The rarest and least known of all the GTAs was the SA (supercharged) version, presented at the third exhibition of racing cars in Turin in March 1967 and then re-presented at the Geneva Motor Show that year. 

The abbreviation SA was an acronym for Sovra Alimentata which indicated the presence of two coaxial centrifugal compressors fixed to the sides of the carburettor group, which, via a watertight collection box, fed the two double-bodied Weber 45 DCOE, typical of the GTA (129). 

The car was designed for group 5, which in the permissive interpretation of the FIA rules left the possibility of massive elaborations on production models. This concept, exploited at the time by English and German tuners... often changed the lazy family sedans into real racing cars for the racetracks and prompted Alfa Romeo to add to the GTA (already crowned queen of the European Tourist Challenge Championship) its 'force-feeding' sister. 

Engines with compressors were already in the DNA of the Casa del Portello, and the memory of the Alfetta 158 and 159 indicated the choice of the turbocharged engine to increase power as a natural solution. 

The authorship of the project at the Centro Esperienze belonged to the engineer Gianpaolo Garcea.  The lack of enthusiasm with which Carlo Chiti accepted the idea is known but as a good commander on the battlefield, he tried to translate the project into a useful racing weapon. 

In truth, the project was already obsolete at birth, and the idea of dramatically increasing the power of the small aggregate and putting it in a position to compete with the Ford or Porsche bigger displacement engines, was destined to be short-lived right from the start. However, GTA SA passed like a meteor leaving a long trail of unanswered questions, mysteries, and uncertain data, which still today are the subject of curiosity of historians and simple enthusiasts. 

Very little is known about the GTA SA starting from the number of specimens assembled, the geometry and origin of the engine, its permanence on the tracks and its fate after retiring from competition. 

The SA represents perhaps the rarest and most mysterious object in the entire history of the Giulia 105 racing cars. Already during the winter of 1966, the “strange” GTA had been noticed on the Balocco track, from which Dorino Zeccoli was trying to extract the secrets of its wild nature and decipher the peculiarities of its character, which only years later he would describe as “a tragedy, the a car in which a driver of today would certainly not have set foot”. Coming from Zeccoli, the driver who more than any other knew the secrets of Alfa Romeo cars, this judgment takes on even greater weight. 

According to Zeccoli, the GTA SA was an unpredictable machine, difficult to drive, delicate and fragile, but also full of charm, and in easy situations, it became the winning opponent. GTA SA hid its talents well and from the outside nothing, not even the acronym, betrayed the difference from the other GTAs. Only the well-trained eye could notice the rear wheel tangs made to accommodate the wider Dunlop 65SC-5.50 M. The absence, sometimes noted, of the shield in the grille was not a distinctive sign, but only an instance observed on some occasions. Looking for the differences, only a pressure gauge on the dashboard would have been noticed and until the engine compartment was opened, the appearance of the SA remained unchanged compared to the other racing GTAs. 

Instead, once the bonnet was raised, the differences leapt to the eye. The already limited space of the GTA engine compartment was drastically reduced with the "additions" to the carburetor side of the entire complex structure of the compressors and the airtight box that fed the GTA SA engine. As already mentioned, the place of the air intake box of the racing GTA was replaced by the airtight box, with at the ends two turbines, of small diameter (7cm). Schematically, the operation of the supercharging system followed a simple logic. A piston pump, located on the intake side of the engine, driven by a chain from the engine, sent high pressure oil (80 kg/cm2) to two coaxial turbines with two centrifugal compressors which rotated at a maximum speed of 95,000 rpm. The compressors sucked in air from the outside and sent it compressed into the carburetor feed box, forming the mixture which, through the admission ducts and the intake valves, reached the cylinders and this air-petrol mixture was not only " sucked" by the pistons but was also "compressed" by the compressors. This resulted in a considerable improvement in the volumetric efficiency. With a high number of revolutions, the turbines created a pressure of 0.6-0.7 bar, which pushed the mixture into the usual Weber 45 DCOE 14, which in turn injected it compressed at 16.8 bar into the combustion chambers, generating a power of 220-240 HP. 

This tremendous driving thrust in theory had to be available immediately and in a linear way... given that the feed pump was connected directly to the crankshaft precisely to avoid the classic "gap" of the turbines. The jump in power, however, only occurred (and often in an unpredictable way) from 3000 up to 7500 RPM. To remedy the immediate drawbacks, such as the very high temperature of the oil that moved the pump and the danger of detonations, caused by the inevitable small leaks, an additional radiator was installed to cool the oil, with the appropriate creation of a injection of water into the pipes to the combustion chambers, to avoid possible fires during operation. Nonetheless, the SA proved to be an unreliable car and prone to catching fire easily. 

The other fly in the ointment was its behavior, which Zeccoli described as an "unpredictable burst of power, which arrived suddenly and without warning, making the SA difficult to steer when cornering or in maneuvering situations". 

