Get instant digital access.
Subscribe Now!

Pressure Cooking

It’s not hard to see why racers and hotrodders are drawn to turbochargers like moths to a light: In theory, at least, getting more go from a turbocharged engine is as simple as turning up the boost. A turbocharger is simply a very efficient way of increasing airflow through an engine by pressurizing the intake manifold, and so it follows that the higher an engine’s manifold pressure, the more power it will make, all else being equal, of course.

Without getting into all the messy details, this is true up to ridiculously high torque (and horsepower) numbers—or at least until the short block gives out. Turbocharged engines have the potential to make power at levels limited only by the builder’s bank account.

That’s the same reason why forced induction is banned in many forms of amateur-level competition outside of drag racing. In professional road racing series that allow turbos, such as the Grand-Am Cup and SCCA’s Speed World Challenge, organizers long ago learned that restrictions are necessary to prevent a boost-driven arms race. Without some limits, the end result would look something like the NHRA, with engine builders assembling ever more exotic engines designed for astronomical manifold pressure and correspondingly high torque.

But despite the reputation of turbocharged engines as powerful time bombs—particularly those that didn’t start out boosted—carefully designed and built turbo engines are no less reliable than other engines. In fact, home-built turbocharger systems assembled from collections of junkyard or aftermarket parts have the potential to produce a lot of reliable power for very little money. The keys are doing your homework and planning ahead.

Boost Theory

The biggest obstacle standing between a stock, naturally aspirated engine and high-boost power levels is proper control of fuel and ignition. Turbocharged engines have different fuel and ignition requirements than normally aspirated ones, and some way of modifying or getting around the stock engine controls is necessary. Without proper mixture and timing control, an engine will either not make power, or will run dangerously close to detonation and holed pistons from a lean mixture and advanced timing.

A turbocharger uses an engine’s exhaust gases to drive a compressor that forces air into the engine’s intake manifold. This compressed air is denser than the air in the surrounding atmosphere, so each time the intake valve opens, more air enters the engine than would otherwise be able to get in. So at a given rpm, a turbocharged engine is taking in more air than the same engine without a turbo, and will require more fuel than its non-turbo counterpart. The turbocharged engine will be producing more power as well.

Turbo engines also generally like much richer mixtures than do normally aspirated engines because a slightly rich mixture helps reduce the knock sensitivity due to the boost level and fuel octane combination used. Some engines are more sensitive to mixture strength than others, but while atmospheric engines might be happy at an air/fuel ratio near the chemically perfect (or stoichiometric) ratio of 14.7:1, forced induction engines like an air/fuel ratio that is much richer for both power and reliability.

According to Bill Cardell, noted Miata turbo tuner and developer of the Flyin’ Miata turbo kits, fuel mixture is very strongly related to boost pressure. “Fuel mixture should be tied to MAP [manifold absolute pressure], so at cruise or light load you’re running close to stoichiometric,” he explains. “At boost, on the gas available in most of the country [91 octane premium], you’ll need to be closer to 11:1 or 12:1.”

Much leaner than that, and knock and high piston crown temperature start to become an issue, while overly rich mixtures can make an engine feel sluggish with poor off-boost performance. Many factory turbo cars dump more fuel than necessary into the air stream for other reasons, however. Cardell explains it as a strategy for warranty work avoidance: “Original manufacturer-installed turbos are all running 10:1 under boost to keep the catalytic converters alive, which is not particularly good for power.”

The second important factor in turbo engine control is ignition advance. Thanks to the high cylinder pressures seen in turbocharged engines, they are more susceptible to pinging than non-boosted engines and must use less timing advance. As boost increases, the ignition timing needs to be retarded further and further, because the engine becomes more and more likely to ping. As Cardell says, “it’s not that timing needs are reduced under boost; you’d like to run as much as you can, but if you do you’ll get detonation.”

The exact amount of boost-driven ignition retard depends on many factors, including combustion chamber design, temperature, fuel octane, rpm and cam timing, and has to be determined for each engine.

Choosing an Engine

Any engine can be turbocharged successfully, although obviously some make better raw material than others. Turbocharged engines make their high power partially through high cylinder pressures, and it stands to reason that engines with sturdy rotating assemblies and low compression ratios are more suited to turbocharging than fragile, high-compression engines. Of course it all depends on your goals: A 200-horsepower, 2.4-liter engine running 8 psi of boost is a very different animal from a 350-horsepower, 2.0-liter engine at 30 psi.

