Back in February of 2020, our longtime Lotus customer reached out about potentially building an Elise from the ground up. The idea was to build a car to his exact specs, a reimagined Lotus Elise. The plan was to buy a brand new tub chassis from Lotus and hand-pick, or fabricate, custom parts for one goal: his perfect “streetable” track day car. Most of the drivetrain and miscellaneous bits would be taken from his current 2011 Elise.
Our customer’s Lotus is actually a very special one because it is number 15 of 15 of the final edition Elises, meaning it was the very last Elise to enter America as a road legal car. If you’re a Lotus purist, you may not want to continue reading.
After some long conversations about how we would go about this build, he decided to pull the trigger. The ordering process was relatively simple. All we had to do was show proof of ownership, wait a few months, then a huge crate was delivered to the shop. Since this is a “replacement part,” the tub came engraved with the same VIN as his original Elise. We would have to cut the VIN out of his original tub and send it back to Lotus.
Honda K20
If you’ve been following our shop, you know that we’ve put a Honda K-series motor into our shop car, including a handful of customer cars. Naturally, we suggested the idea of putting a K motor into this build. After doing his own research, he was sold on the idea under two conditions:
It has to be street legal
It has to have A/C
Since our prior K-swaps were race cars and/or not subject to emissions testing, these were two areas that we’ve never paid attention to. Yes, there are street legal K-swapped Lotus cars out there, but you probably won’t find one in California with a BAR Referee label. California is the only state permitted to issue emissions standards and it is not easy to modify your car here, let alone swap a motor.
When you build a race car, you follow a rulebook to make sure it’s legal for its class. This build would be similar in that aspect, except that we’re making it legal for the street.
The Donor
The Engine Change section of the Smog reference guide states that, “The installed engine must be of the same model year or newer than the model year of the recipient vehicle.” With that, our engine of choice was the K20Z3 out of a 2011 Civic Si, and being a “later” model, this would complicate things exponentially. We both agreed that having a donor car would ensure that we would have everything needed so that the motor would be in a stock certified configuration, and support all of the certified OBD II functionality. After months of searching for salvaged Civics, our customer decided to buy a clean title 2011 Civic Si with relatively low mileage to avoid any “unknowns.” Before actually making the purchase, he brought it in for a pre-purchase inspection.
What we cared most about was the engine’s compression and transmission. Both checked out, so he went forward with the purchase and had the car delivered to our shop.
Before disabling the car, I wanted to do a few things while we had a smog ready car.
Get all emissions monitors “ready”
The engine ECU will run a series of self-tests to determine if the emissions system control system are working properly. We can easily reset these by disconnecting the battery. We will have to drive the car under specific conditions to get the monitors “ready.” The purpose of this test is to confirm that we are able to get all monitors “ready,” that way, if we are not able to do this in the Lotus, we will know that it is unlikely a hardware issue.
It only took about 15 minutes of driving to get all except the EVAP ready. Typically, the EVAP is not required to be complete for a smog test, but since the car is going through the BAR inspection, it is required. According to the factory repair manual, the EVAP test can be done with the factory scan tool. I was able to source a Honda specific diagnostic tool that did the job.
Testing the Lotus gas pedal
Both the Civic and Lotus are drive-by-wire, so I had an idea to use the Lotus accelerator pedal (pedal position sensor), instead of trying to retrofit the Civic pedal position sensor into the Lotus. Both cars use the same style 5v reference sensor, so the only way to find out was to give it a try.
Left: Lotus pedal
Right: Civic pedal
I was able to 3D print an adapter bracket and then make a jumper harness to plug the Lotus pedal into the Civic. The car started right up and I was able to drive it without any fault codes.
bypass the immobilizer
One of the complications with using the 2011 Honda ECU is that it has a built in immobilizer, and it isn’t your average immobilizer that you can bypass by jumping a few wires. When a matching key is put into the ignition switch, the key sends a coded signal to the immobilizer unit, which then sends a coded signal through a relay box (which also has an immobilizer unit built into it), and finally to the ECU. If the signal is correct, then the ECU will trigger the fuel pump relay. Bypassing the fuel pump relay was easy. What the factory manual didn’t tell us is that the ECU also cuts off the signal to each ignition coil. This isn’t something that we are able to bypass without tampering with the ECU, so we decided that the entire Honda immobilizer system would have to go into the Lotus.
