K24A2 dyno testing day!

David Calzada

Last week we did a full day of independent dyno testing that I want to share. Mikey from MSpec Tuning in Schaumburg, IL was gracious enough to give us 12+ hours on the dyno to do back-to-back testing. The goal was to show off what can be done with a stock K24A2 engine (2004-2008 TSX) in the Miata chassis, as well as test out a couple new parts. All tunes are done with the common 50 degree VTC intake cam gear swap instead of the 25 degree stock TSX unit. Each combo was fully tuned, we weren't just bolting on parts and doing dyno pulls.

The best part is that the first two tunes are going to be available pre-flashed into a stock ECU at a very reasonable cost. There will be more info on that in the coming weeks.

First, we did a baseline tune with the stock K24A2, our Street header (the 4-1 design), and the Skunk2 Ultra Street manifold we now carry. Ignore the RPM signal loss that spiked the torque figure up a bit. Here's the graph showing 222whp and 182wtq, revving to 8000 rpm.

Next, we pulled the car off the dyno and swapped on our Race header (4-2-1 design). As you can see, this header smoothed out the midrange dip in power, and carried a similar powerband up top (You can see that the 2 hp difference is just from that little 8000 rpm spike in power on the 4-1 tune). The 4-1 header has 1.75" primaries, and the 4-2-1 has 1.88" primaries, so even with the huge primaries it does well on a stock motor. The Race header really flexed it's design more when we tested it earlier this year on a K20/K24 combo with some big cams, where it made more power over all other headers by about 3 hp.

Also, we removed the torque plot on the graph because rpm signal loss. We had the same peak torque but did pick up some midrange in the same area that we increased hp.

Next, we wanted to test a possible new product. Skunk2 sells some small spacers for their Ultra line of intake manifolds. For the Ultra Street, their spacer only increases the plenum volume by 0.5L. Considering that the manifold with no spacer is only 1.81L, we knew we wanted to bump the volume up considerably, as other big plenum manifolds have shined on this car before. Instead of stacking $400 worth of Skunk2 spacers, we had a 2" thick test spacer machined, bumping up our total plenum volume to 3.2L:

So, did it make power? On a stock K24A2, not at all! The red line is with no spacer, and the green line is with the spacer added. This was pretty surprising to us.

Next, we pulled the car off the dyno to swap in some different camshafts. After speaking with our friend Jeremy at Drag Cartel, he recommended that we test a set of his 3.2 cams. Cams were installed, and a proper valve lash adjustment was done before strapping the car back on the dyno. We previously installed upgraded valve springs in this engine because we knew we'd be testing some different combos. The results were interesting. Although we lost a little midrange, from 6000 to 8200 rpm, the top end screamed and we gain 13whp. Torque stayed the same around 180.

I think a set of the smaller 2.2s may be a better fit for this engine with the K24 head, as I'd like to see a bit more midrange, but the car feels fantastic in this configuration and will be a faster car for any form of racing. Based on an engine we tested last year, I think a simple K20 head swap and these 3.2 cams should easily put the power output to 240+whp.

Lastly, as a test control, we pulled off the manifold for the last time and removed the spacer. We knew that the spacer wasn't giving us gains with stock cams, but we were curious to see the difference with the cam swap. Red is with the spacer, blue is without. You can see, the spacer is giving us about 3 hp the entire time the engine was in VTEC. Again, I bet the difference would be even more significant with a K20 head, and certainly with a built engine.

Based on this test, we can reasonably conclude that the more air our engine can move, the more helpful that additional plenum volume will be with this manifold.

This last graph shows the power gains from the day. The blue line shows how it rolled in, and the red line shows how it left.

Overall, we're really pleased with the results, and I think this gives customers some nice options and data to help plan out their builds. We do plan on testing a K20 head on this engine. The K24 head seems to be holding us back a bit, and K20 heads are readily available on the cheap.

Thanks for reading!