← Gallery
Community work recordPorsche 964

Porsche 964 3.8L Varioram engine build

Moriartyhlms documents a two-year 3.8L engine rebuild and Varioram conversion, from case machining and a head-stud installation error to DME and harness troubleshooting. Stalling and fitment issues remained after tuning.

Edited by Future Classic ·

Select a photo to view its source post. The source may require sign-in.

An engine build spanning more than two years

Moriartyhlms spent more than two years rebuilding and upgrading his 964 engine. The combination includes Mahle 3.8L machine-in pistons and cylinders, Carrillo rods, 993 Supersport cams, a GT3 oil pump, Varioram intake, 993 fuel injection and a custom-tuned 993 DME. He had just returned from dyno tuning with Steve Wong and described it as running mostly correctly, while also documenting several unresolved problems.

The owner’s 964 after the build.
The owner’s 964 after the build.
The rebuilt engine on a stand.
The rebuilt engine on a stand.
The at-the-wheels dyno chart supplied by the author.
The at-the-wheels dyno chart supplied by the author.

Case machining and a head-stud installation error

For the machine-in cylinders, the author says the case needed machining and that a shop cut O-ring grooves into the case spigots, staggering the grooves so the walls would not become too thin. The rings were 993 RS parts. The author also records a serious assembly error: measuring ARP studs to the domed end rather than the last thread left too little engagement. About 75 miles after the rebuild, a loud pop, noise and power loss led to a tow; the right-side top-row head nuts had pulled off. The author says he reset the studs to the correct depth and installed new nuts.

The case-spigot machining shown with the 3.8L discussion.
The case-spigot machining shown with the 3.8L discussion.
Cylinders and exposed head studs.
Cylinders and exposed head studs.
The engine with a cam cover removed.
The engine with a cam cover removed.
A stud shown in the installation-depth discussion.
A stud shown in the installation-depth discussion.
A close-up inside the cam housing.
A close-up inside the cam housing.
The cylinder-head assembly shown during inspection.
The cylinder-head assembly shown during inspection.
The car being recovered on a flatbed.
The car being recovered on a flatbed.

Varioram electronics and harness work

The author’s refurbished 993 DME would run the engine but, in his account, did not operate Varioram correctly. Troubleshooting took eight months, roughly seven engine drops and multiple sensor replacements. A salvaged 993.618.123.03 worked with a custom chip. The author suspects the refurbisher had supplied a different or incorrectly represented unit, but does not prove that diagnosis. He also describes adapting an 88-pin harness to 55 pins by comparing wiring diagrams, calling the work tedious and recommending a new harness instead. The separate 993 engine harness needed changes to the oil-pressure and transmission speed-sensor wires.

The refurbished Motronic unit with handwritten identification.
The refurbished Motronic unit with handwritten identification.
The replacement Motronic unit shown by the author.
The replacement Motronic unit shown by the author.
The opened DME harness connector and wiring.
The opened DME harness connector and wiring.
Harness work alongside the removed engine.
Harness work alongside the removed engine.

Wiring, ignition and compression-ratio problems during tuning

The author reports poor early running alongside DME and wiring problems. Dyno tuning took two days; he describes a distributor installed one or two teeth out and detonation attributed to high-compression pistons, having not calculated the compression ratio during the build.

A/C clearance remained imperfect with custom lines

The original 964 A/C lines did not fit around Varioram, while 993 lines did not fit the 964 engine bay. Custom lines and a compressor plate were made, but the fitting/plate still contacted the oil filter. The author was having an oil-filter-console spacer fabricated; another supplier’s lines had interfered with the spoiler mechanism.

Continuing stalling and exhaust fitment problems

Stalling persisted with the lightweight flywheel and 993 DME. The author was experimenting with the idle-control valve and considering a mid-weight flywheel. He liked the B&B exhaust under some conditions and believed it helped power, but also reported drone and poor fit. Its bracket needed modification, the mufflers sat unevenly, and he trimmed the bumper after heat damage.

Based on the source author’s work record.

Discussion & reactions

Open discussion