The Toyota Supra established itself as a track-day favorite and a fitting successor to its tuner-king older brother, the Mark IV, but the A90 Supra hasn’t quite been proven in wheel-to-wheel club racing circles. The growing aftermarket for the Supra has helped the new GR become an appealing prospect for many track rats and Time Trial drivers, but its complexity and propensity to run a little hot have made it hard to envision as an endurance racer. Still, some enthusiastic NorCal racers are giving it a go.

Two years ago, Rich Farler of Doteki Auto Solutions began building a few Supra-specific items for the track day community. He had no plans to approach a real race weekend, but a series of conversations with Michael Baxi, his customer, turned business partner, turned media manager, got him thinking that there might be a market for the GR Supra in more respectable parts of grassroots motorsport. It didn’t take long to pick the event from which success would spell undeniable promise for the platform: the NASA 25 Hours of Thunderhill.

Turning the unproven GR Supra into an E0 contender in such a prestigious endurance race would be “a moonshot,” as Farler put it. But as he assessed the situation before him, he felt there was a chance it could work, not that he was wholly convinced from the get-go. The mountain of work ahead of him was daunting, and if the numerous partners could not adhere to Doteki’s strict deadlines, the project might not be worth starting.

It was Cole Viernes, an industry veteran with close ties to the better hired guns on the West Coast, who helped give Farler the confidence needed to push on. “When meeting Rich, I felt his passion in making things better in racing — I wanted to be part of that journey,” he said.

Viernes recommended a few capable drivers who would extract the most from the heavyset platform. With Edgar Lau, Johnny Huang, and Chris Hill as the hired guns, Farler could focus on solving the mechanical issues ahead of him. If they weren’t successful, it wouldn’t be because of a lack of driving talent.

“I’d gotten to understand the Supra’s appeal through time attack, so I was eager to build one for wheel-to-wheel. I knew there was plenty of potential there,” Lau recalled.

Although not as developed nor as light as some of the established E0 contenders, the Supra’s superb powerplant could give it the edge to shine in endurance racing. After all, it’s always easier on tires if all the passing is done on the straightaway. Plus, the recent changes in the rules might suit the Supra.

Other platforms like the E46 are tried and true, but their greatest strength might be limited somewhat this year. Hoosiers have been banned, so the lightweight cornering machines will be deprived of their greatest advantage — at least that’s Doteki’s thinking.


Farler had invested quite a lot into parts development by this point, and volunteering his ‘21 Supra as the build candidate left no one doubting his commitment. Over 25,000 hard miles — many of those on the track, Farler had gotten some sense of the Supra’s shortcomings.

If the heavyset Supra’s flaws could be distilled down into two categories, they would be heat management and oiling. “Increasing mechanical grip has been hard on this car,” Farler began. “Guibos, driveshafts, differentials, and transmissions all struggle under the increased thermal loads that come with more grip. Plus, being more of a luxury grand tourer, it’s not robust enough to handle real racing. A 20-minute session is one thing, but. …”

A trip to a local professional race shop allowed Farler to inspect the GRMN version of the car — the one which becomes the GT4 variant. “When I compared mine to the GRMN, I could spot a few differences from just a quick glance. The size and number of heat exchangers, the diameter of the lines, and the things that are bypassed are a few notable differences.”

Oiling is another matter. Thankfully, this ‘21 model has a better oil pump than earlier cars. Its internals are made of an aluminum alloy compared with the flimsier composite parts inside the prior years’ versions.

However, the same complicated oil cooler/filter combination is still present, which is made of failure-prone composite plastic. It also handles the burden of sharing its cooling potential with the coolant running through the same apparatus.

To solve heating problems at the engine, they had to address the awkwardly designed intake/charge cooler combination. Like several other significant powerplant items, BMW tried to kill as many pairs of birds with a single stone as possible. The B58’s charge cooler is an integral part of the intake manifold, and the OE design causes some bottleneck and reversion issues.

