Finding dramatic improvements from small, sometimes intangible or even free modifications is often disconcerting to us-like finding huge power gains from just a little bit of ECU tweaking on a fairly stock vehicle. Don't get us wrong. We welcome the improvements, it just bothers us a bit. It's enough to make us wonder: If they're that easy and cheap, why weren't they implemented in the first place? We often try to figure out why the manufacturer might not have addressed a particular issue on the assembly line or on the drawing board. It often boils down to ugly things like emissions concerns, cost-cutting measures and plain old corporate practicality.
When you think of the '06 Civic Si, it's unlikely you'll conjure up images of accountants brawling it out with engineers in an effort to produce a vehicle economically on par with one of the company's lesser competitors at the expense of its technical traits. No, we have no qualms when it comes to Honda's commitment to quality, performance and attention to detail. We're not the only ones saying it either. This is arguably the best Si we've been handed in over six years. It also looks like there was a brawl, at least the engineers won. But when we picked up 30 hp in the Si's midrange all from just a simple ECU reflash, we were nothing short of confused. Impressed, but still confused. Hondas typically just don't see gains like this when it comes to ECU tuning. At least not ones like our test Si, with just simple bolt-ons. Either Honda skimped somewhere or Hondata's done something incredible.
We've learned a couple of things about Honda's latest K-series since we first got our hands on the Si. Although quite similar to the RSX Type S' K20A2 mill, the K20Z3 differs in more than a few ways. For one, Honda outfitted the Si engine with a chain-driven balancer in an effort to smooth out engine operation. We can tell from our experiences with the Prelude H22A engine that this added civility comes at a cost-a few horsepower to be exact. So Honda, in what seemingly appears to be an attempt at retrieving that lost power, outfitted the Si with one of the most impressive intake manifolds we've tested on the K-series engine yet. Most importantly, we've learned that the Si's ECU architecture is quite different from its K brethren. Its drive-by-wire throttle, which is controlled completely through the ECU, is the culprit. Now the ECU directly controls the vehicle's idle, cruise control system and throttle. Say good-bye to Honda's long-time TPS (throttle position sensor) and IACV (idle air control valve). As such, Hondata's popular Windows-tunable K-Pro doesn't work here. So Hondata did the next best thing: It developed a reflash compatible with the Si's K20Z3 engine paired with either of the latest RRB computers.
Hondata's reflash does three things, three very important things. First, it lowers the VTEC engagement point to a more useable 4,500 rpm-something we've seen pick up midrange power in the past on VTEC Hondas and is kind of a given when it comes to easy places to find lower-end Honda horsepower. With the Si's aggressive cam profile coming into action at 4,500 rpm, you can just watch the dyno curve take a turn for the better, almost straight up. How did Honda miss this? Just before 6,000 rpm, all this translates into our 30 extra horsepower we found. Sweet. Power gains are minimal at best from there on out and all tapers off just before 8,500 rpm-about the same as what we would have expected from the stock ECU with just a redline increase. But Hondata does something a little out of the ordinary when it comes to the VTEC switchover point: It's not really 4,500 rpm, at least not always. They created a window between 4,500 and 5,800 rpm in which VTEC may be activated. Below this threshold, it's always off. Above it, it's always on. But inside that 2,300rpm VTEC window, valve lift and timing are all determined based upon throttle. At part throttle, the camshafts make use of their milder side, while at wide-open throttle their maximum lift and duration are taken advantage of. Did Honda miss something here? Yes, but it was arguably intentional. The Si, like all K-series Honda engines, meets low emissions vehicle status requirements. So it's likely they weren't willing to push the envelope much further with this K engine in hopes of a bit more power at the expense of their coveted LEV-2 status. Who can blame 'em? Besides, this wouldn't be the first car we've seen detuned straight from the factory.
Second, when it comes to duration, Hondata addresses this here as well by means of advancing the intake camshaft angle. The K20Z3, like other K-series engines, employs two different forms of technology when it comes to the intake camshaft: Honda's tried-and-true VTEC combined with VTC (Variable Timing Control). VTC continually adjusts intake camshaft phasing electronically while VTEC continues to do what it's known best for-alter camshaft lift, duration and timing via its third lobe system on both camshafts. The good news for i-VTEC (VTC and VTEC) owners is that the VTC part can be altered within the ECU, which is exactly what Hondata did. In stock form i-VTEC allows for a maximum of 50 degrees of timing adjustment to be made. To get the best results power-wise, we really only need about 20 to 25 degrees of adjustment throughout the rpm range, but Honda implemented additional leeway for emissions reasons. While we like i-VTEC for the performance benefits it gives us, it's really responsible for why the Si burns so damn clean. Speaking of emissions, the good news is that Hondata's reflash makes little difference, so California readers can rest easy.
Third, Hondata also raises the rev limit from 8,000 to 8,600 rpm. Right now this change makes little difference since our power difference at the top end is negligible. We suspect swapping in a set of '06 TSX cams may prove different though. The added lift and duration that both the intake and exhaust cams possess should increase our potential for top-end power a bit. For now, our power gains remain in the 4,500 to 6,000rpm zone, which is after all a reasonable place to spend a lot of time at, especially since this is a daily driver. We should also note that our test Si was already outfitted with an AEM cold air intake, a DC Sports race header and a DC Sports exhaust system. Without these breathing upgrades, our 30hp gain realized from the reflash would likely be less-much less. Since the reflash finds much of its power from the advanced intake cam timing, a free-flowing intake and exhaust is pretty important.
While we were plugging in one of Hondata's reflashed ECUs, they suggested we try one of their heatshield intake manifold gaskets. Never one to turn away a product for testing, we obliged. Honda intake manifolds soak up a lot of heat. They're made of aluminum and are spaced away from the cylinder head with nothing more than a paper-thin gasket. The Honda-issued gasket seals well, even under high boost, but does little to prevent the intake manifold from taking on the heat of the combustion process and isn't reusable. Hondata's heavy-duty plastic gasket serves as an insulator and doesn't allow as much heat transfer. We measured the intake temperature at 135 degrees F at the end of the runners, right next to the cylinder head, prior to installing Hondata's gasket. This might not sound like a lot, but intake temps are directly proportional with horsepower-at least that's what we used to think. Hondata's gasket quickly dropped our temperature to a cooler 118 degrees. We thought this 17-degree drop surely would have translated into one or two additional horsepower, but this wasn't the case.
To be fair, we even revisited XS Engineering's dyno a second day to keep our outside temperature consistent. This didn't work either. The results: Hondata's gasket does its job, it dropped intake temperatures almost 20 degrees near the head and roughly 8 degrees inside the runners and near the plenum. This just didn't translate into a horsepower increase on our Si. But, to be fair, we have to give it up for the Hondata heatshield gasket. We have in fact seen it produce power on other vehicles and, at the very least, $50 is a small price to pay for a reusable gasket. That's reason enough for us to want one.