K-series swaps have a particular shape to them: the engine side is well understood and well supported, and almost all of the difficulty is in the car the engine is going into. That is the opposite of how most people budget the job.
The framing that saves time: a K-swap is mostly a chassis wiring project. The engine has a known sensor set and a mature aftermarket. The recipient car has a cluster, a harness, and a list of things it expects to still be talking to — and that list is where the hours go.
The general electrical principles are in our main Haltech wiring and CAN bus guide. This page covers what is specific to K-series swaps.
VTEC and VTC are two separate things
The K-series has both VTEC lift control and VTC intake cam phasing, and they are not the same system. Both need control outputs, and VTC additionally needs a cam position input configured correctly to close the loop.
The common failure here is partial: VTEC gets wired because everyone knows about VTEC, and VTC gets left because it is less discussed. The engine runs. It just does not do what the engine is capable of, and because there is no warning light for "your cam phasing is not configured," the owner never finds out what they are missing.
Configure both. The hardware is on the engine either way.
Drive by wire varies by engine code and year
Some K-series engines are cable throttle, some are drive by wire, and it depends on the specific engine code and year rather than on "K-series" as a category.
This collides badly with the nature of swaps. K engines end up in an enormous variety of chassis, and the recipient car essentially never came with the correct pedal. People improvise, and improvising around DBW is where it stops being a convenience problem.
A Haltech DBW setup depends on a dual-track accelerator pedal and a dual-track throttle position sensor being cross-checked continuously against each other. That plausibility check is the thing that catches a broken wire before the throttle does something you did not ask for. Fit a single-track pedal, or wire only one track of a dual-track pedal because the second looked redundant, and you have removed the safety mechanism while leaving a car that appears to work perfectly.
If you are running DBW, the redundancy has to be genuinely intact rather than nominally present, and that is established by deliberate verification during commissioning — not by the car driving away successfully.
The recipient chassis is most of the job
This is the part worth planning properly, because it is where the time and the frustration actually live.
Whatever car the K went into has its own harness, its own cluster, and its own expectations. If it is an older chassis, that is mostly conditioned signals — a tachometer feed at the right type and level, a speedometer source, temperature and warning lights. Routine work when planned, irritating when improvised.
If it is a newer chassis, it is a communications problem. The cluster, and potentially ABS and other modules, are listening on the vehicle bus for engine data that the factory ECU used to send. They do not gracefully lose a feature when it stops; they fault. Solving that means broadcasting what they expect, which means a Haltech vehicle CAN profile if one exists for that chassis or custom frames via Generic CAN if it does not — and GCAN requires a Nexus.
We have built purpose-made looms for exactly this situation on the chassis we see most, including an RX-8 K20/K24 swap engine harness (TBS-00184) and a Series 1 RX-8 chassis jumper (TBS-00317). The logic generalises to whatever chassis you are working with: a designed loom beats splices that become intermittents two seasons later, and the chassis side is where the design effort pays.
Grounding in someone else's car
Worth stating because swaps make it worse. The ground architecture in a swapped car is usually inherited rather than designed: some of the original chassis grounding, some of whatever the engine harness brought with it, and a few points chosen for convenience during the install.
That is how you end up with sensor returns sharing a path with fans or the starter, which produces readings that move with electrical load — an idle that hunts once things are hot, a throttle position that shifts, a wideband that reads lean under load. All of it looks like a tuning problem and none of it is.
On a swap the grounds have to be designed rather than inherited. Ground architecture is a topology decided as a whole, and a swapped car almost never has one — it has three partial ones layered on top of each other.
K-swap misconceptions
"The engine is the hard part."
The chassis is the hard part. Budget accordingly.
"VTEC is wired, so the cam control is done."
VTC is a separate system needing its own output and a cam position input. Doing one and not the other is the most common half-finished K-swap configuration.
"Any pedal will work with this throttle body."
DBW depends on a matched pedal and intact redundancy. This is a safety system, not a nicety, and it fails silently when compromised.
"I will sort the cluster out at the end."
On a modern recipient chassis the cluster question determines your ECU, because custom CAN frames need a Nexus. That makes it a before-you-buy question.
"The grounds are fine, the car ran before."
It ran with a different engine, a different harness, and a different set of loads. Swaps invalidate the original ground design.
Questions we get on K-swaps
Do I need DBW, or can I use a cable throttle body?
Depends on your engine code. If you are running DBW, run it properly with a matched pedal. If you have the option of a cable setup and it suits the build, that is a legitimate simplification.
Will my cluster work?
On an older chassis, usually, with conditioned signals. On a modern one it depends on the CAN frames and therefore on your ECU choice. Ask before buying.
Do you build K-swap harnesses for chassis other than the RX-8?
Yes. The RX-8 looms exist because that combination comes through often enough to justify a product. We build for other chassis as custom work — send us the details.
Planning a K-swap?
Tuned By Shawn is a Platinum Haltech dealer. We build harnesses in house and tune remotely over Zoom, and on swaps that pairing matters more than usual: the chassis integration decisions and the calibration are the same conversation, and they go badly when they happen months apart between two parties who never speak.

Share:
Haltech Subaru Wiring: AVCS, the Immobilizer Trap and CAN Clusters
Haltech Rotary Wiring: FD Series, Staged Injection, Sequential Twins