CT-26 Twin Entry Turbocharger

The stock CT-26 twin entry turbocharger in the 3S-GTE is probably the most misunderstood component in the MR2. It's probably due to the fact that turbochargers, while apparently easy to understand on the surface, don't behave in intuitive ways. It is not uncommon to hear that the CT-26 puts out nothing but hot air above 15 psi. Many have upgraded their CT-26 with a 50-, 56- or even 60-trim TO4E compressor wheel only to find (if they even bother to check on a dyno at all) they are getting very little to no extra power beyond what a stock CT-26 gives them and certainly nothing like the 400-500 rwhp these compressor wheels are capable of producing. Mystified? Let's recover some basics and get to the real truth.

One of the first things to keep in mind is there there are actually several CT-26 turbochargers. The larger 3.0 liter Toyota Supra used the CT-26 with a larger single entry turbine housing. Because the higher displacement Supra engine can flow more exhaust gasses to get the compressor spun up than the 2.0 liter 3S-GTE, the single entry housing is larger and less restrictive than its twin entry sibling. This makes the two turbochargers completely different beasts for reasons I will cover later. There is also a ceramic turbine wheel variant of the twin entry turbo which uses a slightly smaller but lighter and less restrictive turbine wheel.

Toyota must have the compressor map for the stock CT-26 compressor, but they haven't distributed it as far as I know. However, a quick glance at the CT-26 tells us that it has a standard looking compressor wheel and housing such as is found on most turbochargers. Given that, we can bet big money that its compressor map is similar to the hundreds of other compressors out there such as the one for the TO4E 40-trim compressor shown below. We can also safely assume that the Toyota engineers who matched the CT-26 to the 3S-GTE knew what they were doing. Undoubtedly, they calculated how much air the engine would ingest at various RPMs and boost points and selected a compressor whose efficiency island was well within these pressure and flow rate points.

OK, but what about those really bad efficiency areas along the edges of the compressor map? Won't the compressor go there when we turn up the boost? The answer is: no. Look at an average compressor map and compute a line through it tracing the compressor's operating point as you raise or lower the boost. What does that line look like? When you raise boost, two things happen. First, you move straight up the vertical axis of the compressor map because raising the boost raises the pressure ratio. Second, you move towards the right because a higher pressure produces a higher density ratio in the intake manifold air which forces more air into each cylinder increasing the air flow through the engine and compressor. If that move towards the right didn't happen, a lot of compressors would hit their surge limit as the boost increased since surge lines also move towards the right as you go up. Notice as you trace this line that it tends to stay in the efficiency island and, if it pops out at the top, moves to an area whose efficiency is always within 5% of the middle island. Aha! Here is where the compressor turns into a blast furnace and your performance goes south. Not so. Let's work the math on this.

Assume that the CT-26's efficiency island is in the 70% range. It is probably better than this since most modern compressors are, but it is OK to underestimate for the purposes of this exercise. All compressors produce hotter air as the pressure ratio goes up. This is not an artifact of efficiency or inefficiency. The ideal gas law states that the temperature of any volume of gas will increase as it is compressed into a tighter space. The temperature at the outlet of any compressor is determined by the following formula derived straight from the ideal gas law:

Tout (in F) = ((Tin (in F) + 460) * (Pressure Ratio^0.283)) - 460

This formula tells us the temperature increase assuming an ideal (100% efficient) compressor. In real life, all compressors operate inefficiently and thus heat up the air more than the minimum that the ideal gas law requires. The real outlet temperature after factoring in compressor efficiency is:

delta T actual = delta T ideal / efficiency

Let's calculate the outlet temperature of a CT-26 boosting to 17psi at sea level on an 85F degree day:

Tout = ((85 + 460) * (((14.7+17)/14.7)^0.283)) - 460 = 217.4

Thus, the ideal change in temperature (delta T ideal) is 217.4 - 85 or 132.4, so the actual change in temperature is:

delta T actual = 132.4 / 0.70 = 189.1

So the actual outlet temperature is 85 + 189.1, or 274.1F. Fortunately, this air doesn't enter the intake manifold without first passing through the intercooler. The temperature drop across the intercooler is determine by the following equation:

T IC drop = (T IC in - T ambient) * IC efficiency

For the stock MR2 intercooler, the efficiency is around 55%. This decreases the turbo outlet temperature by:

T IC drop = (274.1 - 85) * 0.55 = 104

So the manifold temperature will be 274.1 - 104 or 170.1F.

Now we use the same equations to calculate the manifold temperature assuming that the compressor efficiency at 17psi has moved off the efficiency island and dropped to a worst case imaginable figure of 65%. The ideal gas law temperature increase remains the same as above, but the actual temperature change increases due to lower efficiency:

delta T actual = 132.4 / 0.65 = 203.7

This makes the new outlet temperature hotter:

Turbo outlet temp = 203.7 + 85 = 288.7

And also the intercooler will have to deal with a greater temperature:

T IC drop = (288.7 - 85) * 0.55 = 112

Manifold temperature = 288.7 - 112 = 176.7

So, even if the efficiency of the compressor drops by as much as 5% as we raise the boost, the difference in the manifold temperature increases by less than 7 degrees. With a slightly more efficient IC such as a 65% effective sidemount, manifold temperature increases only one degree per one percent decrease in compressor efficiency.

