MazdaSpeed Dyno Testing
by Chris Black
The first thing I needed to do was get a good baseline for the car. Dyno testing a car with a top mounted intercooler can be tricky and it adds another variable to our normal equation. It is tough to get a fan that will supply the necessary, or ‘correct’, airflow to a top mount intercooler. Secondly, it is hard to know if you are duplicating results that will typically be found on the road, or if you are ‘over cooling’ the intercooler with your fans, if there is such a thing.
Regardless, there are several tests that need to be done. First, we need to make a good base line run with the car. We need to do this without my intercooler fan rigging, and with it to see if it will make a difference. Also, during these tests we will want to check the effectiveness of the intercooler by looking at the temperature of the charge air before and after the cooler, with and without the fan. The next test that we will need to do is to raise the boost! We have had some experience with something similar on a Mazdaspeed 6 so we are only testing to see if the results are the same. The final test that I have planned at this point is to remove the exhaust past the downpipe and run the vehicle on the dyno to see if there are any plausible gains to doing an exhaust system for this car in stock form.
Test 1: Base Line Completely Stock Car, No Intercooler Fan VS with Intercooler Fan
As can be seen below there is no discernable difference in power between the two runs. The boost on the Mazdaspeed3 is not always consistent depending on which rpm you start the run at. After battling with the car for a little while, I was able to reproduce the same run from the car each time and conduct
![]()
Test 2: Charge temperatures
In order to do the next tests, I needed a thermocouple in the actual air stream before and after the turbo. Using some thermocouple wire rated for the temperatures we should be seeing, I plugged it into a converter that allows any ordinary multi-meter to read the temperature in millivolts. I installed this wire by disconnecting the intercooler coupler, sliding the wire in, bending it so that it would be in the middle of the airstream and then re installing the intercooler coupler and tightening the hose clamp. The pliable nature of the coupler allows me to do this without any boost leaks for an accurate test.
With my trusty thermocouple installed in the intercooler ‘piping’ right after the turbo, I measured the air charge to be about 230 degrees Fahrenheit on my first run. The second run on the no longer cold vehicle netted a result of 265 degrees.
Using the same methods I moved to the cold side of the intercooler ‘piping’. On the cold side of the intercooler, with no intercooler fan on a car that was already hot, I saw the charge temps rise from 100 to 170 degrees by the end of fourth gear.
The next tests required me to temporarily install an intercooler fan on the car, in order to directly cool the intake charge on the other side of the intercooler. While I was setting this up and before I could actually get into the car and set up the test, the intercooler fan would be running and therefore cooling the intercooler to the fullest amount (ambient temperature). This is not the best method for testing, because it does not mimic the conditions that are available on the road. Surprisingly enough, the runs did not have very much, if any, difference at all.
![]()
The results that I got from the thermocouple and the results that I got from the dyno only make sense when you start to look at the way that the ECM controls the systems that are onboard. The results I had on the intercooler outlet were:
No IC fan: 100 to 170 degrees
With IC fan: 70 to 120 degrees after cold soaking
With IC fan: 80 to 130 degrees at close to normal operating temperatures
No IC fan: 70-120 degrees after cold soaking
With a decrease of about 50 degrees you would expect to see a noticeable horsepower difference. In some circles this kind of temperature difference would be computed at a 5% gain to the horsepower and torque level. So where did the 14hp go?
Data Analysis 1: Throttle Position vs. Intake temperature
In order to answer this question, we are going to have to look at some of the data from the OBD II logger I am using.
![]()
If you take a look at what is going on above you can see a very erratic red line. If you take a look at the key on the left hand side you will see that this red line corresponds to the throttle position angle. Notice, this is not the throttle position from inside the car, this is the actual throttle valve position, as controlled by the ecu. For this entire run, the throttle pedal inside the car is being held to the floor. The yellow line indicates boost pressure and the green the RPM range. At what appears to be about 2000 RPM the throttle valve closes rapidly to about 50% holds there until full boost, which is just after 3200 RPM. At this point the throttle valve opens very rapidly, but not to 100%. Honestly, there is little to no difference in horsepower in throttle position openings above 75% on a forced inducted car.