With the aid of Jeremy Harshman at Harshman Racing Engines, they began building the first true runner-style intake plenum. Their design employs a true intake-runner-to-plenum design on the outlet to the cylinder head, which helps to channel reversion, as does the its increased volume.

After adding the plenum, the gearbox started throwing torque limitation codes, forcing Farler to reduce boost.

On the dyno, this plenum gave an otherwise stock B58 an increase in torque by 40 pound-feet across the entire rev range. “Designing this was tough. We realized that the taper, runner, length, and other aspects had to be a very specific size to produce the right sort of airflow. Additionally, being a  BMW product, we were limited by the amount of real estate available,” Farler elaborated.

Crucially, the plenum design, when coupled with a CSF charge cooler, brought post-throttle delta charge temperatures down by more than 150 degrees Fahrenheit. “We also increased the hose diameter to the cooler and removed the bottlenecks in the factory intercooler lines going to our new unit. Increasing flow always helps, considering this is a dual-pass intercooler design,” Farler added.

The B58’s cramped design meant that the addition of their new intake prevented access to the OEM oil filter housing. Farler began building an oil filter relocation kit and mounting it as low as possible to keep unwanted oil from flowing back into the engine.

Increased levels of grip fueled their concerns about oil starvation. To maintain oil pressure after relocating the assembly, they had to change the AN lines stretching from the cooler to the block from -10 to -6 at the inlet side. “Shrinking the AN sizes increases the differential pressure while maintaining adequate oil pump output flow, which was always a delicate balancing act,” said Farler.

That was half of the oiling issues the team dealt with. The addition of a larger Pure800 turbo and increased boost pressure meant they had to address oil temperatures, which, even in stock trim, had always been marginal.

The Thermal Mass Exchanger’s odd construction brings coolant and oil through the same unit. Divorcing the coolant lines from this assembly stabilized the coolant temperature, and the oil cooler was more effective than when it was responsible for two fluids, reducing inlet-to-sump temperature drastically.

Behind the front bumper, Doteki mounted a Vibrant oil cooler to the bumper beam. With a few minimal snips of the front bumper inlets, they were able to design a low pressure area to accelerate air past the oil cooler while channeling the air properly.

Transmission cooling is a must with the factory ZF eight-speed automatic, which is regarded as one of the Supra’s weak links. The Time Trial driver claims it can support the 400 pound-feet of torque without any modification, but the endurance racer must ask: For how long?

By replacing the factory cooler with a CSF cooler twice as large, Farler took the first step in helping the ZF last 25 hours. The next step involved some custom mapping to reduce some of the preload between shifts, which, although extending the gearchange duration slightly, yielded a net benefit in terms of consistency and durability.

Ultimately, it’s a complicated BMW product that doesn’t enjoy the influence of tinkerers. FlexRay, the communication system that links the ECU and the GCU, has not enjoyed the continual removal of unnecessary modules. When the system starts to get warm, the FlexRay forces some premature shifts in the higher gears. Just add this to the list of electronic gremlins that they’ve had to deal with in the last few months.

Second Wave

The power-down characteristics of the Supra have received some criticism from racers and journalists. In an effort to remedy some of these issues, they had to address power delivery and rear suspension behavior. Because the Pure800 turbo shifts the powerband to the right on a graph and softens the torque onset, a linear power delivery doesn’t mean so much if the traction available is insufficient, so the team got thinking of a replacement for the unpredictable e-diff.

Partnering with KMP Drivetrain, Doteki co-developed a fully mechanical LSD — the first to be fitted to a production GR Supra. Building upon the diff’s benefits, Doteki installed a tubular subframe built in conjunction with Verkline. “This setup minimizes toe and camber deflection and promises more compliance at the rear,” Farler stated.

To complement the mechanical improvements, Farler and Jeremy Harshman of Harshman Racing Engines designed their own supports to a giant Zebulon rear wing. Though  it generated the downforce needed to settle the rear at speed, it produced a little too much — more than the composite hatch could support.

“We designed the new mount to disperse the load in the correct area, but the back half of the trunk still flexed more than predicted, causing the composite to crack. This forced us to create a more robust unit that locks the bottom of the lid to a pedestal when closed,” Rich described.