Well, that's the math. What about real life? The following is a TEC³ datalog of my CT-26 and GReddy sidemount IC combo doing a 4th gear freeway pull up to ~17.2psi (223.5 kiloPascals) on a 75F degree day. What we see is intake manifold temperatures actually dropping during the pull because the IC doesn't cool the air much when it flows at low volume allowing manifold temperature to rise near the level of the engine bay (this was the fourth or fifth pull in a ten minute interval, so the engine and the engine bay were quite warm). When the boost starts, high air volumes from the turbo push their way forcefully through the intercooler shedding heat before entering the manifold and cooling it as well. The bottom line is that a CT-26 with appropriate intercooling produces very safe, usable boost above 15psi.

The CT-26 compressor has been found to produce 22psi of boost pressure at 3.5K RPMs if the actuator hose is blocked and the turbo allowed to build maximum boost. The CT-26 will not continue to make 22psi throughout the entire RPM range. In fact, by 4K RPMs, the boost pressure begins to drop quickly to reach 12-13psi at 6K RPMs. Is the compressor incapable of flowing enough air to feed the engine as RPMs rise? That sounds like an plausible theory and many have upgraded their CT-26s with larger compressors such as the TO4E 50-, 57- and even 60-trim wheels capable of flowing enough air to produce over 400 rwhp. Unfortunately, none of these upgrades have produced more than several peak horsepower more than what is possible with the stock compressor and, in fact, these compressors are so heavy and take so much more time to spin up that they never match the stock compressor's peak torque. The result is a less drivable car making about the same power for the price of a rebuild. Obviously, not the best investment.

The real problem with the CT-26 at higher boost levels is that the small turbine housing with twin scrolls to maximize the exhaust energy concentrated on the turbine wheel at low RPMs is actually "chocking" the engine. Once the exhaust system has been opened as far as it will go, the only restriction preventing exhaust from escaping the cylinders is the exhaust valve, the exhaust manifold and the turbine. If two boost gauges are attached to the 3S-GTE, one measuring intake manifold pressure and the other exhaust manifold pressure, what is seen is that at boost levels of around 10psi (what the stock MR2 Turbo was designed to operate at) the boost pressure in the intake manifold and the exhaust manifold are practically equal once the CT-26 spools up. This is a very desirable situation and speaks highly of Toyota's ability to match a turbo to their intended application in the MR2. At 15psi of intake manifold pressure, however, the exhaust manifold pressure climbs to over 20psi. This is not desirable. What is happening is that the exhaust gases are being held back by the turbine and, once that starts to happen, the turbine is unable to extract any additional power from the exhaust to push the compressor harder. The only way to produce more power in this situation is to allow the exhaust gases to move more freely across the turbine. There are several ways to do this: 1) get a larger turbine housing that permits more exhaust to flow across the turbine, 2) get a smaller turbine wheel that allows more exhaust to pass across it or 3) "clip" the existing turbine wheel to allow more exhaust to pass through it.

There are very few options (at least in the USA) for changing the turbine housing or wheel. The single entry CT-26 turbine is easily available, but it does not bolt to the stock MR2 turbo elbow, so serious modifications are required to use it. The most widely available alternative is to clip the turbine wheel. Clipping the wheel does reduce its ability to extract energy from the exhaust, but since there is so much exhaust at the higher RPMs where the stock turbine wheel chokes the engine, clipping is a reasonable choice. Dyno runs with a 12 degree clip have shown ~15 rwhp improvements in peak horsepower and moved the choke point from around 5.5K to 6K RPMs. There is a slight increase in turbo lag, but this can be kept to a minimum if the compressor wheel is kept to a modest size (since there is really no need to go with a large compressor wheel as we determined in the 3S-GTE turbo sizing primer).

At the current time, my suggestion for a twin entry CT-26 upgrade is to add a TO4E-46 trim compressor wheel and perform a 15 degree clip on the turbine wheel. Those who have ceramic turbine wheels (only used in Japanese spec 3S-GTEs) cannot clip due to the brittleness of the material, but the smaller ceramic turbine is already a little less restrictive than the metal one and naturally supports about 10 rwhp more that its unclipped sibling. For all out performance, there is a CT-26/35R drop-in hybrid available from Extremeboost which couples a low-friction ball bearing central assembly and large GT35R turbine housing with an A/R of 0.85 (the stock CT-26 turbine housing has an A/R of 0.49) and a 56-trim compressor wheel. I have not yet seen a dyno chart of this hybrid turbocharger in action but it should be able to produce well over 300rwhp on a 3S-GTE. In fact, it may even require a complete fuel system upgrade to safely run at higher boost levels on the MR2.