Engines that share significant parts with a factory-turbocharged variant make the best candidates. The Nissan SR20 family, Buick V6, Ford 2.3-liter and Mazda Miata/323 engines all fall into this category. They are very easy to turbocharge because the factory has already done most of the development work--just find the right parts and go for it. Other sturdy engines like the Nissan KA24 and single-cam Honda engines also work very well, but they require a bit more planning and extra fabrication to get to the same result.

The lack of factory parts help shouldn't dissuade anyone from attempting to build a turbo engine. The mechanical "bolt stuff on" part is pretty straightforward, at least on paper: Find the parts, fabricate some plumbing, and bolt it in.

As long as you have done your homework and can handle the level of fabrication required, most conversions are not terribly difficult. Take the time to choose the right parts and think carefully about layout and fabrication, and you're halfway there. The other half comes after assembly: Proper tuning and testing are essential to building a successful turbo engine.

If you keep your horsepower goals modest and take the time to tune your engine conservatively, you probably don't need to do anything to the internals. Most modern engines will handle boost just fine, as long as the engine is never allowed to experience pre-ignition or pinging.

Engine Building

If you like to spend money, or if you're rebuilding your engine already, plan ahead for the highest power level you can afford to run. Keep your budget in mind and shoot for a power level that keeps you within it.

Don't spend money on exotic cams and fancy valve train parts for a street turbo engine build--turbos can add a large amount of torque, so a high rev limit isn't really necessary. Do spend money on decent head gaskets, sturdy pistons and head porting; the more air the engine flows, the more power it will make at any given boost level.

When it comes to deciding on a compression ratio to run with a turbo, there are two schools of thought; each has its professors and students.

The first theory is to run as much compression as possible, but only as much boost as absolutely necessary to achieve the desired power level. The other group is the "low-compression, high-boost" camp: Build an engine with compression of less than 8.0:1 and boost the heck out of it with a huge turbo. A good argument can be made for either side.

High-compression, low-boost engines may not have the ultimate power potential of low-compression, high-boost engines, but they tend to have better off-boost performance. The high compression gives the engine better pull out of low-speed corners, but limits the amount of boost and timing that can be run before detonation causes problems. A low-compression engine, on the other hand, can run more boost and timing with the same fuel, and will be able to make more high-rpm power.

Of course, an engine's combustion chamber shape and design have a large effect here--a more detonation-resistant combustion chamber design will allow more compression and timing at a given boost level. Even the cylinder head material matters, as aluminum transfers heat out of the combustion chamber faster, which tends to reduce detonation-causing hot spots and allows higher compression ratios and more timing advance.

Tuning and Setup

To properly build a turbo setup without blowing apart the engine on the first run, you need some way of monitoring the air/fuel mixture and boost pressure.

The cheapest way to take care of the mixture monitoring issue is with a simple "narrow-band" oxygen sensor and a digital volt meter. This works in a pinch, but unfortunately narrow-band sensors are neither very accurate nor very useful at mixtures other than stoichiometric (14.7 pounds of air to 1 pound of gasoline). Since turbo engines run best at very rich mixtures, narrow-band sensors aren't very helpful.

So-called "wide-band" oxygen sensors, such as the widely available Bosch LSU, are a much better choice. They are very accurate over a wider range of air/fuel ratios, assuming they have been calibrated.

The control hardware for these sensors is more complicated (and expensive) than a simple meter, but if you are handy with a soldering iron, there are several kits that shouldn't take long to build, and that will significantly reduce the total cost of a wide-band system. If you plan on taking the car to a dyno for final tuning, an in-car oxygen sensor may not be necessary except as a general warning of dangerously lean mixtures.

Incidentally, consider dyno tuning a fairly important step in building a bucks-down turbocharged engine. An hour or two of dyno time and the advice of a good engine tuner are definitely cheaper than broken parts, as well as easier than fighting with intermittent drivability issues. A dyno pull is always good for showing the weaknesses in a setup, whether it is boost creep (when wastegate flow is too low and boost cannot be controlled), or fuel delivery problems that only show up under high load conditions that are hard (or illegal) to maintain on the street. Dyno time is cheap enough now that there is really no excuse for doing without.

Some engine builders like to monitor exhaust gas temperatures in addition to exhaust oxygen percentage. Exhaust gas temperature (EGT) is a great tuning tool, but temperatures vary so much between engines that a baseline number has to be determined at peak power for every engine. Once the base exhaust gas temperature for good mixture strength and power has been established, the reading can give the tuner a lot of information about the combustion process.

Unfortunately, just bolting an EGT thermocouple (temperature sender) into the exhaust manifold without tuning on a dyno will not yield much information. Generally speaking, EGT ranges from 1200 degrees Fahrenheit to 1800 degrees, but there are no hard-and-fast rules. Factors including ignition timing, cam timing and intake design can all affect EGT. The right temperature for maximum power in your engine will not be the same as any other engine, or even your engine with very small changes.

