In the 1980s, Chrysler took its around-90-to-100 horsepower 2.2 liter four-cylinder engine and pushed it to 142, then 150, and finally 174 horsepower, topping at an impressive 210 pound-feet of torque—all quite good for an engine never designed for such treatment.
How could they beat that in the 1990s, while improving what they already had?
Meet the insane Turbo III
The “T3” was only ever used in three cars, all of them quite impressive. The quickest of the three was the big Dodge Spirit R/T sedan, made in 1991 and 1992; the best-handling was the Dodge Daytona IROC R/T, made in 1992 and 1993; and the rarest may be the Stratus R/T, made only in Mexico.
Generating 224 hp from 2.2 liters, this engine was a thrill to drive, and not just because of the horsepower rating. All those complaints of turbo lag and uneven acceleration really hit home for this powerplant, which shifted from the 2.2’s basic hundred-or-so horsepower to full-on-boost-madness almost instantly. Floor the gas pedal, feel the pause, and then shoot forward: that was life with the Spirit R/T unless you were prepared. Peak horsepower was 224 at 6,000 rpm, with 210 pound-feet of torque at 4,800 rpm; but that was with the original restrictive exhaust. Many owners replaced it with a larger, mandrel-bent exhaust and performance mufflers, yielding greater power.
Development on the A-522 project started in March 1985; while the Turbo IV came out earlier, the Turbo III was in progress first, hence the seemingly backwards naming. The difficult came not from the turbocharger itself, but from the conversion of the 2.2 liter single-cam motor to a dual overhead cam setup, with four valves per cylinder—unique among Chrysler engines, and unusual anywhere—with, just for s-and-giggles, distributor-free ignition. In many ways it was an experimental prototype engine people could actually buy.
Air pumping was far better than other 2.2s, largely due to the cross-flow porting, with the intake side facing front. Individual intake manifold runners, a 52 mm throttle body, divided ports, and large valves (1.4 inch intake, 1.28 inch exhaust) also helped; valves seated with a 45° contact. Boost peaked at 11 psi, well above the old Turbo I’s 7.2 psi, though it may seem tame today.
The cams were made from a common casting and had equal lengths; they were set up alongside the valves to reduce the assembly height. As with Hemi heads, the spark plug was right at the center of the combustion chamber (in fact, the heads were a shallow pent-roof design). The pistons were forged aluminum, with scalloped tops fortunately providing clearance to the valves even if (when) the timing belt snapped or lost its sprockets.
Internal changes to the turbocharger assembly were made to increase airflow, but otherwise it was similar to the original unit. The engine used the same connecting rods as the Turbo II, double weight sorted before assembly for accurate balance. In addition, the air cleaner was changed from round to oval in shape.
One engineer mistakenly used iron plugs in the aluminum engine’s water jacket holes, possibly to save a little money on tooling, which caused many 1991 heads to crack; that was fixed as soon as it was discovered, and there was a recall to replace them with aluminum plugs.
Lotus Engineering was called in for the head design, given their experience. Some blamed them for using overly-strong chrome-vanadium steel valve springs to prevent float at the kind of high revs the computer wouldn’t let the engine reach anyway (redline came at 6,700 rpm). That led to a very high tension on the timing belt, which in turn led to relatively frequent removal of the timing belt teeth by the sprocket. Any mechanic who failed to put the correct pressure on the timing belt would doom it to an extremely short life—think weeks rather than years.
One engineer, though, wrote that the springs were not the issue; but “tow roping,” or the timing belt going into negative tension, was. This in turn was caused by low valvetrain friction (from roller rockers), coupled with the dual cams. When the exhaust valve rocker went over the nose of the cam, there was no friction to slow it down, causing the exhaust cam sprocket to go faster, cutting tension in the belt between the sprockets. The single-cam version didn’t have this problem because exhaust valves were coupled with intake valves. The later Neon dealt with it by having an automatic belt tensioner. Other dual-cam four-cylinders used bucket tappets, which had more friction.
