Automobile Forums  Where Car Enthusiasts Go to Discuss Cars Online

Ask your questions and fight for your choice of turbo or no turbo, as discussed in Jason Camissa's "Truth in Numbers" feature, here.  

It appears that no measures were taken to ensure adequate cooling air to the Mazdaspeed3's top-mounted intercooler. The airflow from the fan at the front of the car is going across the top of the IC but not through it. When the car is on the road, air is drawn from the front of the car and directed through the IC. Wouldn't this affect the results? A search of Google shows that a common technique among tuners of other cars using a TMIC is to attach a box fan to it. This would have put the TMIC on an even footing with the front-mounted IC on the GTi, and perhaps more accurately reflected the nature and amount of the power available when the car is being driven. Was this issue addressed in setting up the dyno?  

Hey there,

You're right, we did not have a fan blowing down through the MazdaSpeed 3's TMIC.  Like the other cars, it had a high-volume fan in front of the car to ensure adequate engine cooling and intake air at ambient temperature.

Sure, adding a fan directly on top of the TMIC could have (in theory at least) added a few peak ponies to the MazdaSpeed's numbers.  But in fairness, that might have given the 3 a competitive advantage relative to the other cars that didn't have a fan attached *directly* to their intercoolers.

Far, far, far more importantly:  Worrying about the peak numbers is missing the point of looking at a dyno graph. If we iced the intercooler, added a couple more fans, and waited until it was -20'F outside (which, in Michigan, should be about... next week), we could have gotten another couple peak WHP out of the car.

But unlike tuners, maximizing the peak number wasn't our goal.  The gain we would see is likely no more nor less than the normal variance you'd see between two stock MazdaSpeed 3s anyway.

All of us would have loved to see our favorite car of the bunch put out ridiculous power. But I think as a whole, we did the test fairly to all of the cars, and the results should accurately reflect real-world conditions.  And by "results" I mean the shape of the curve, *not* its peak number.  Forcing air down the intercooler might have got us a few more peak hp, but wouldn't have changed the shape of the curve, which is IMO the most important piece of information about the MS3's dyno run.  

I was looking over both the torque and HP curves of the Mazdaspeed3. The drop in both torque and HP after 5500 RPM seemed to me a bit odd, considering I really have not seen this kind of drop that early before red line on any other dyno tests of a car in good mechanical condition. In your feature you mention that the turbo runs out of breath, but does that not usually result in a flattening of the torque curve rather than a dramatic drop in torque as seen on the dyno chart? Also, I am curious to know what gear the car was tested in, since it is my understatndin that torque in first and second gear is limited by the computer.  

Hey there,
The dropoff in power over 5500rpm *felt* pretty odd to us, too -- which is why I was so anxious to see the dyno results.  Interestingly enough, Car & Driver just published their test of the MS3, and in it they mention that Mazda told them to shift well before the redline to get the best acceleration runs.  So obviously this wasn't a fluke on our vehicle.

A big drop like that is indeed unusual.  We can't know for sure what's going on, but boost is almost certainly the culprit.  Either the turbo's run out of breath or the computer is pulling back boost.

The computer can limit torque in 1st through 3rd, so all runs were done in 4th gear.  

Jason  

An update on the TMIC cooling thing...

Mazda contacted us said that they've managed to get repeatable 242 whp and 282lb-ft out of their cars by using a ($1000) 50-mph fan blowing on the front of their cars, and a powerful house fan bungee-tied to the top-mount.  Ouch, that's a lot of power!

The product development engineer who gave me those numbers knows his stuff... enough so to admit that the GTI's intercooler also wasn't getting the air-flow that it would at high speeds.

Again, our goal wasn't to get peak numbers, just to look at how differently the cars made their power.  

Of course, I'll be curious to see how much MS3 buyers can get their cars to put down when they resort to all the old tricks - icing the intake plumbing, spraying water on the intercooler, etc.  But that certainly wasn't our goal here!

Jason  

I am also curious what gear they did the testing in. The Speed3 computer limits the max boost output limiting the power in 1st, 2nd, and I think 3rd gear. Just another variable to look into.  

Excellent stuff. Thanks for following up with Mazda. And while I appreciate what you're saying about advantages and all, it seems to me that you gave the GTi an advantage since the fan was at least blowing through its IC instead of over it. A $1000 50 mph fan is not crazy expensive or  crazy powerful and how powerful could a house fan be? My point is that this is not trickery to get max peak numbers. It's just trying to approximate real-world airflow as much as possible. I'll definitely run my first dyno with a box fan bungeed on the TMIC.