Exhaust Manifold

Designing a home-brewed turbo setup requires more than just theory and tuning; some hardware has to be considered as well. These pieces mount the turbocharger itself to the engine and route the force-fed air to the right places.

Turbo exhaust manifolds don't have to be exotic. Especially on a street car, the most important characteristics of a turbo manifold are heat resistance and strength. These factors mean that thick, cast-iron manifolds have a lot to recommend them for many applications.

If you can use a stock or ported turbo manifold from another application, give it some thought--while they aren't as sexy as tubular headers, manifolds cast from a high-nickel iron alloy will last longer than just about anything else. Of course they are heavy and hard to modify, which eliminates them from consideration in many applications.

Cast iron is a great material, but if your engine doesn't have a factory turbo manifold or one that can easily be made to work, there are a few options for fabricating a manifold. The best-known suitable material comes in the form of thick-walled cast steel tubing shapes known as "weld-els," which are available from industrial supply houses. Weld-els make very good turbo manifolds. All that is required are manifold flanges, which are available from several sources.

Don't be tempted by regular mild-steel header tubing, as it won't stand up to the heat generated by the turbocharger. Medium-wall stainless tubing works well, but many amateur welders find it harder to use than weld-els.

As far as manifold design goes, simple log manifolds are on par with well-designed equal length manifolds. However, cast units are less prone to cracking due to heat expansion.

Ducting Air

Turbo engines require ducting to transfer compressed air from the turbocharger to the intercooler or intake manifold, and from the intercooler to the throttle body. This plumbing must be able to stand up to the expected boost level without blowing apart, and the tubing must withstand the considerable heat of the compressed air coming out of the turbocharger.

PVC tubing has been used successfully in the past because it is cheap and comes in many diameters and bend radii, but its heat resistance is not very good. It can be used for cold-side (before the turbo) ducting in a pinch, but there are many better materials for permanent installations.

Mandrel-bent mild steel tubing, like the simple U-shapes sold at most muffler shops, works very well for intake tract plumbing and is easy to use. Pre-bent aluminum tubing in the right sizes (2-inch to 4-inch diameters) is harder to find and more difficult to use, but it is much nicer looking and weighs less.

The biggest problem with using steel and aluminum bends to fabricate plumbing is keeping the rubber couplings connected to the ends of the hard tubes. Lots of shops have bead-rolling machines, and they can bead the ends of the tubing for a fee. However, the least expensive method of hose retention is to weld a thin bead around the end of the tubing. The bead does not even have to go all the way around the tube, as several short strips of weld work just as well to retain a hose clamp.

There are a couple of ways to deal with the flexible couplings like those used between the intercooler inlet and outlet, throttle body, turbo and similar. The best, and the most expensive, is fiber-reinforced silicone hose. It is thin, strong and resistant to high temperatures. This material also comes in just about any size.

If silicone hose couplings are too rich for your blood, head over to your favorite truck parts dealer and have them cut you some segments of diesel radiator hose. It is thicker and less flexible than silicone, but strong enough for most installations.

Oil and Water

The central shaft that connects the impeller (exhaust turbine) to the compressor wheel in a turbo spins at a pretty high rpm--well past 100,000 rpm in some cases. To survive these harsh conditions, the center bearings of the turbocharger require a good supply of clean, cool oil. In addition, the seals on either side of that bearing cannot cope with much pressure behind them, so getting the oil back out is very important.

The oil pressure feed line to a turbo does not have to be very large. An AN -4 (1/4-inch inside diameter) hose taken from a good pressure source like the oil pressure sender port will usually do the trick, and is the size used by a lot of production turbo installations.

The drain line should be short and as straight as possible. It should also be large in diameter, at least 1/2 inch (AN -8) or so. Most factory turbos drain into a special port on the side of the block or the oil pan, which are the easiest places to attach a large fluid fitting.

Because a loss of oil pressure will result in instant death for the turbo, oil feed lines are not a place to skimp. Small-diameter AN lines and fittings aren't nearly as expensive as a rebuilt turbo or new center section, so use good hoses and hardware if possible. The drain line is also important, so it's imperative to use the right material. A good truck oil line should be sufficient (bigger is better), but be sure that in never travels "uphill."

Turbo center bearings are cooled by the oil flowing through them, but most factory turbos also have water-cooled center sections to help reduce oil overheating and coking problems. On a street car, the water cooling can significantly prolong the life of turbo bearings without any special care. Center sections cooled only by oil simplify installation, but at a possible cost in longevity unless the car is always idled down.