The real issue, or the answer to our lack of horsepower from above, lies in the throttle position in the last 2000 RPMs of our pull. As can be clearly seen, throttle angle is tapering off and the line directly above it, boost pressure, is following suit. The reason why the horsepower drops is not only because of the reduced throttle angle, but it is also a product of the pressure differential that it is creating. Pressure differential is what the blow off valve uses to sense a closed throttle and vent the pressure back to the intake tube in a re-circulated system. The boost pressure would most likely drop in the higher RPMs anyways, but it would also most likely follow the trend in-between the RPM ranges of 3200 and 5000 RPMS, dropping to something slightly over 12 or 13 psi most likely. A stock Speed3 drops the boost pressure all the way to 5lbs by the end of the run! If you take this info into consideration with the intake temperatures that are listed above, it is easy to see where our extra power is going. Although the charge temperature is much cooler, all of the energy and pressure is being vented!
Test 3: Install the boost controller!
One of the best things about factory turbocharged cars is the ability that you have in the aftermarket to make a few small changes and totally wake up the car in comparison to the factory tuning. For the test I will be using one of the manual boost controllers that we use regularly at the shop. Hooking it up was not the easiest thing to do. The stock boost control method involves a pressure port off of the compressor housing, which is linked to an internal wastegate actuator. The WGA has a second smaller port on it which is designed to bleed off the pressure that is supplied from the compressor housing. This port is attached via a vacuum line to the boost control solenoid.
In order to switch it to the boost control method I wished to use, I needed to cap off the secondary and smaller port, and interrupt the stock line from the compressor before it got to the wastegate. This line would serve as my pressure source for my boost controller. From the other port of my boost controller the line would attach to the WGA.
From previous tests on a Mazdaspeed 6, we knew that the boost cut was set to about 17-18 lbs. I assumed the Speed3 would be roughly the same and set out to find the max boost possible.
As depicted in the chart below, raising the boost definitely gained us some horsepower. Initially it didn’t look like the ecu was too happy with us during spool up (could also be the throttle). Once the car was in full swing you can see where it gains the most… everywhere! Run 7 is our reference run from earlier. Run 12 is a run on about 17lbs of boost. Run 13 is a run on 18 lbs of boost. Do not pay attention to the maximum numbers that are listed for this run. It hit boost cut and it messed with the way that the dyno reads, giving it a spike that is much higher than the actual horsepower that it made. You can see in some areas where it gained up to and over 30 ft lbs of torque! Although the boost still drops off in the top end I managed to get it to carry just a little bit longer and make quite a bit more horsepower
![]()
Test 4: Removing the Exhaust
Its time to see if the exhaust is a restriction or if it is properly sized from the factory. The easiest and most cost effective way to do this is to take if off of the car and put it back on the dyno.
In order to see the changes that were made, I opted to show just the torque figures so that it will be easier to see the differences. Surprisingly it is not that huge of a difference on a completely stock car. During most of the power band there is little to no difference. During the spool up area of the turbo there is a sizable difference. From the data it looks as though there is about a 200 to 250 RPM increase in spool time, and as a result, at the same RPM there could be about 30 more ft lbs of torque. The other gains can be seen just over 5800 RPM. The power is extended a little further, presumably because it is more free-flowing than the stock exhaust at high rpms. There is also a gain in torque here at a maximum of 20 ft lbs. It should be mentioned that Run 3 is one of three runs that were identical. I only am showing one for ease of viewing.
![]()
Conclusions and first impressions:
By now, I’m sure it is apparent that we need some tuning software for this car. I think the gains that we can make with the AccessPort are going to be AWESOME. First of all, we will be able to control the throttle opening (hopefully), and remove the boost limit (raise it to a safe level). At this point we will see the gains from the exhaust and the intercooler. This thing will be a rocket.
All of those gains will be BEFORE tuning. I’m sure Mazda has left quite a bit on the table (as I’m sure all manufacturers do for safety reasons) and we will be able to extract ridiculous numbers out of the car compared to when it was stock.
The tests above outline the ability of some simple aftermarket modifications effect on the power output of a completely stock MazdaSpeed 3. While some of them may not be producing the best results at this power level or state of tune, I can assure you that once they are properly tuned in, they will make the differences that you would expect.
After seeing what the car does now, and seeing the things that holding it back, it is easy to see the great potential of the car. With a few, well thought out upgrades, we will certainly be looking for a little more traction!
