RX-7 (Second generation)

An RX-7 Drawn for America The first RX-7 had been a light, cheap, live-axle coupe that sold on novelty. Its replacement, launched in Japan in October 1985 as the Savanna RX-7 and coded FC3S, was deliberately heavier and more conventional, aimed at the market that had bought most of the first ones. Design work ran through Mazda's North American studio in California, the Porsche 944 was the openly stated benchmark, and the finished car carried a quoted drag coefficient of 0.29 at roughly 250 lb more than the car it replaced. Commercially it worked at once. The FC took Motor Trend's Import Car of the Year for 1986, the Turbo II made Car and Driver's Ten Best list for 1987, and the RX-7 took more than a quarter of Mazda's United States sales in its first full year, at 56,203 cars.

The 13B in Two Tempers The 13B twin-rotor Wankel measures 654 cc per chamber, 1,308 cc nominal, and appeared in two states of tune that have aged very differently. The naturally aspirated 13B-DEI used six-port induction and 9.4:1 compression for 146 hp, rising to 160 hp with the Series 5 revision for 1989, which also brought thicker rotor housings and a higher redline. The 13B-T dropped compression to 9.0:1 and added a single turbocharger with a twin-scroll housing, a split intake manifold and an air-to-air intercooler fed by a bonnet duct: 182 hp and 183 lb-ft as the Series 4 Turbo II, 200 hp and 195 lb-ft as the Series 5. Japanese cars were quoted in PS - 185, then 205, and 215 in the Infini series. Precisely when the twin-scroll arrangement arrived is not agreed: encyclopedia and magazine accounts attach it to the Turbo II from launch, while Grassroots Motorsports credits it to the 1989-and-later cars alone.

Toe Control at the Rear Axle The engineering the FC is remembered for is not the turbocharger but what Mazda did behind it. The Dynamic Tracking Suspension System replaced the old live axle with independent semi-trailing arms whose hub carriers sat on compliant bushings, allowing the rear wheels to toe out slightly under light lateral load and to toe in once cornering forces passed roughly 0.5 g. It is passive four-wheel steering achieved with rubber rather than actuators, with optional Auto Adjusting Suspension varying damping alongside it. Whether it was a good idea depends on the tyre. On period-width rubber and fresh bushings it does what was intended; on worn bushings and modern tyre sections it produces exactly the transitional unpredictability it was meant to cure, which is why race preparers fit toe-control bushings that disable the rear steer outright.

Convertible, Anniversary and the Limited Cars The FC3C convertible arrived in Japan in August 1987 and the United States for 1988, the only factory drop-top in the RX-7's three generations, with a rigid removable panel over the occupants, a folding rear section and an integral windblocker claimed as a first. Mazda tied its launch to twenty years of rotary production. Above it sat a run of commemorative cars: 1,500 10th Anniversary Turbo IIs for 1988 in monochromatic Crystal White; the American GTU and GTUs of 1989-90, lightened naturally aspirated cars carrying the Turbo's four-piston front brakes and a viscous limited-slip differential to mark the model's IMSA record; four annual Infini series in Japan of roughly 600 cars each; and a 1991 Winning Limited marking Mazda's Le Mans victory with the 787B. The R1 and R2 designations frequently attached to the RX-7 name are not FC cars at all - they belong to the third-generation FD.

Falling Sales, and the Generation in the Middle On track the RX-7 was doing better than in showrooms: Mazda counts 100 IMSA wins across twelve years, and the RX-7 GTO took four wins and both titles in 1991. Sales told the opposite story. United States volume fell from 56,203 in 1986 to 38,345, then 27,814, then 16,249 by 1989, and 6,006 in 1992. By then the FD was arriving, and the attention went with it.

Apex and side seal wear is the defining risk, and it is a wear pattern rather than a fault - low compression from age or neglect means a rebuild. Coolant seal failure is the more damaging cousin: it lets the engine overheat, and overheating warps rotor housings, turning a seal job into a housing job. Buyer's-guide costings put an FC rebuild at roughly $4,000-$9,000 for seal work and $4,500-$10,000 where housings are involved. Supplier pricing in August 2026 sits in the same territory: Atkins Rotary lists a rebuilt 13B at $4,099-$4,699 on used rotor housings and $5,299-$6,699 on new, with the turbocharged 13B at $4,999-$5,199 and $6,499-$6,999 respectively. Beyond the engine: flooding on cold start; overheating in traffic from a clogged radiator, missing shroud or failed fan control; turbo seal wear showing as smoke and low boost; detonation under boost from lean running or heat soak; oil leaks from the front cover, sump and rear main; second-gear synchro wear; and electrical faults traced to poor earths. The eccentric shaft bypass valve that feeds the seals only operates once warm and is worth replacing as routine maintenance. Regulator records for 1986 cars cover front brake caliper corrosion where road salt restricts pad movement, and in Canada a misrouted evaporative purge hose.

Start with a warm compression test read per rotor face rather than as one number. Specialist guidance puts a healthy 13B above 100 psi per face with the faces even; below about 85 psi indicates seal wear and a rebuild in prospect. Hot-start behaviour matters as much as the gauge - a long crank on a hot engine points to weak seals or flooding. Then work through the cooling system, because a rotary's tolerance for overheating is close to nil and one event can warp a rotor housing: radiator, fan shroud and clutch, thermostat, coolant loss. Establish which series the car is, since Series 5 cars from the 1989 model year carry the higher outputs, thicker housings and higher redline and price accordingly. On Turbo II cars check turbo shaft play and smoke under boost, and expect to replace the vacuum hose network, which is extensive and brittle. Rust is structural here: sills, arches, strut towers, hatch channel and door jambs, and the rear hatch drain plugs that clog and cause it. Check the front subframe for accident damage, as it resists straightening. Underneath, feel for play in the rear lateral rods and assess the DTSS bushings, worn on most survivors. Second-gear grind is common. Body parts are the quiet expense: nose cones have been quoted around $800. Prefer documented engine work over claimed engine work.

As of August 2026 classic.com separates the FC into submarkets that sit a long way apart, and the spread is the story. The Turbo II carries a Market Benchmark of $24,093 against an average sale of $24,729, on a rising trend, with a high of $39,250 recorded on 29 October 2025 for a 1988 10th Anniversary car and a low of $843 in June 2022. The naturally aspirated base coupe benchmarks at $11,460 with an average of $11,623 and is trending down, its recorded results running from $5,000 for a 1986 in May 2025 to $18,257 for a modified 1987 in October 2025. The convertible sits lowest, at a $10,663 benchmark and a $10,931 average on a rising trend, between $3,750 for a 1988 project in August 2024 and $19,999 for a 1991 in March 2026. Across the whole generation classic.com's average sale is $15,681. Auction rooms have been less forgiving: a 45,000-mile automatic 1991 convertible offered at Mecum Kissimmee in January 2024 failed to sell against a high bid of $12,000. For orientation against the generation that followed, classic.com's third-generation FD average sale is $48,593 as of August 2026, with a 1993 R1 recording $127,000 in August 2026 - which places a strong FC Turbo II somewhere near where the FD market begins.