The coolant ports in the turbo center section are usually female NPT threaded, and are pretty straightforward to fit. The lines don't have to be large in diameter, as AN -6 or -8 lines are all that is needed in most cases.

Even engines with water-cooled turbos demand more from their oil than do naturally aspirated engines. If the engine in question does not have an oil cooler, it's definitely a good idea to include the cost of a simple cooler in a turbo conversion. At the very least, put an oil temperature gauge on your engine and monitor it religiously until you can be sure that the oil stays below 240 degrees Fahrenheit, preferably between 180 and 210 degrees. Much hotter than that and oil lubricating ability drops off very quickly.

Boost Control

If you simply bolted a turbo to a car's engine and hit the road, there would be no control of the boost level. Boost would rocket up to the highest level allowed by the intake and exhaust restrictions of the engine and the turbo's capacity and stay there, leaving a trail of broken parts on the road behind.

In the days before wastegates, designers of turbo engines avoided this situation by using restrictive plumbing to reduce boost. The 1964-'67 turbocharged Corvair, for example, used a tiny side-draft carburetor that restricted boost to around 6 psi. It worked great at the time, because there was no other good way of regulating boost. The problem with this approach is that it causes the turbo to run less efficiently than possible while heating the intake air even more.

Modern turbo engines don't suffer with restrictions. Instead, they limit boost by using a small valve called a wastegate to divert part of the exhaust gases around the turbine impeller and back into the exhaust stream. The wastegate opens when boost hits a preset value.

Most wastegates are flapper valves operated by a diaphragm canister. The most common OEM variety are cast integrally with the turbine housing and turbocharger outlet. While most have a preset boost level, they can easily be fooled by a simple bleed orifice in the line that feeds pressurized air to the diaphragm. Wastegates can also be controlled by complicated electronic or mechanical boost control valves that allow boost to build quickly before stabilizing.

The cheapest way to implement a boost control is to choose an OEM turbo with an integral wastegate. There are two problems with integral wastegates, however. The first is flow: The gases blowing out of the wastegate hit the main flow of exhaust coming out of the turbine, and cause turbulence right where you don't want it.

The second problem is their ability to hold a preset boost level. Some integral wastegates are too small for engines with very high exhaust flow, and do not allow gases to bypass the turbine fast enough. The excess exhaust continues through the turbine and boost increases uncontrollably.

Both of these problems can be avoided with an external wastegate, though at a price. For a street-driven turbo engine, an internal wastegate is usually the best choice, but several manufacturers sell high-flow wastegates that can be welded onto an exhaust manifold and have enough flow to tame the worst boost creep problems.

A blow-off valve is located in the intake tract of a turbo engine and is often mistaken for a boost control device. Its purpose is not boost control, but rather to eliminate boost spikes from occurring between shifts that can cause the turbo compressor to stall.

When the driver slams the throttle closed for a shift, the rapidly moving column of air between the turbo and the intake slams into the closed throttle plate and reverses direction. This momentary reversal sends a pressure wave through the intake ducting back to the turbo compressor, stopping or slowing it for a fraction of a second. The compressor must then accelerate to provide boost again after the shift.

A blow-off valve eliminates this problem by popping open when the pressure in the intake manifold (on the engine side of the throttle plate) drops below atmospheric. One side of the valve is exposed to the outside air, while the other side is connected to the turbo pressure ducting near the throttle plate. When the valve opens, the pressure wave exhausts into the air and the turbo compressor continues to spin.

Blow-off valves work very well, but they are not the best for engines with mass airflow sensors. With a mass airflow sensor, the momentary pressure release occurs after the air has already been measured. This air loss fools the ECU into thinking that there is more air in the engine than there actually is, and the mixture momentarily goes rich. One other problem with some MAF-based systems is that many blow-off valves will remain partially open under vacuum, allowing unmetered air into the engine causing the engine to go lean.

With this kind of engine control system, a recirculating valve is used. It works the same as a blow-off valve, but the released air is returned before the turbo inlet, but downstream of the mass airflow sensor.

Turbo Selection

Selecting the proper size turbo for a given engine seems more difficult than it is. There are a lot of ways to approach this, depending on whether you intend to use an off-the-shelf turbo from the junkyard and do a quickie rebuild, or buy a new or rebuilt custom piece.

Any good turbo book (such as "Maximum Boost" by Corky Bell, or Hugh MacInnes's classic "Turbochargers") can walk you through the process of calculating your engine's airflow demands, target horsepower and turbo flow requirements. With this information, you can either read a turbo map yourself or find a local turbo shop to help you through the process.