Again, I appreciate your efforts to delve into the totally different ways the GTi, Si, MS3 engines make power. Whenever, I step on the gas in my MS3, I think I made the right choice, but can easily understand how others will find that the GTi or Si do a better job for them.  

The Civic Si is a sophisticated.  The Mazda is crude.  Scalpel versus Battleaxe.  They are both great additions to the choices that are out there for enthusiasts on a budget.  I prefer the Honda's approach but it always seems they miss the power/torque marks by about 20%.  The Civic Si sedan appearance is great with just a hint of agressiveness.  The others look like atomic potato blobs.  The Si would be perfect with a normal dash option and another 30 BHP.  

HELLO GTI in the house veedub!. The Si does look good but The gti and the DSG rule! the gti dosent cost that much more and is more suave than the fast furious Si. I would choose the GTI but the si is still not bad. Potao blobs!! The GTI's "Fast" could kick the Si into next year.  

My buddy has a 07 MS3, I am surprised at the dyno numbers that I have seen thus far. The reason I say this is we just had the car dynoed @ 89 degrees and 40% humidity in South Florida just recently and the car put down 260FWHP and 260TQ. The car only had the catalytic removed and one of the resonators thats it! I have the dyno sheet to prove this. By the way the results I have seen have all been from Dynojets which is the same  dyno we used. I will post the dyno sheet if anyone is interested and I will keep you posted when the other mods go in soon such as a 3 and a quarter inch TMIC, CAI, and a new BOV to the market. Once we get the numbers from having the car breathing stock, then we will toss in an upgraded turbo kit and post those numbers as well. I almost forgot the first pull the guy took the car to redline in third gear and the car put down 245FWHP/259TQ, second run performed we did a fourth gear pull and this time the car was taken to 6,000RPM and the car put down 254FWHP/259TQ. The third and final run was performed and this is when the car put down the numbers mentioned earlier (260FWHP/260TQ). Gotta love those Wednesday motors! Take care all!  

I am the owner of an MS3 so am a little perplexed by the "stated" results for the three dyno tests - not the charts themselves, but what they supposedly represent. In looking at the charts I noticed how similar the torque and HP tracking were between the MS3 and the GTI, the two turbocharged vehicles, (although the GTI's was more sustained) and how different (linear) the torque curve was for the SI. I also noticed that the rpm axis begins at 2000 for the SI, but begins at just over 1000 for the MS3 and GTI which, if you don't catch this, makes the turbos look pretty weak compared to the naturally aspirated engine. My question is what do the torque and HP curves really mean in daily driving or in street racing? If I know that the sweet spot for the MS3 is between 3000 and 5000 RPM and can up or down shift to keep it in this range -what does a 4th gear dyno test tell me? If I shift from 1st to 2nd at 5 MPH, to 3rd at 15, and to 4th at 25 then the torque and HP numbers would be sad indeed, but who drives this way? To me this "truth in numbers" comparison seems more like a "make the numbers say whatever you want" comparison.

Although the Mazda may have a shorter curve than the GTI or SI, it holds 90%+ of its peak torque from 2900 RPM to 5100 RPM or 36% of the range from idle to redline (750 RPM to 6750 RPM).  This 220+ foot lbs. of torque may not last as long as the torque curves for the GTI or SI, but both of those motors have that amount of torque available to them for exactly 0% of the time – with the SI not even coming close.  My lawnmower may hold 95% of its torque from start up to blow up that doesn’t make me want to start driving it.  I would always pick a shorter, but significantly higher torque and HP curve to a longer, but lower curve.

As to the idea that turbocharging (or supercharging?) are cheating I don’t understand this way of thinking.  Getting more (sustainable) power from smaller and lighter engines seems smart - not unfair.  Both of these approaches go back as far or farther than the internal combustion engine itself.  The V8 engines of yesteryear were woefully lacking in power, but technology  made great strides in increasing their output.  Being down on advances would have us driving 120 HP V8’s getting 10 MPG – not something I would be interested in!  As for the brief mention of turbo lag, the ½ second difference between pedal down and power seems preferable to, for example, resuscitating the SI up from its flat line to 6000 RPM .  The 2.0T DIGI and 2.3 DISI engines of VW/Audi and Mazda, respectively, have all but eliminated lag and unless I’m mistaken, both made Wards Top Ten Engines list.