First Test 

The setting for the Supra’s first test would be appropriately grueling for any car expected to run for more than 25 consecutive hours. At Thunderhill East, the ambient crept past 105 degrees Fahrenheit, while the track was nearing 150 degrees Fahrenheit. Hardly ideal conditions for an unproven turbo car.

“It’s not often you see any car pounded lap after lap in that heat without any changes to the power delivery,” said Lau.

Fortunately, all the rear end tweaks were felt. “The demonic snap oversteer is gone,” Lau began.” It puts the power down well, though I’m thinking we should soften the diff’s pre-load to improve turn-in and minimize oversteer at track-out. It also leans a little too much and overwhelms the outside-rear tire in long corners, like Turn 2. Most importantly, we have all been able to adapt to it.”

Additionally, the D2 Competition brakes did not provide the bite and modulation drivers needed for utmost consistency, though some of that was attributed to light seeping at the master cylinder. “I had to be cautious. I couldn’t just step on them. I had a hard time telling where the first bite point is, so I couldn’t modulate them as I would’ve liked to,” Edgar reported.

Toward the end of longer runs, the gearbox had a habit of short-shifting through the higher gears, which forced the drivers to get creative with lifts and shift times to try to keep it in the right gear.

Still, considering it was the first extended test for the car, the performance was confirmation that the platform was stout and the team was well on its way. A consistent session of laps in the 1:57 range on a set of undersized Nankang CR-S with eight heat cycles suggests they’ll be running competitive times with all the bugs sorted out.

Second Promise

For the Sonoma test, perhaps the most notable modification was their hotly anticipated Wilwood eight-piston brakes. Built in conjunction with Wilwood, Doteki actually reverse-engineered Wilwood’s E46 kit to bolt onto the Supra. Wilwood’s engineers inspected the bore sizes and were happy to see the two were a surprisingly good match.

The kit uses Doteki’s own full-floating hats and 7075 aluminum radial mounts, 13-inch rotors at all four corners, and Wilwood’s trusted Aero 6 caliper: six-piston in front and four-piston in the rear. The kit will be available in early 2024.

The other notable changes were to the suspension. First, they increased spring rates at the rear, and, along with revised damping, helped the car style nicely out of corners, reduced roll, and eliminated most of the frustrating snaps that kept the drivers from completely trusting the Supra’s tail end at Thunderhill several weeks prior.

The average lap times were auspicious. Clocking consistent 1:46s with a best lap of 1:45.32. Like the previous test’s time, this 1:45 was not indicative of the car’s potential because a host of minor issues kept them from putting in successive hard laps.

“Nobody’s gone this far into tearing apart a factory Supra before,” Farler noted. “The CANBUS and FlexRay errors kept them scratching their heads in the pits for hours. This, unfortunately, is the reality of operating a production BMW after removing a number of OEM modules — it just isn’t happy.”

Aside from electronic gremlins, Farler and the crew had a relatively pleasant time. No overheating, moderate wear of the Bridgestone RE71s, and fuel burn at seven gallons over two 20-minute stints left them feeling calm and confident.

If driver Johnny Huang had one issue with the car, aside from the lack of front downforce, it was that the brakes weren’t biting as he would like. In fact, the mismatched pair of sprint pads in the rear and endurance pads up front prevented the fronts from ever getting up to temperature, and so Johnny had to treat the left pedal with some caution in the heavy braking zones.

They’ve ticked all the big boxes. All that’s left is adding Zebulon’s front splitter and a set of well-matched Wilwood pads, and some tweaking to resolve those remaining electrical gremlins, which, realistically, may always linger in the background.

What remains for the team prior to the 25? Well, some longer runs to ensure the car can keep its temperatures within reasonable limits. If that’s guaranteed, and it continues sipping fuel at a moderate rate, its turbo power may outweigh its considerable weight and complexity.

Images courtesy of Doteki Auto Solutions and Brett Becker

Join the Discussion