If the budget stretches to accommodate a new turbo, then by far the easiest way is to simply call one of the custom turbo manufacturers. These companies can easily recommend a turbocharger that matches your engine combination, driving conditions and some other bits of key information. There are a lot of small, local turbo shops that specialize in building custom turbos from OEM parts to fit a new application, and a quick check of the Internet will tell you who is trustworthy and whom to avoid.

Most of the OEM turbos that are widely available are not particularly good from the standpoint of maximum horsepower, but they work well if they are used within their range of acceptable performance. When most manufacturers design a turbocharged engine, the factory turbo is generally on the conservative side. Small turbos have better drivability on the street for the average driver, but they can choke high-end performance.

If you find an OEM turbo that seems "about right" for your home-brewed application, check with tuners who specialize in the engine on which the turbo was originally found. They can help identify the limitations and power potential of that particular piece.

Make things easier by sticking with turbos that have a standard manifold flange. Some OEM turbos (certain Mitsubishi and Dodge pieces, for example) use a nonstandard flange. Using one of these turbos will make later upgrades much harder, since you will have to redesign the manifold.

Cooling the Charge With Intercooling

Turbocharging the incoming air does have one major inefficiency: increased air temperatures on the pressure side of the system. Depending on many factors--including pressure ratio (boost as a percentage of ambient air pressure), ambient temperature, airflow and compressor design--the air coming out of a turbo can be a hundred degrees or more above ambient, and therefore less dense than would otherwise be the case. This counteracts the effects of high boost and results in less power.

Most turbo installations solve this problem by using an intercooler to drop the temperature of the pressurized air going to the engine. Intercoolers can be either air-to-air or air-to-liquid, with an intermediate heat-removing liquid and pumps to circulate it through an external radiator. An air-to-liquid intercooler has the potential to cool the intake charge better than an air-to-air intercooler can, but the added weight, complication and expense of the fluid pumps, plumbing and radiators make them less attractive for use in simple home-built systems.

Air-to-air intercoolers come in a baffling array of sizes and shapes. A fabricator who is handy with aluminum should be able to weld up just about any shape necessary using commercially available cores and sheet aluminum end tanks.

Most OEM intercoolers are compromised in some way for packaging reasons, but some are better than others and make good donors for custom-welded coolers. In addition, turbo diesel trucks have enormous intercoolers, and some of them can be adapted to cars without too much trouble.

There are two characteristics of intercoolers that are important from a performance standpoint. The first is heat transfer efficiency. An ideal intercooler would drop the temperature of the incoming pressurized air to the temperature of the ambient air. The greater the differential between the temperatures at the intercooler's outlet and inlet, the better the intercooler is doing its job. Larger intercoolers almost always have better temperature characteristics, since the larger mass will absorb more heat, and the larger surface area will contact more of the airflow.

The second characteristic of an intercooler is the pressure drop from one side to another. An intercooler that loses 2 psi from inlet to outlet is somewhat self-defeating, since it would have to cool the air significantly to make up the 2 psi that was lost. Too much pressure drop also forces the turbo to work harder for the same power level.

Pressure drop across an intercooler should be kept to 1 psi or less. Intercooler design plays a large part in pressure loss: A long intercooler with air entering and exiting from each end will cause more pressure loss than the same intercooler with air moving from top to bottom.

Fuel and Air

As a turbocharger crams more air into an engine, a subsequent increase in fuel is also needed. Fortunately, there are a lot of ways to get more fuel into a turbocharged engine to make up for the increased airflow.

To make life easier, however, try to stay away from carburetors. Yes, they are tunable and, yes, there are successful carbureted turbo setups, but they can be more trouble than they are worth, since neither method of turbocharging carbureted engines works very well.

"Suck-through" carbureted turbo engines, which have the carb (or carbs) placed in front of the turbo, work well, except that most turbos lack seals against the vacuum on the throttle side of a carburetor. A turbo will flow a steady stream of oil into the engine without these seals. An electronically fuel injected engine places the throttle plate downstream of the turbo, so it never sees vacuum, and the seals do not have to seal oil against negative pressures. Also, the fuel that is carefully metered by the carburetor into the incoming air stream will be centrifuged out by the compressor, leaning the overall mixture that reaches the engine.

"Blow-through" turbo installations feed pressurized air through the carburetor into the engine. This works great if you can seal all the possible leaks in the carburetors with something like the plenum used on the Maserati Biturbo. Otherwise, you could be looking at a high-pressure gasoline fountain in your engine compartment.

Electronic fuel injection suffers none of these problems and allows for complete control of the air/fuel mixture at a given engine load. All of the aftermarket parts available these days make installing an EFI system on a formerly carbureted car pretty straightforward: Provide a high-pressure fuel line and return line, and you are well on your way. If the engine started out fuel injected, then you need only find a way to increase fuel flow beyond stock levels.

