Direct "Dear Dave" tech letters to email@example.com. Coleman will share mind-numbing details, earth-shattering revelations, and technical nerdisms in this space each month.
TemptationI have a '92 Miata with a dead engine, so I'm planning on having it swapped for either an SR20DET or an F20C. I've also seen a rotary swap before, but I'm worried about reliability issues. Which engine would be a better choice for my Miata? And what parts and procedures would be required for the swap?Chaiyaphruk ChotikasupaseraneeWaugullewutlekauh,* CA
Instead of answering your question I'm going to, as usual, explain what a bad idea it is to even ask it in the first place. It's no mystery why Miata engine swaps are so tempting. The car's handling is second to few, but the engine is completely lacking in nutsack. However, stuffing in non-Miata engines is notoriously tough.
Other than determining if the engine hole is big enough, the most fundamental issue with any rear-drive engine swap is making sure there is no interference between the oil pan, the crossmember, and the steering rack. There are very few swaps where this doesn't become an issue, and in most cases, this is where the most fundamental mistakes are made.
The Miata has a rear-sump engine, meaning the lowest part of the oil pan is toward the back of the engine. The bulk of the crossmember and the steering rack mounts sit under the shallower front portion of the oil pan. Both the SR20 and F20C are front-sumpers, meaning the deep part of the pan is right where the crossmember and rack want to be.
The smart thing to do here is convert to rear sump, either by making a new pan or figuring out how to mount the original pan backwards. A slightly easier solution is to modify the crossmember and move the steering rack. But since steering and handling were probably the main reasons you wanted to use a Miata chassis in the first place, don't ruin them by moving the steering rack.
The next problem is the powerplant frame. A Nissan Silvia (like most other rear-drivers) has two engine mounts, a transmission mount, a driveshaft carrier bearing mount, and four differential mounts.
The Miata, by contrast has just two engine mounts and two differential mounts. The transmission and differential are rigidly connected with a powerplant frame that both simplifies mounting and makes the entire drivetrain one rigid structure to give that crisp, responsive feel that makes so many people forgive the Miata's lack of power.
Neither the Nissan nor the Honda transmission has a tail housing designed to accommodate the powerplant frame, so you'll have to make a transmission crossmember and figure out how to mount the nose of the differential. This is not an insurmountable challenge and it's not as likely to ruin the car as moving the steering rack, but it will still soften one of the Miata's virtues.
In spite of your fears, Mazda's own rotary engine is better for these two mounting issues. The third-gen RX-7 has a similar rear-sump/front rack layout and it also uses a powerplant frame. The RX-7's frame is mounted on the driver's side of the driveshaft, though, versus the passenger's side for the Miata, so it still isn't easy enough to really be worth doing.
The reason none of these hurdles are worth jumping is that the Miata's engine is so easy to turbocharge. There are several good kits on the market and the engine is fundamentally robust and takes to turbocharging very well. With only 200 to 250bhp, a Miata will feel wildly powerful. Your best bet is to either swap in a less tired engine or do a simple rebuild on yours, then bolt that spinny thing to the exhaust. You won't be sorry.
CA vs. SRI just picked up a 200SX RS13 with a CA18DET. I intend on doing a complete job for serious drifting. I have a budget of about $20,000. I've been following your S13 project and need some advice. Do I keep my motor, or swap for an SR20DET? I'm looking for a reliable 350 to 400 wheel-hp. The best gas we have here in India is 97 octane (UK rating) and the ambient temperature is 68 to 104 degrees F.Mohammed Imran KhanBangalore, India
If you were starting with a US-model S13, I'd recommend the SR20. CA18s went out of production in 1991, so buying a used CA18DET means getting at least a 15-year old engine. Buying an SR20DET for a few more bucks gets you more displacement, a slightly more robust design, an engine with much more aftermarket support, and at least a chance of getting an engine that's only 10 years old.
Since you already have a CA18, though, an SR20 swap is a lot more money for a relatively small benefit. You're better off putting that money into the engine you have. Because the tuning world so completely embraced the SR20, there is far less collective knowledge about the CA18DET's capabilities, but general wisdom is that a healthy stock motor can deal with something around 300 to 350 wheel-hp before things go kablooey.
You might think you need more for drifting, but I can point to plenty of AE86s that say otherwise. I say stick with the CA, go for 300 wheel-hp, and spend the rest of the cash on suspension, crash repair, and tires. Drifting is a competition of driving skill, not horsepower. Skill takes practice, and practice costs more money than you're probably going to budget for. Build the car for $10,000 and spend the other $10,000 drifting the crap out of it until you're actually good.
Square Peg, Round HoleI was wondering if you would be able to tell me the way to fit a 3.4-liter Ford Taurus SHO V8 into a 1990 CRX. I had the thought a while ago and would love to hear the redneck rumble from a teenie-weenie CRX. I just got the CRX, so before I make an ill-fated attempt at shoehorning this engine in, I would like to know the headaches involved. I know everyone under the sun is putting B-series and K-series in these things, but I would like to be a bit different.Eric BernardManchester, NH
What the hell are you... oh, never mind.
The SHO V8 only came with an automatic. In my world, that's a non-starter, but-in my world-that engine is too big and heavy for a CRX no matter what transmission it has. I don't know how things work in your world, but just in case you were thinking a CRX transmission would bolt to an SHO engine, you should know that you'll be stuck with the mushbox. If that's not a problem, you should also know that the engine probably won't fit, so don't take the following as any kind of encouragement.
Center the drivetrain side-to-side so the inner CV joints on the left and right side are equal distances from the wheels. Front-to-back, you want to mount the drivetrain so the inner CVs are in line with the front-wheel centerline. Vertically, you want the inner CVs as close as possible to the height of the center of the front wheels when the car is at its target ride height. If you manage to hit all these mounting targets, torque steer will still rip the wheel from your hands any time you hit the gas, but at least you'll be secure in the knowledge that you did what you could to minimize torque steer caused by axle angularity.
After buying custom axles with Ford inners and Honda outers, you'll need to fabricate all the mounts from scratch. This will be even harder than normal as Ford and Honda mount their engines in fundamentally different ways.
Honda uses mounts at the extreme ends of the drivetrain (front of the engine and the back of the transmission) to support the weight of the engine, then uses a single mount in the rear to resist the engine's torque. The SHO, as far as I can tell without looking too hard at a Taurus, seems to mount at the corners of the drivetrain with two down low near the front of the engine, attached to the subframe. To mount the engine Honda-style, the SHO would need a front timing cover strong enough to support the whole engine. Since it mounts to the sides of the block, it won't have that.
If the prospect of living life in D hasn't deterred you and the fact that the engine is bigger than the car isn't a problem, I'm hoping that the lack of any practical way to bolt the engine to the car will finally convince you that you should just drive a Taurus instead.
*Not his real city. In future, please tell us where you're from, otherwise we'll just pick a place.
Miata turbo kits are plentiful and downright cheap in comparison to engine swaps, or you can build one yourself like the one shown above. Our colleague, Ryan McKay, put this GT-series turbo kit together for about $1300 (withoutthe ECU).
The venerable CA18DET motor is old, but reasonably strong and worth keeping if your S13 already came with it.