*Written by Chris Black, a tuner at AMS.*
Eric
by Chris Black
The first thing I needed to do was get a good baseline for the car. Dyno testing a car with a top mounted intercooler can be tricky and it adds another variable to our normal equation. It is tough to get a fan that will supply the necessary, or ‘correct’, airflow to a top mount intercooler. Secondly, it is hard to know if you are duplicating results that will typically be found on the road, or if you are ‘over cooling’ the intercooler with your fans, if there is such a thing.
Regardless, there are several tests that need to be done. First, we need to make a good base line run with the car. We need to do this without my intercooler fan rigging, and with it to see if it will make a difference. Also, during these tests we will want to check the effectiveness of the intercooler by looking at the temperature of the charge air before and after the cooler, with and without the fan. The next test that we will need to do is to raise the boost! We have had some experience with something similar on a Mazdaspeed 6 so we are only testing to see if the results are the same. The final test that I have planned at this point is to remove the exhaust past the downpipe and run the vehicle on the dyno to see if there are any plausible gains to doing an exhaust system for this car in stock form.
Test 1: Base Line Completely Stock Car, No Intercooler Fan VS with Intercooler Fan
As can be seen below there is no discernable difference in power between the two runs. The boost on the Mazdaspeed3 is not always consistent depending on which rpm you start the run at. After battling with the car for a little while, I was able to reproduce the same run from the car each time and conduct
Test 2: Charge temperatures
In order to do the next tests, I needed a thermocouple in the actual air stream before and after the turbo. Using some thermocouple wire rated for the temperatures we should be seeing, I plugged it into a converter that allows any ordinary multi-meter to read the temperature in millivolts. I installed this wire by disconnecting the intercooler coupler, sliding the wire in, bending it so that it would be in the middle of the airstream and then re installing the intercooler coupler and tightening the hose clamp. The pliable nature of the coupler allows me to do this without any boost leaks for an accurate test.
With my trusty thermocouple installed in the intercooler ‘piping’ right after the turbo, I measured the air charge to be about 230 degrees Fahrenheit on my first run. The second run on the no longer cold vehicle netted a result of 265 degrees.
Using the same methods I moved to the cold side of the intercooler ‘piping’. On the cold side of the intercooler, with no intercooler fan on a car that was already hot, I saw the charge temps rise from 100 to 170 degrees by the end of fourth gear.
The next tests required me to temporarily install an intercooler fan on the car, in order to directly cool the intake charge on the other side of the intercooler. While I was setting this up and before I could actually get into the car and set up the test, the intercooler fan would be running and therefore cooling the intercooler to the fullest amount (ambient temperature). This is not the best method for testing, because it does not mimic the conditions that are available on the road. Surprisingly enough, the runs did not have very much, if any, difference at all.
The results that I got from the thermocouple and the results that I got from the dyno only make sense when you start to look at the way that the ECM controls the systems that are onboard. The results I had on the intercooler outlet were:
No IC fan: 100 to 170 degrees
With IC fan: 70 to 120 degrees after cold soaking
With IC fan: 80 to 130 degrees at close to normal operating temperatures
No IC fan: 70-120 degrees after cold soaking
With a decrease of about 50 degrees you would expect to see a noticeable horsepower difference. In some circles this kind of temperature difference would be computed at a 5% gain to the horsepower and torque level. So where did the 14hp go?
Data Analysis 1: Throttle Position vs. Intake temperature
In order to answer this question, we are going to have to look at some of the data from the OBD II logger I am using.
If you take a look at what is going on above you can see a very erratic red line. If you take a look at the key on the left hand side you will see that this red line corresponds to the throttle position angle. Notice, this is not the throttle position from inside the car, this is the actual throttle valve position, as controlled by the ecu. For this entire run, the throttle pedal inside the car is being held to the floor. The yellow line indicates boost pressure and the green the RPM range. At what appears to be about 2000 RPM the throttle valve closes rapidly to about 50% holds there until full boost, which is just after 3200 RPM. At this point the throttle valve opens very rapidly, but not to 100%. Honestly, there is little to no difference in horsepower in throttle position openings above 75% on a forced inducted car.
