SU Carbs on a Miata 1.6L
Yes, I put carbs on my Miata as a lark back in '92 or so. I found a pair of rebuildable HS4 carbs in my pile of Brit-car parts and adapted them to my daily driver. These carbs came from a 1.8L MGB engine and were "close" to what would fit the needs of the 1.6L Miata, so that was enough to justify the exercise. I had Graham at Jim Downing's shop (where we sold the Sebring Superchargers out of) make up a beautiful TIG-welded aluminum intake manifold, complete with cross-over pipe/balance tube which is critical for idle performance. Throttle linkage was easy, and I put a manual choke cable on it, mounting the knob where the cigarette lighter hole was in the dash (my favorite part).
Some notes from the installation:
1. The 1.6L Miata engine pulls around 177 CFM at full song (naturally aspirated) at 7,250 rpm and according to the SU factory specs, the HS4 would be just fine
2. Putting carbs on a Miata eliminates the rev limiter (fuel cut at 7,250 RPM), so it ran to 8,500 quite nicely (albeit at even higher CFM rates, so the carbs began to wheeze a bit up there). The loss of low-end torque through the loss of the factory tuned intake system was partially offset by the nature of the SU carbs (see comment #4 below for why).
3. I had to trick the ECU into some sort of Throttle Position Sensor (TPS) signal, so I found a plunger TPS (simple linear potentiometer - this one from a 1990 Mazda B2000 truck at MazMart) and mounted it to the side of one of the carbs, actuating it with a small aluminum cam profile I made that fit on a nearby throttle shaft stub. As the throttles opened, the cam depressed the TPS's plunger. Using the wire carrying the 5.0VDC reference signal that goes to the stock TPS, we could feed a linear 0 to 5 VDC signal back to the ECU so it would advance the ignition timing correctly. This is the only trick you'd need to accomplish if you were to do this with a different engine.
4. Throttle response was greatly improved, since the SUs are "constant depression" units which keep the carb venturis' air velocities very high at part-throttle operation and prevent the "open door" issue a butterfly throttle's opening alone will do to an intake tract's flow velocity (throttle blade-to-intake valve "air colum velocity" drops terribly immediately following a rapid throttle opening, thus killing any inertial benefit of a high-velocity air stream with regards to cylinder filling).
5. Power felt about the same (never dyno'd it other than my calibrated tush) and the sound was fantastic.
There, much more than you probably wanted to know.
If you're wanting to do this on a Toyota 20R, you might want to go with the HS6 carbs to be closer to the right CFM flowrate. Tuning is easy - there are reams of information online on jetting SUs.
A photo from my book:
Lo-res detail of the TPS: 
Please let me know if you have any comments or questions on what you're trying to accomplish.
Norman G