We started off by removing as much of the wiring from the relay box unit as possible while allowing the car to start and run without any fault codes. Out of about 210 wires, we narrowed it down to to 8 wires. The rest of the immobilizer components would also have to be tranferred over and stay wired as-is.
CAN bus Modules
We found that unplugging any of the modules on the CAN bus, such as the instrument cluster, EPS, or VSA modules would set a fault code for CAN malfuction. Since we don’t plan on using any of these in the Lotus, we are either going to have to transfer over all of these modules, or figure out a way to trick the ECU into thinking that they are still plugged in.
After countless hours of learning how to retrieve the CAN data, comparing messages in spreadsheets, and unsuccesfull attempts at transmitting the correct messages into the bus, I learned that Hondata has a CARB version of their FlashPro, which allows us to disable all of these modules, except the instrument cluster. This meant that they had already figured out what I was trying to do all along. We will just have to keep the Civic instrument cluster plugged in and hide it under the dash.
Having this “easy way out” was a feeling of relief and defeat at the same time, but not all of the invested hours were lost. The things I learned the with help of others will be useful for future CAN bus projects.
Check Engine Light
An AIM MXP Strada dash will be used in place of the Civic instrument cluster. Unfortunately, a check engine light is not an available channel in AIM’s configuration for this ECU, and a functioning CEL is required to pass the inspection. Since the CEL is triggered by the ECU over CAN bus, and not an analog input, we decided to use the bulb input from the stock cluster to trigger a CEL on the MXP. All we had to do was desolder the bulb and solder a lead between the bulb input to an unused pin on the cluster connector. While we were at it, we also did the immobilizer indicator light.
Tub Prep
We went back and forth on whether we were going to install any sound deadening in this car. It’s a compromise between weight and road/engine noise. We decided that we were going to put heat reflective tape on the firewall on the engine side to keep heat out of the cabin. It looks a little bit cleaner than the factory carpet-like material, and if sound deadening is desired in the future, it wouldn’t be difficult to apply on the interior side.
We will be transferring the single side mount oil cooler over from his original car, so now was a good time to remove the factory oil cooler lines.
A few months after receiving the tub, the owner dropped off his Elise to start transferring parts over.
Comments from the owner
This project is my attempt at playing a bit of automotive development, if you will. Basically, the theme of the build is what would happen if the US Elise was developed further. Something like an Elise GT3, where the car is still streetable, but much more hardcore. Therefore, I started planning for this build with some rough core targets and also a general core theme for how the vehicle should feel. My goal is to have something that feels factory built, instead of a car that has been modified by its owner, which is what the current build of my Elise is starting to feel like. We'll be examining a lot of subsystems as the build goes on.
The main theme will be lightweight and reliability. In terms of lightness, more is obviously better, but it needs to have a presentable interior, actual glass windows, and won't have me rethinking my life choices when I'm trying to drive this car in 100 degree weather. Therefore, I decided on a curb weight (full fuel) target of 1,900 pounds with a target weight distribution of 42/58 (though 40/60 is more realistic).
In terms of reliability, it's admittedly weird that I'm engine swapping. Engine swap and reliability generally go together like lots of alcohol and good decisions. However, I believe that the Elige platform is actually quite well suited to engine swapping due to how open the rear of the vehicle actually is and how little in the way of electronics and other integration is needed, compared to other contemporary vehicles. In the end, what I've heard about Trackspec's K swaps gave me confidence that it'll be reliable. We'll see.
The truth is, having grown up an F1 fan during its high revving naturally aspirated era, I always wanted a naturally aspirated Lotus. However, all the Final Editions were supercharged and, well, I was just drawn to this specific car. We can debate about Theseus' ship until the cows come home, but nothing about Lotus ownership is logical. In the end, I think it's appropriate that one of the last US Elises will end up serving as a further development of the US Elise platform.
I've considered just converting the car to naturally aspirated for many years, but I'm not a big fan of the power curve of the stock 2ZZ and something always felt wrong about "down speccing" the Elise. The lack of aftermarket parts for NA 2ZZ builds meant that, in the end, I just couldn't build the naturally aspirated Elise that I wanted with the 2ZZ, without significant sacrifices. The K20 allows me to build the naturally aspirated Elise that I want.