Many people have successfully "fooled" their stock EFI system by installing larger injectors or increasing base fuel pressure. As long as the change is no more than 10 percent or so, this can work well. At idle and low speeds, the car's ECU--using feedback from the oxygen sensor--determines that the mixture is richer than expected and shortens the injector pulsewidth to compensate. At higher loads, the built-in fuel tables, combined with the larger injectors or higher fuel flow, give the engine more fuel with the same pulsewidths.

The problem with this strategy is the limited correction possible. Turbocharging a stock engine will probably outstrip the ability of larger injectors to supply enough fuel with the stock programming at very low boost levels. (See "Going With the Flow" in the May 2004 issue of GRM for more details.) Something more complex must be done to compensate.

One of the best ways to increase fuel flow under boost--without affecting off-boost performance--is to use a rising-rate fuel pressure regulator. Normally, fuel pressure decreases with decreasing manifold pressure and increases with increasing manifold pressure. Each 1 psi change in manifold pressure or vacuum results in the same amount of change in fuel pressure.

Rising-rate regulators increase fuel pressure faster than manifold pressure. They can be set up so that under vacuum they work linearly, as the ECU expects. However, under boost, they increase fuel pressure more rapidly than the manifold pressure increases. This has the effect of increasing injector flow just when necessary without affecting flow at low-throttle openings and cruise.

Incidentally, stay away from questionable practices like modifying vane airflow meter springs and modifying the engine coolant temperature sender signal. Both of these fool the stock computer, but will kill drivability in all but one set of conditions.

If you want to maintain your stock ECU and don't have a way to reprogram it, consider modifying the MAF/MAP signal with one of the aftermarket boxes such as the GReddy e-Manage or A'PEXi S-AFC. For moderate changes, these systems can "bend" the signal enough to keep the engine fueled.

Once again, though, such signal modification is fairly limited--to increase fuel flow, you have to make the engine's ECU think that more air is entering the engine than is actual fact. Since the ECU bases ignition timing and other factors on airflow, it will likely advance the ignition too far, causing pinging. Signal modification works within a few percentage points, but you can quickly run into a situation where the required signal modification does more harm than good.

Another ECU-external method of adding fuel flow is with additional injector controls. There are a couple of different types on the market, but only the most sophisticated are worth the effort. The additional injector (injectors) must be staged in gradually to keep the proper air/fuel ratio, which is where the older units did not do so well. Another strike against these systems is their price, as it isn't much more expensive to install a basic aftermarket EFI system or adapt an OEM system to an engine.

The most sophisticated solutions to the fuel control issue require a working knowledge of electronics and software, but the skills aren't hard to learn. One way to change an engine's fuel-control strategy is to modify the code in the existing ECU.

If your car is one of certain models of Honda, Ford, GM, Nissan, VW or Mitsubishi, you are in luck, since there are commercial and noncommercial tools available for modifying the factory ECU programming. A successful result with some of these tools requires more than a passing familiarity with electronics and computer software, but this is one of the best ways to modify a stock ECU to cope with forced induction.

If your ECU is not supported by any of these tools, there are a few other ways to get programmable fuel injection control. Some people have successfully adapted older GM ECUs to other engines. These computers use a MAP sensor to determine load, so they work very well with forced induction. Add to that the fact that the software code is very well understood and easy to modify, and you have a good budget solution for some applications.

The biggest problem with adapted or stock OEM computers is the steep learning curve involved in reprogramming them. If you're willing to tackle the challenge, then it can be the cheapest and arguably one of the best ways to deal with turbo conversions. Unfortunately, many of the software tools are less than user friendly, which does not help.

The most flexible option for controlling the fueling of a turbocharged engine would be an aftermarket EFI system. While the cheapest commercially available systems cost upward of $1000, there is at least one that can be built at home for much less: The MegaSquirt computer designed by Bruce Bowling and Al Grippo. This system has received a lot of buzz in recent years, and deservedly so. For less than $300, you can build your own EFI system, including the wiring harness.

Totally programmable, the MegaSquirt system makes a pretty attractive choice for a turbocharged engine build. It has its limitations, including no completely satisfactory way to control ignition timing, but is an excellent choice for a true low-buck turbocharged engine. (We're currently installing a MegaSquirt system on our own Volvo 142 project car; look for the articles detailing that process to debut later this year.)

Ignition Considerations

Another upside to using programmable factory or aftermarket ECUs is the ability to control ignition timing. As previously mentioned, turbo engines need to have their ignition timing backed off as the boost increases. There are a few sophisticated ways to do this, and a few simple ways.