The real issue, or the answer to our lack of horsepower from above, lies in the throttle position in the last 2000 RPMs of our pull. As can be clearly seen, throttle angle is tapering off and the line directly above it, boost pressure, is following suit. The reason why the horsepower drops is not only because of the reduced throttle angle, but it is also a product of the pressure differential that it is creating. Pressure differential is what the blow off valve uses to sense a closed throttle and vent the pressure back to the intake tube in a re-circulated system. The boost pressure would most likely drop in the higher RPMs anyways, but it would also most likely follow the trend in-between the RPM ranges of 3200 and 5000 RPMS, dropping to something slightly over 12 or 13 psi most likely. A stock Speed3 drops the boost pressure all the way to 5lbs by the end of the run! If you take this info into consideration with the intake temperatures that are listed above, it is easy to see where our extra power is going. Although the charge temperature is much cooler, all of the energy and pressure is being vented!
Test 3: Install the boost controller!
One of the best things about factory turbocharged cars is the ability that you have in the aftermarket to make a few small changes and totally wake up the car in comparison to the factory tuning. For the test I will be using one of the manual boost controllers that we use regularly at the shop. Hooking it up was not the easiest thing to do. The stock boost control method involves a pressure port off of the compressor housing, which is linked to an internal wastegate actuator. The WGA has a second smaller port on it which is designed to bleed off the pressure that is supplied from the compressor housing. This port is attached via a vacuum line to the boost control solenoid.
In order to switch it to the boost control method I wished to use, I needed to cap off the secondary and smaller port, and interrupt the stock line from the compressor before it got to the wastegate. This line would serve as my pressure source for my boost controller. From the other port of my boost controller the line would attach to the WGA.
From previous tests on a Mazdaspeed 6, we knew that the boost cut was set to about 17-18 lbs. I assumed the Speed3 would be roughly the same and set out to find the max boost possible.
As depicted in the chart below, raising the boost definitely gained us some horsepower. Initially it didn’t look like the ecu was too happy with us during spool up (could also be the throttle). Once the car was in full swing you can see where it gains the most… everywhere! Run 7 is our reference run from earlier. Run 12 is a run on about 17lbs of boost. Run 13 is a run on 18 lbs of boost. Do not pay attention to the maximum numbers that are listed for this run. It hit boost cut and it messed with the way that the dyno reads, giving it a spike that is much higher than the actual horsepower that it made. You can see in some areas where it gained up to and over 30 ft lbs of torque! Although the boost still drops off in the top end I managed to get it to carry just a little bit longer and make quite a bit more horsepower
Test 4: Removing the Exhaust
Its time to see if the exhaust is a restriction or if it is properly sized from the factory. The easiest and most cost effective way to do this is to take if off of the car and put it back on the dyno.
In order to see the changes that were made, I opted to show just the torque figures so that it will be easier to see the differences. Surprisingly it is not that huge of a difference on a completely stock car. During most of the power band there is little to no difference. During the spool up area of the turbo there is a sizable difference. From the data it looks as though there is about a 200 to 250 RPM increase in spool time, and as a result, at the same RPM there could be about 30 more ft lbs of torque. The other gains can be seen just over 5800 RPM. The power is extended a little further, presumably because it is more free-flowing than the stock exhaust at high rpms. There is also a gain in torque here at a maximum of 20 ft lbs. It should be mentioned that Run 3 is one of three runs that were identical. I only am showing one for ease of viewing.
Conclusions and first impressions:
By now, I’m sure it is apparent that we need some tuning software for this car. I think the gains that we can make with the AccessPort are going to be AWESOME. First of all, we will be able to control the throttle opening (hopefully), and remove the boost limit (raise it to a safe level). At this point we will see the gains from the exhaust and the intercooler. This thing will be a rocket.
All of those gains will be BEFORE tuning. I’m sure Mazda has left quite a bit on the table (as I’m sure all manufacturers do for safety reasons) and we will be able to extract ridiculous numbers out of the car compared to when it was stock.
The tests above outline the ability of some simple aftermarket modifications effect on the power output of a completely stock MazdaSpeed 3. While some of them may not be producing the best results at this power level or state of tune, I can assure you that once they are properly tuned in, they will make the differences that you would expect.
After seeing what the car does now, and seeing the things that holding it back, it is easy to see the great potential of the car. With a few, well thought out upgrades, we will certainly be looking for a little more traction!
*Written by Chris Black, a tuner at AMS.*
Eric
