PART IV: UPGRADING THE TURBO SYSTEM

This is the fourth and final installment in our Project RX-7 Turbo series. In the past three months, we've
optimized the performance of our 1988 Mazda for SCCA C Stock Solo II competition, then upgraded the
suspension for A Street Prepared action. The final step is to upgrade straight line performance for Street
Prepared competition. In keeping with the nature of our project, we want to increase power without
sacrificing new car reliability; the simplest way to do this is to increase the turbo's efficiency by adding
Cartech Performance Systems' intercooler upgrade kit.

In order to develop products for upgrading the Mazda Turbo II, Cartech spent ten months living with the
factory turbo under various conditions and at varying levels of modification. As with all their development
projects, Cartech development manager Todd Wilson went through an extensive exercise in determining
baseline turbo, intercooler, fuel, air inlet, ignition and exhaust system efficiency prior to making any changes.
This included instrumenting and measuring most or all of the following on the stock vehicle operating on 92
octane fuel:

Ambient air temperature
Barometric pressure
Air/fuel ratio*
Vacuum ahead of compressor inlet*
Compressor discharge pressure (intercooler inlet)*
Compressor discharge temperature (intercooler inlet)*
Intercooler discharge pressure (before throttle)*
Actual pressure in intake after throttle plate
Exhaust manifold pressure before turbo*
Exhaust backpressure after turbo, ahead of all converters and mufflers, plus individual readings ahead
of each individual point of restriction
Exhaust gas temperature*
Tailpipe temperature*
Actual cooling system temperature
Underhood temperatures at air filter inlet, behind radiator, air inlet to i.c., air discharge from i.c.
Timing changes related to boost pressure - at varying levls betwen 0 & stock max boost

Once these baselies are complete, a plan is outlined to concentrate effort an those areas weakest in the base
car's various systems. In the case of the Mazda RX-7 Turbo II, the results showed that the following major
areas could be improved.

PROBLEM AREAS & SOLUTIONS - EXHAUST BACKPRESSURE

The stock exhaust has four points of restriction. First and second are the two monolith convertors on the
turbo outlet; third is the main catalytic, and fourth are the dual rear mufflers. Starting from the rear, the
measured amount of pressure ahead of the mufflers showed that the factory produced a good-flowing
muffler with only 2 psi of restrictions. The main cat showed a tolerable 3.5 psi restriction, and the two
monoliths rendered an additional 5 psi restriction.

So what does backpressure have to do with performance? Simply put: If the burned exhaust can't flow out
the tallpipe, then a reversion of outward flowing gas occurs at the combustion area, along with not allowing
maximum inlet flow, this increases combustion temperatures and lowers horsepower. Even more directly,
lowering this backpressure in a turbo car results in a noticeable increase in boost response.

Cartech offers the autocrosser, for off-highway use, a high flow header from the back of the turbo to the "Y"
intersection of the stock exhaust. This bolt in/out replacement eliminated over 80% of the total exhaust
restriction, and wakes up the boost response dramatically.

INTERCOOLER EFFICIENCY

The most critical concerns In any turbo design are maintaining a proper air to fuel mix in the combustion
chamber, and keeping the combustion temperatures low enough so that the air/fuel mix does not explode
ahead of the flame front (i.e. ignition spark). When this explosion due to too small a fuel versus air mix or too
high a combustion temperature occurs, it is called 'detonation.' Detonation is several hundred times more
violent than normal spark-ignited, controlled combustion the results of this violence can, in instants,
necessitate several thousand dollars worth of overhaul repairs.

Even though the stock car operates at a very modest 5.5 psi of boost pressure, the RX-7 benefits from a
sophisticated electronic feedback sensor capable of listening for detonation and adjusting the ignition
advance/retard level to protect the engine several hundred times every revolution. Cartech found that the
large injectors and the aggressive fuel curve afforded by the stock EFI rendered a correct air/fuel mixture to a
level of about 8 psi without lowering discharge temperatures; with other improvements, it was possible to
run safely to the 11 psi range before additional modification to the fuel system was required.