The simplest system uses a smog-era vacuum advance canister or an aftermarket equivalent. Mounted on a distributor (assuming the engine has one), some vacuum advance units can be set up as pressure retard units, backing off ignition timing as boost comes up. Unfortunately it is very difficult to modify the amount of ignition retard at any one point along the curve without changing the whole curve and losing power at some boost/rpm points or allowing pinging at other points.

The second simple way to retard timing with boost is with a box similar to the MSD Boost Timing Master. This unit does the same thing as a pressure-retard diaphragm, but electronically. Newer versions allow the user to specify a boost point to start retarding the timing, as well as the amount of timing to take out per pound of boost. The Boost Timing Master's only failing is an inability to change timing based on engine speed as well as boost.

The best solution is to use an integrated fuel/ignition ECU and plot the timing curve based on manifold pressure, throttle position and rpm. Most OEM ECUs can be programmed this way, as can higher-end aftermarket systems. Any method you choose will require dyno time to set the baseline timing and boost-dependent retard levels.

Go Forth and Boost!

Turbocharging is an incredibly complex subject, and there's plenty more information out there. For more information, check out the two books we mentioned earlier, or spend some time talking to engine builders and the owners of local turbo shops. Don't expect them to reveal all their secrets, but you will be surprised at how easy it is to build a turbo engine.

In the case of most modern cars, a conservatively tuned, intelligently designed turbocharged engine won't require significant modifications, unless you are shooting for very high power numbers or drag racing. There are a lot of cars running around with rudimentary turbo setups that make good power without causing broken engine parts.

Plan ahead, shop smart, and you can't go wrong.

Donor Cars

In case you hadn't noticed, during the last 50 years most car manufacturers have produced at least one turbocharged engine, so there is a large base of parts to choose from in cobbling together a home-built turbo system.

Turbos, plumbing, intercoolers and even complete fuel injection systems pulled from turbo cars languishing in junkyards can provide the basic elements of a home-brewed system. The following list of donor vehicles doesn't cover every option out there, but it should give you a good idea of what is out there that's worth scavenging.

Audi turbos

All Audi turbo engines have KKK turbos that use nonstandard flanges. They are somewhat scarce, but can be used on other engines with the right manifold. Some newer Audis have very nice intercoolers that are suitable for custom applications.

Buick V6 Turbo

These cars are rare and valuable, but they do have decent, large intercoolers that work up to 250 horsepower or so. Their Garrett T3 turbos have nonstandard flanges, and the high-impedance injectors are not particularly large (30 lb./hr.).

Chevrolet Corvair

Don't be tempted, unless you're trying to turbo another Corvair. Old turbo designs are very inefficient, and these Airesearch turbos are truly tiny.

Diamond Star/Mitsubishi turbos

Mitsubishi's 4G63T-powered cars have a couple of different Mitsubishi turbos, some with nonstandard flanges. Early Starions have very nice, small intercoolers. Eclipses, Talons and Lasers have smaller, less-efficient intercoolers. They can be used, but only up to 200 horsepower or so.

Mitsubishi turbo injectors are large (380 or 440cc/min.) and happen to share the dimensions of some Honda injectors. As a result, these pieces have been used in home-brewed Honda turbo installations with some modifications.

Dodge Cummins diesel pickup

The Holset turbo used on these is enormous, but it has a standard T3 bolt pattern. If you have an engine that can use one, they are very good up to 400 horsepower or so. The intercoolers are heavy and not particularly impressive. The air ducting is also very large.

Ford Thunderbird Super Coupe

These supercharged V6 engines have smallish intercoolers that have been used in custom applications. Injectors are high-impedance and are rated at 32 lb./hr.

Ford TBI V8 engines

Many 1980-'89 Ford and Mercury cars (Crown Vic, T-Bird, Mustang, Marquis and Capri) with throttle-body injected V8 engines use two large, 42 lb./hr. low-impedance Bosch injectors fitted inside the throttle bodies.

At the salvage yard, break the thin casting on the top of the throttle body to get to the injectors—the injectors are greenish blue, and their part number ends in "400." These fit most Bosch applications just fine, but two O-rings must be used because the top does not have a machined-in retainer. Injectors this size should be good for 85 horsepower per injector.

Even better, Mustangs and Capris with H.O. throttle body-injected, 5.0-liter engines have enormous 55 lb./hr. injectors—good for 100 horsepower per injector. These injectors are gray, and their part number ends in "403."

Don't be fooled by the similar-looking injectors found inside V6-powered cars from the same era. These have a part number ending in "402" and flow only about 33 lb./hr.

