so I decided to try ignition compensation by gear. I quickly glanced at the K-Manager documentation: "To retard the ignition, enter a negative value." Retard... Retard.. okay, so that probably means a lower number, which I needed. So I entered -1 for third, and -2 for fourth and fifth. I also adjusted fuel, wondering if I was making a mess of things doing it all on the fly, and also what temperature compensation had already done with my starting line IATs in the mid-120s.

Anyway, I really blew my second run when I couldn't get it into 2nd, but the rest of the way it was fine. Datalog showed that total knocks were down substantially, and that gear ignition compensation had actually worked.

Then we got into tropy runs which I both love and hate. I am still on a quest for lower ETs, but since I also want to win, I usually get off the gas when an opponent is obviously beaten. First trophy run: 13.9 coasting through the finish line since my opponent, a nice E36 M3, had apparently misjudged his dial-in. Second trophy run: a twin turbo 280ZX who had a reaction time of 0.001, then ran under by almost half a second. I had an average run, but beat him because he ran under. Third run against a Prelude who also couldn't even come close to his dial-in, so I got off the gas early and did another 13.9 as I coasted through the line at 90mph. Semi-final against an Acura who had beaten me before. He ran a near-perfect run, but I beat him with a 13.8 on a 0.08 reaction time and went off the gas early. The runs now had no more knock and AF in the high 11s. Boost, however, was only in the mid to high 7s and actually decreasing a bit after 8000 rpm. Could it be that my belt was slipping a bit with Ryan at Comptech torqueing it the way he thinks is right as opposed to what the instructions say?

For the final run I lowered my dial-in to 13.7 because it was cooler now, in the 90s, and I did not want to break out. I got to race the dreaded middle-aged lady from Folsom who drove that 18.0 second Kia and generally ran 18.05s like clockwork with near-perfect reaction times. Needless to say, she didn't have wheelspin problems. The 4.3 second headstart was a bitch, I spun the wheels (as I did on every launch yesterday, despite the Falken 615s), caught her at the finishing line, inches too late. 13.8, 105 mph trap.

Fun evening all in all. Tuning the car majorly while waiting to stage is a handful. The knocks, AF problems, lower-than-expected boost and sky-high IATs kept me busy. My inability to find the proper launch infuriating as it probably added half a second to each run. The only other RSX there, I/H/E/Hondata with a good driver, did 14.6-14.8. The winner in the other import class was a S2000 who did 14.2s and beat a E46 M3 (the new one) who did 13.9s in the finals. Maybe Sac Raceway just isn't a very fast track.

So now it's back to the drawing board. My 3.3 inch pulley only does mid 7s to low 8s in boost. My launches bite; still in the 2.3s 60 foot. The calibration needs some real work; no way can the official Comptech Stage 2 calibration be used for the 3.3-inch pulley without drastic changes.

Posted by conradb212 at 2:03 PM

September 19, 2005

It's amazing that third party performance kits work at all

In general, superchargers are lauded for their simplicity and reliability. The blower itself is guaranteed for 100,000 miles and doesn't need any maintenance at all. Combine that with a company famous for its engineering, experience, and customer support, such as Comptech USA in El Dorado Hills, Calif., and you have an unbeatable combination. Well, yes and no. It really depends on your expectations. When you consider that it costs automobile manufacturer billions to design a new platform and millions to create even small parts, and that it then takes millions of test miles and a couple of years of actual production to work out all (or most) of the bugs, it's amazing that performance kits, like superchargers, work at all. I mean, the design budget is minuscule, and often only a few dozen of an item are sold. Yet, customers have very high expectations and are quick to diss a company or product if it doesn't perform flawlessly.

This morning I witnessed an example of the struggle relatively small tuning companies go through. I stopped by Comptech to have them replace my 3.6-inch blower pulley with a 3.3-inch one in order to increase boost from 6 to 8.5 psi. One of Comptech's experienced mechanics did the work. He found a couple of parts that showed wear and replaced them. Once everything was back together he started the car to see if everything worked as it should. It seemed that way until he pulled the throttle cable a bit to see if the belt was nice and tight under load and changing rpm. Well, as he let go of the throttle and the engine rapidly dropped revs, the belt sort of slackened on top of the power steering pulley for a split second. It looked like it hopped just a bit. Needless to say, you don't want that. The belt might come off.