Dyno operator Ed Poplawski told me that they looked at a Maserati head as well; that had buckets operating the valves. The Lotus rocker arms were expensive to make, which may explain why Turbo III production was always limited.
In any case, performance with the Turbo III was quite impressive for the day; the Spirit R/T, a five-passenger sedan with little “performance DNA,” could do 0-60 in under 6 seconds, bone stock, easily beating many V8-powered cars. Getting there was a little tricky, because while the engineers made many changes to deal with the extra power, torque steer was a definite issue; and as noted earlier, power went from “almost none” to “wow” in a moment as the turbo suddenly started spinning up and the cylinders were shoved full of air/fuel mixture.
Year by year changes
The 1990 engines, both turbo and single-injector, had a new single board engine computer (SBEC) replacing the old dual-board computer. Engineers again tried to chase down pesky valve cover gasket oil leaks, endemic to the series of engines, using a new stamped steel cover with a single-piece gasket. They also added balance shafts to its 2.2 liter turbo engines, trying to make them smoother; used sintered metal intake and exhaust guides to improve lubrication; altered the pistons to improve pin support; and upgraded the water pump innards. They also started making “Shelby” VNT engines. Now, the Turbo I was still used in the Lebaron, Daytona, Shadow, and Spirit, with manual transmissions; and in the Lebaron and Daytona automatics.
Other changes for the 1990 cars included a new air cleaner box, using an integrated pressure bypass valve to vent turbocharger pressure on sudden throttle release, a better PCV oil separator, and a cleanable oil vapor element. Sequential fuel injection finally reached the Turbo I, too. All these changes were likely part of the company’s internal cultural change effort, which pushed responsibility downwards and dramatically accelerated the pace of change.
With the release of the Turbo III in 1991, Chrysler improved on the 2.5 Turbo I, adding low-speed boost and changes to fuel and timing control; the net result was another 2 hp and 30 pound-feet of torque, reaching 152 horsepower and 211 lb-ft at the same engine speeds. Improvements to the computer controls caused faster reactions, adding to the benefits.
The Turbo IV did not make it to 1992; the 2.5 turbo was alone, and would remain on its own in its final year in the 1993 cars. Even the 2.2 itself only made it to the 1994 cars. All the turbo four-cylinders were quietly replaced by Mitsubishi V6 engines, mainly missed by people who owned cars with the turbo-fours; with their higher power and lower fuel use, the turbocharged four-cylinders were better for any manual-transmission car and many automatics than the Mitsubishi replacements. What’s more, the four-speed automatics that would have helped them to strut their stuff for slushbox buyers were already in production.
The next Chrysler turbo engines would also be four-cylinders, and they would both arrive in the 21st century—but that’s a story for a later time. So is the launch of the Neon four-cylinder, a 2.0 liter engine rated at 132 horsepower with one cam and 150 horsepower with two—the second number just about matching that of the Turbo I, but with far better economy (albeit far less torque) and costing Chrysler far, far less money.
The 2.2/2.5 “slant four” engine did not die completely when they left American cars; First Auto Works of Changchun made them from 1990 until after 2000. Even in the United States, the 2.5 liter four-cylinder, albeit without a turbocharger, stayed in use in the Dodge Dakota until after the 1996 Dakotas were all built. From there, the Dakota used the AMC 2.5 liter four-cylinder, producing 125 horsepower courtesy of multiple-point injection. The next Chrysler turbo-four would arrive in the 2002 PT Cruiser and Neon SRT4.
See Chrysler’s 1980s turbocharged engines
David Zatz started what was to become the world’s biggest, most comprehensive in 1994 as he pursued a career in organizational research and change. After a chemo-induced three-year break from Web sites, during which he wrote four car books covering Vipers, minivans, and Jeeps, he returned with some friends to create Mopower.com, which is intended to end up as an enduring partnership. Contact him at (973) 925-6058.