Ford Probe GT (first generation models)

These have small IHI turbos and small, restrictive intercoolers. They have been used in other applications but can't be recommended.

Ford Power Stroke diesel pickup

These pickups have enormous turbos that aren't much use for conversions. They do, however, have big, beautiful intercoolers that will tame the hottest turbo outlet temperatures. If you can find a place to mount one, these are among the best production intercoolers available.

GMC Syclone and Typhoon

You're unlikely to find one of these in the junkyard. If you do find one, however, take everything you can get your hands on and sell it on eBay.

The air-to-liquid intercoolers are nice, but hard to use in other applications. They have large fuel pumps. The ECU is easy to reprogram, but the same ECU (part No. 1227749) can be found in early-'90s Quad 4-powered cars.

Isuzu NPR diesel delivery trucks

These box trucks have big, heavy turbos that aren't much use for most small engines, but they have very nice intercoolers. The intercoolers are slightly smaller than those found in Power Stroke pickups, but may be easier to package. They should be good for up to 350 to 400 horsepower.

Mazda 323 GTX, Mercury Capri Turbo

While rare, these can provide small IHI turbos and many other parts for Miata turbo conversions. The entire engine fits into a Fiesta pretty well.

Mazda RX-7 Turbo II

The turbos used on RX-7s are smallish, but they have good intercoolers. The injectors on some turbo Mazdas are huge, low-impedance injectors, up to 550cc/min. Early RX-7s have good oil coolers, but inspect them very carefully and clean them thoroughly.

Merkur XR4Ti, Ford Thunderbird Turbo Coupe and Mustang SVO

The Merkur XR4Ti, Ford Thunderbird Turbo Coupe and Mustang SVO all share the same basic turbocharged 2.3-liter, four-cylinder engine, and the salvage yards should be full of these cars. Most have a Garrett T3 turbo which uses a very common bolt pattern. Some automatics have a smaller exhaust housing, which causes boost to build earlier (and run out sooner).

Late Turbo Coupes have an IHI turbo that is smaller but has a standard bolt pattern. These IHI turbos have been used as upgrades on the Ford Probe and other cars originally fitted with a smaller version of the same turbo.

The manual-transmission T3 turbos have been successfully used up to around 250 horsepower, while the automatic T3 and IHI turbos will generally run out of steam around 225 horsepower.

The top-mounted intercoolers used on the Thunderbird and SVO are small and not particularly efficient, although some people have used them as cores for custom air-to-liquid coolers. From 1985 on, these cars had large, 36 lb./hr. low-impedance injectors that have been used in many applications.

Mopar 2.2-liter turbos

These cars have various small Garrett T3 and Mitsubishi turbos that do not all have a standard bolt pattern. They have smallish injectors and intercoolers, and are getting quite rare in the junkyards.

Nissan 280ZX, 200SX and 300ZX Turbo

These cars were fitted with various Garrett T3 and T25 turbos that are good for 200 to 225 horsepower depending on the model. The flanges are standard, but the exhaust outlets may not be.

Porsche 944 Turbo

Late 944 Turbos have very nice intercoolers that are mounted with the radiator. They flow very well, and have been used successfully by many people up to power levels of 350 horsepower or so. The end tanks are aluminum and easily modified.

Porsche 944 Turbo injectors are large (35 lb./hr.) but have an odd resistance and may not work in all applications. The turbos themselves are KKK units, which have an uncommon bolt pattern.

Saab 900 and 9000 Turbo

Saabs have smallish Garrett turbos on all but the very newest models. Early 900s have decent-flowing but small intercoolers.

Saab 9000 intercoolers look very much like Volvo 740 intercoolers, complete with plastic tanks. They are large but don't flow very well. Later Saabs have very nice low-mounted intercoolers that are good for up to 200 horsepower or so.

Volvo turbos

Early Volvo turbo (740) engines have a smallish T3 turbo with a standard flange. They have very large intercoolers that are efficient but flow fairly poorly. The intercoolers are good to perhaps 200 horsepower. Most have plastic end tanks so they are hard to modify. Later turbo Volvos, including the 850, have larger Mitsubishi turbos.

Join Free Join our community to easily find more articles.

Comments

View comments on the GRM forums
Flattop
Flattop New Reader
5/5/15 3:22 p.m.

Great read, Thanks for the article. I need more on this subject as I am not turbo knowledge blessed. Good thing for me there is a Nashco in my crowd.

Hungary Bill
Hungary Bill SuperDork
5/5/15 5:05 p.m.

Seriously a great read, thanks!

Our Preferred Partners
uxGbgAppzan6YhMJJjQmb12Ll6exOUPu