Up to 3000 rpm the engine was suffering and below the performance of normal use, but when passed to the supercharging regime, it revealed itself to be a beast of enormous power, capable of downgrading any other car. At the time, in the absence of the "wastegate" valve, the compressors were difficult to manage, and the SA, otherwise identical to its 70 HP less sister, was unable to exploit the braking power of the engine, and in any case showed poor stability in sudden braking. 

What was certainly not lacking beyond 3000 rpm was the enormous torque which pushed the 780 kg weight of the SA with incredible ease in every gear ratio at lightning-fast accelerations. 

Another innovative element that appeared on the GTA SA was the transistor ignition, the true progenitor of future systems, which from 1968 replaced the classic Marelli S119 distributor. But the real Achilles' heel remained the easy overheating in medium and low speed conditions. Excessive heat dissipated better at higher speeds, aided by a good cooling system, but in circuits with low speeds and low revs engine overheating was the order of the day. 

Another problem consisted in the frightening consumption of petrol, which reached 32 liters per 100 kilometers under racing conditions and penalized the SA in long races, forcing it to stop frequently. 

The torque that the SA developed sometimes caused easy slipping of the rear wheels even in high gears, and the more frequent use of the brakes in endurance races (given the poor performance of the engine braking effect), exposed the whole braking system to the “fading”.

But the real mystery in which the story of the Giulia GTA SA remains shrouded begins with its technical content, never fully disclosed (or the data has been lost), so much so that over time two schools of thought have developed, which see the SA with different eyes. 

Most Italian authors, describing the SA engine, cite the bore and stroke values as 86x67.5 mm, having in mind a super square engine, which perfectly corresponded to the research of a brilliant engineer who was undoubtedly Chiti. 

Given the period of SA engine development which coincides with the absence of the 1300 powerplant, with which the SA would in theory share the piston stroke, crankshaft, bearings, connecting rods and other mechanical components, some are looking into the experimental engine developed for the F2 (86x68.5 mm), the base from which all the specific mechanical components have been transplanted. The cylinder head could only be the classic "double ignition", slightly adapted to the purpose; for the rest, the bulk of the changes concerned the crankcase, where the oil pump that fed the compressors was housed. The shortened stroke and the increased bore were the logical way to reduce the speed of the stressed piston, and many are inclined to believe that this was the solution adopted, given that even the specialized magazines of the time highlighted this innovative technical side. But given that the compression ratio was considerably lowered (8.5:1) and that it was not possible to adopt the classic GTA pistons, this opened the bore chapter, which really contains many unknowns. 

On an engine intended to run continuously hours and hours for the entire length of long races, thinning the shell of the barrels to a much lower thickness than usual, with the subsequent weakening of the housing of the barrels (holes significantly enlarged) meant exposing the crankcase (the entire engine) to a critical weakening, absolutely not able to fulfill the duration task under a regime of high stress. Since everyone agrees that in the case of the GTA SA engines only separate barrels were always and only used, and that the single sleeve (perhaps more robust and rigid) made its appearance only on the GT Am, the presence of detailed descriptions and many precise data cited for an unlikely solution certainly remains a mystery. 

Perhaps, we must take into consideration the presence of two different variants of the same engine, already at the time. In favor of this thesis, the different provenance of mechanical components can also be found (Portello and Settimo Milanese, with respective markings) of the various engines that have survived to this day. However, it seems that all the engines recently examined (and here we are at the second school of thought) are based on the classic geometry of the GTA, i.e. 78x82 mm, and this figure speaks in favor of those who stated that the engine block was the traditional 10532.01.010.99 elaborated for the GTA. Given the presence of the same values on the classic racing GTA, it is to be assumed that particular pistons were adopted, considering the lowering of the compression, which only with the boost factor approached the declared 10.5:1. 

In this case things were simpler, and using the 10532.01.053.99-cylinder head and sodium-cooled valves...the rest of the GTA kit was logical and easy. The mystery of the number of engines assembled, however, has never been solved. No one can say with absolute certainty how many engines and how many cars were assembled in the SA version. Opinions quite agree on the hypothesis that no more than twelve engines were produced for racing use, while for the number of cars the estimate does not exceed ten GTA in SA trim. 

Observable data regarding certain chassis numbers list the following cars, present in competitions between 1967 and 1970: AR 613015, AR 613016, AR 613056, AR 613069, AR 613470, AR 613929, AR 613919. After 1969, almost all the GTA SAs were reconverted to the classic aspirated version, i.e. with atmospheric pressure carburetors and the last certain appearance of a GTA SA dates back to 8 March 1970, when the GTA SA (AR 613069) driven by Christine Beckers placed second overall in the uphill race in Condroz. 

Shortly afterwards it was converted into an experimental version with the 1300 engine, but this changed little. 

"The GTA SA was already a legend…"