In measuring the intercooler efficiency the stock car showed a heat rejection of 68% of the incoming
temperature from the turbo above ambient, and a pressure restriction of 1 psi at 5.5 psi of boost. The
pressure restriction is important, since for every psi of boost lost the turbo must produce that much more to
equal the desired boost at the engine; and for every psi of work the Mazda produces, there is a rise in
compressor discharge temperatures before the intercooler of approximately 22 degrees. So, in the case of
the Mazda, the turbo is called on to produce a temperature equivalent to that at 6.5 psi boost in order to
achieve a 5.5 psi pressure in the intake manifold. In addition, a pressure restriction is a percentage of total
boost; so as boost rises, so does the restriction, which further exacerbates the condition.

The reasons for the stock intercooler's lack of efficiency are two-fold. First, the heat rejection surface area is
located at a less-than-optimum position for air flow. In fact, according to a drawing Mazda supplied to Rotary
Rocket (a magazine for Mazda owners), which that magazine published in its December '85 Collectors'
Edition, the designers would have done much better to point the air intake for the intercooler back toward the
windshield. The area they did select receives very poor flow, because it is smack in the middle of a pressure
zone where the air is moving upward. The optimum location would be ahead of the radiator; however, we
can understand Mazda's reluctance to locate it there due to the massive complications that location would
create. The second area of inefficiency, pressure restriction, is even more fundamental; the intercooler lacks
internal flow area. This area, which on the stock intercooler measures right at six square inches, is again
forced by the location and resultant physical size. There are two common ways to increase the flow area on
an air to air unit: one is to increase the size of the complete intercooler; the other is to change the internal
aerodynamic shape of the heat transfer core. A third solution involves redesigning the system as an air to
water unit.

Cartech's solution began with building, installing, and testing three different air to air intercooler units. All
three maintained the stock location, but the smallest more than doubled the total intercooler size. The
pressure losses came down, but the heat transfer was only modestly improved. A serious effort to locate the
intercooler up front was proposed; but it did not look like this could be accomplished without requiring the
consumer to own a body shop so he or she could reconstruct the Mazda's sheet metal. So, in order to
simplify the Installation, a more complicated plan for intercooling was necessary, thus Cartech arrived at this
plan for the world's first twin-pathway air to water intercooling unit.

Cartech has built air to water intercoolers for a number of years, but the typical single path air to water unit
has an efficiency that is generally accepted to be in the mid-70% range. This is good, considerlng that the
given flow area in this method of intercooling is much greater, and the result is typically a much lower
pressure loss (a norm of about . 1 psi). At 70%, however it is not good enough to suggest that the owner of a
new Mazda Turbo II should scrap his or her original unit,

So the Cartech engineers arrived at the idea of producing an intercooler plenum that bolted into the position
of the stock intercooler, but had two completely separate pathways (two separate intercooler core units)
inside. The twin-pathway design would offer the user a combined efficiency of 86%.

Air to water units have an air to air intercooler welded into a plenum. These units flow the water through the
pathway normally carrying the pressurized air, while blowing the pressurized air across what is normally the
frontal or heat rejection surface area. A small electric pump then moves the water from a reservoir to the front
of the car where the heat rejection coolers are located. So the twin pathway design contains two intercoolers
in the plenum, two reservoirs, two pumps and two coolers running one after the other.

TURBO COMPRESSOR EFFICIENCY*

The rotary engine is an interesting application for turbo use, in that it produces an extremely high amount of
exhaust heat and velocity. This is largely due to the two cycle function and the associated firing of the
combustion chamber at closer intervals than are found in four cycle piston engines. Cartech learned to
exploit this high exhaust gas velocity by utilizing very large turbos on applications for the non-turbo Mazda;
they also worked hard to find ways to modify the stock Mazda turbo unit for a similar advantage.

The reason it was felt that a larger turbo would be beneficial is that the stock turbo has a measured efficiency
of only 49% at the peak flow rate of the motor. A 49% rating shows a little better pumping efficiency than that
of a positive displacement supercharger (commonly considered to be peaked out at about 42%); this is
nowhere near the current 70 to 75% which is considered "state of the art" in terms of turbos.

