HB Cosmo Build-Up
1995

The car was then delivered down to Topline Autopainters where the car was totally re-sprayed inside and out with Mazda Classic red.

After painting was complete the car was taken back to Don Beynon Panelbeaters where it was re-assembled.

While the car was being painted I ordered brand new window, door, bonnet and boot rubbers from Japan. A new front Air Dam was also ordered.

When the car was completely re-assembled it was dropped back to RES in April 1995. Work was then done to replace the engine.

While the panel and paint was being carried out, Ray from Wild Vehicle Electrix and myself were trying to decide what sort of EFI computer to use.

The Tilton Twin Plate Clutch from the USA

A custom one piece driveshaft was manufactured to replace the standard two piece.

We started searching for a replacement Mazda Cosmo LSD but couldn't find one.

The RX7 SIV Gearbox that we were using had to be modified so that the gear lever would line up with Cosmo gearstick hole.

Things seemed to have changed since the first fitting of the engine back in 1994 and we ended up re-manufacturing new engine mounts, intercooler and exhaust pipes. These pipes were then sent away to have a heat resistant HPC coating applied to them.

A custom 4 core radiator was manufactured to provide the cooling. The standard oil cooler was to be used.

One month had passed and we had the engine installed with gearbox and new driveshaft. The radiator, intercooler and oil cooler had also been installed.

It was now time to look at the wiring.

After a lot of research we decided to use a factory 20B EFI Computer. This was going to be tricky as no-one had ever successfully used a factory computer in a re-fit to another vehicle. The reason for this was that the 1991 Mazda Cosmo was virtually a rolling computer with the EFI computer controlling just about all the functions of the car including EFI, Air conditioning, Auto Transmission, Emission Control, Exhaust Operation through special valves, Twin Turbo operation.

If you modify the operating configuration of the computer for example take away anyone or more of the functions that it operates it will not function correctly. (discovered later on)

We decided to have a go anyway. We thought that a factory EFI computer would be the best way of controlling the three rotor engines unusual ignition requirements. We also wanted to make use of the factory crank angle sensor and coil packs and do away with having to use a distributor as would have been necessary if we had used an aftermarket computer.

Well the first thing to do was to find a factory 20B EFI Computer. We searched all over the place and finally found a computer and factory Air Flow Meter. We still required the coil packs and ignitors. After another two weeks of searching in NZ, Australia, USA and Japan we finally found all the parts we required.

The wiring loom on the 20B Engine was stripped down to only leave the wires that were necessary. We wanted to give the engine bay the de-loomed look, with no wires showing.

The original loom had around 80 wires in it and when Ray had finished it had around 10.

The 20B computer was mated with the rest of the cars electrical system. The computer, coil packs and ignitors were all mounted and hooked up.

We were just about ready to start the car. We topped up the motor with all the vital fluids and gave it a go.

The car went through two batteries before we got it started. When we finally got it going, it wouldn't idle properly. This was due to various manifold leaks and leaks around the throttle body. After fixing all of these problems, we were able to get the engine to idle but not rev. It was spluttering and coughing and two more weeks had gone past before the problem was found.

The problem was that the EFI computer was not allowing the secondary injectors to operate which was causing it not to rev.

Another month and another computer went by before the cause of this lack of secondary injector operation was cured.

The problem was that the computer was checking for all of its other components such as the emission systems etc and when it couldn't find them it put itself into "limp" mode which is a safety mode that disables the operation of the secondary injectors so that the car can "limp" home without doing damage to the engine.

Ray, once discovering this, modified the EFI computer to bypass this startup diagnostic function and allow the computer to operate the secondary injectors properly.

This allowed us to finally get the car on the road around August 1995.

This was not the end of all the problems though. The Twin Turbos were not operating properly and we had to disable one of them.

We then had overheating problems. The first to go was the factory Oil cooler in favour of an aftermarket Earls Oil Cooler.

This helped a little but we still had overheating problems. We also had problems with the drivetrain. The driveshaft didn't seem to be balanced correctly and it was re-balanced twice.

The Intercooler was not functioning as we would have liked and we replaced it with a High Flow racing intercooler from Australia.

This helped with keeping inlet temperatures down. I was desperately trying to get the overall operating temperatures down. The problem was the lack of air flow. The front of the Cosmo is very pointed and there is not a lot of surface area for adequate cooling.

I managed to get temperatures down by removing the front bumper. Another contributing factor to the overheating problems were that the turbos were not functioning properly. Because one of them had been disabled it was holding exhaust temperatures within the engine.