A Comptech design engineer stopped by and the matter was discussed. Did part of the tensioner catch somewhere and not flex back properly? Did the tensioner move too much? Everything came apart again, and the engineer installed a small metal insert to stabilize the long bolt. The mechanic cautioned that the thread of the bolt might catch in the bolt. The assembly went back in, but the belt still did that little hop when the throttle was released. Was the belt too loose? It was tightened until it bottomed out. The hop was still there. So the tensioner assembly came off again and was taken apart. The engineer took the two springs and tested them on some machine. Aha! They did not have the proper tension. New springs were installed and the tensioner went back in. The hop was still there, which meant that the tension of the springs was momentarily overcome. The assembly was tightened beyond spec and that fixed the problem, but that did not seem an acceptable solution to the Comptech engineer. He came back with two small springs that fit inside the larger springs. They had to be cut to fit properly. Ryan, the mechanic, wasn't too happy as he still felt the steel insert would make the threaded bolt catch. Steve, the engineer, felt otherwise. So the tensioner assembly was fitted with the dual springs, and then tightened (there was discussion as to how much to tighten; the mechanic felt it should be less than what was in the instructions). Anyway, with the new double springs, the hop was gone. No one seemed completely happy as there was agreement that the threaded bolt MIGHT catch on the steel insert, and that might either eat away the insert or dig into the bolt. Nothing terrible would happen, but it just wasn't a totally satisfying solution.

Now multiply that experience with all the things that can possibly go wrong, or not quite right, in an aftermarket supercharger system with all its brackets, fittings, nuts, bolts, belts and whatever else, and consider that the kit then gets installed into cars with very different modifications, often by people who are not professional mechanics, and it's a miracle these kits work as well as they do. The lesson really is that we should never expect perfection and flawless operation. Those who can't stomach that possibility should probably remain stock.

Posted by conradb212 at 2:02 PM

September 6, 2005

Back at the track with new tune and Falken RT-615 tires

Comptech had my Acura RSX for a couple of days so they could dynotune the Stage 2 supercharger Hondata K-Manager calibration for their modified Honda injectors. After I got the car back, there was a nasty noise on acceleration, and the tune didn't look too good. I brought the car back, feeling quite uneasy about the noise. Then again, the worst-case scenario was something wrong with the motor, and that would have given me an excuse to go for forged pistons and stronger rods. Fortunately, the problem was minor: the power steering line was touching the tie-rod, and that caused noisy vibrations. Billy at Comptech had managed to retune the fueling nicely and the car ran well, clocking in at 257.5 whp on Comptech's DynoDynamics chassis dynamometer.

In the meantime, my Falken RT-615s had arrived at Big-O Tires in Cameron Park. I had high hopes for them as their rubber compound is much stickier than the Kumho Exsta 712 tires I had before.

I went back to Sacramento Raceway to test that theory. Once again it was very hot, over 100 degrees. Though I got smoked by a turbocharged Civic in my first time trial, it was immediately obvious that the Falken tires provided much better grip. As before, I datalogged the 1/4-mile runs on my notebook. Whereas before I could hardly apply 30-40% throttle in first, now I could apply 80 and even 100%. I finally broke into the 13s and ran farly consistent 13.9s all night long. That is still pretty slow given my trap speed of over 103mph, so I will have to work on my launch technique. It was a fun evening and I made it all the way into the final trophy run in my class. I lost that when I spun the tires on the worse of the two lanes.

After a few trips to the drag strip, it's obvious that getting good times is a lot more difficult that many people think. And a drag strip is not this perfect, grippy, smooth piece of asphalt; it's an ever-changing patchwork where you never know if you'll get traction or not. Still, it's tons of fun, especially with being able to datalog all runs for post-race analysis.

Posted by conradb212 at 2:02 PM