Just what is compressor efficiency? This is the difference between the actual temperature discharged from
the compressor at the maximum flow rate of the engine, and the calculated thermodynamic ideal. This ideal is
arrived at by an accepted thermal formula. So, to put it another way, efficiency = discharge temp. ideal /
discharge temp. actual.

The next question is: why doesn't the factory install a larger turbo - or, more exactly, why do they tolerate
such a low thermodynamic efficiency? To answer this question, we can only guess that Mazda's choice of
turbos was probably the result of a combination of searching for instant low speed response and
compromising efficiency in order to utilize a turbo size already available from their vendor.

In answer to the question of response, Cartech has found that the properly sized, 70+ % efficient turbos
used in their own applications provide a boost response target allows full boost by 3,000 rpm and continues
to pull strongly all tile way to 7,500 rpm. Conversely, the Mazda factory turbo develops full boost by 2,600
rpm, but falls off noticeably by 5,000 rpm. This is due to the compressor dropping off the efficiency curve,
and the associated rise in exhaust manifold back pressure as the turbo requires more shaft speed to
produce the same boost pressure in proportion to the rise in CFM of the engine at higher rpm. Note that as
boost pressure is raised above the factory level of 5.5 psi, the ability to perceive the turbo "falling off" is
increased.

What Cartech offers as their solution is an increase in the compressor size. This solution was selected
instead of a complete turbo change, in order to keep costs down and to maintain the low speed response
offered by the stock twin scroll exhaust manifold ports. The compressor change results in an increase of
compressor efficiency to about 66%, which is enough to afford a reduction in temperature of about two
degrees per psi. Although this reduction seems small, it is enough to justify raising the boost pressure from
9 psi to 10 psi without an increase in the temperature seen at the engine.

Although this modification involves altering the compressor or 'cold' side only, SCCA Solo Board Chairman
Gregg Lee did say in a telephone conversation that he would not consider it legal for Street Prepared
competition. This is a gray area, however, so check your local club rates. At any rate, this step in the upgrade
process is worth an additional 10 hp or so at most, and can be omitted.

Results

Chassis dyno testing of Cartech's compressor upgrade, thier dual path intercooler system and an off-road
exhaust system produced a rear-wheel horsepower of 239. Before modification the stock car tested ran just
151 hp! This would tend to validate Cartech's conservative estimate of 265 bhp.

Of course, all this technical data is great, but it doesn't answer the most important question: How does this
upgrade kit perform in the real world? With the aforementioned modifications in place, our Project RX-7
Turbo is cornpletely docile in traffic. It doesn't idle differently from the factory car. It's not appreciably louder
than the original car. It doesn't start hard. It doesn't run hot . This is truly a remarkable "hop-up" modification.
High tech performance is very different in the eighties, compared to the relatively crude modifications made
twenty years ago.

How much high tech performance are we talking about? There is virtually no turbo lag. The turbo starts to
come on strong at just over 2000 rpm. Four thousand rpm comes quickly; 6000 rpm comes so quickly, you
don't remember to shift until the you feel the rev limiter. In fact, our upgraded Mazda RX-7 Turbo recorded a
best 0 - 60 of 5.83 seconds. This represents a 19.5% increase over our stock time of 7.25 seconds! This is
true world class performance for a street car that will still commute back and forth to work.

On the Auto-X test course, we did even better. We started with an average lap time of 46.454 seconds for the
stock car; we ended tip with an average time of 43.455.

The Cartech approach to improved turbo performance is to lower the thermal stresses, and then raise the
boost back to a point at or below the stock thermal load. The company feels that this approach is the only
logical way in which factory-type durability can be maintained when turning up the boost. Some turbo
upgrades may promise more, but few manufacturers have invested the thought and planning required to
assure engine safety. The Cartech system is not cheap, but look at it this way: the price of this upgrade does
not have to include the cost of regular engine overhauls!