The first vending machine in history​

The original Hydra-matic transmission was one of the most important innovations in the history of the car. It wasn't the first automatic transmission, but it was the first that actually worked. The great commercial success paved the way for all future automatic gearboxes.

We look back to the origins of the Hydra-matic and its inventor, Earl Thompson, who also developed the first synchromesh gearbox in the 1920s.

Switch

Recently, Ferrari and Lamborghini threw the dice by announcing that they would be phasing out the manual transmission in favor of F1 style sequential gearboxes. This announcement has rekindled an old discussion: have the manual transmission and separate clutch pedal become obsolete?

The shift has also been noticeable in the Netherlands in recent years. More and more models are supplied exclusively with an automatic transmission. Sometimes only the basic motorization is still available with a manual transmission. Due to the shift to electric cars, manual gear shifting will eventually cease. Electric vehicles do not require gears and are therefore already supplied with an automatic transmission as standard. 

Outside of a small group of enthusiasts and truck drivers, the American auto industry has seen the manual transmission, at best, as a necessary evil. The multi-speed gearbox, dating back to the 1890s, has adapted to the limits of early engines with limited power and narrow revs.

​

A gear ratio good enough for a heavy car to accelerate from a standstill would give the engine a run over at 25 km/h, while an optimal gear ratio for normal driving would take a hit if you hit a bump. on must. One of the advantages of early electric cars, despite their very limited range, is that they rarely needed a different gear. This was because the maximum torque is already available from 0 rpm. If the first car manufacturers had found more efficient ways to store electricity, the development of passenger car propulsion would have been very different.

An early sliding-gear transmission was rarely a pleasant experience. Even upshifting required precise timing and patience to avoid gear grinding, while downshifting required 'double clutching' and rpm matching. But few drivers mastered that perfectly. Tachometers were still a rarity on normal cars, so that didn't make it any easier. The strategy of many drivers was to drive them into top gear as soon as possible and keep driving until you had to stop again.

​

Car manufacturers were not blind to this problem and started looking for ways to alleviate this suffering. Henry Ford preferred planetary transmissions, he didn't learn to operate a normal gearbox until the 1920s. The Model T's pedal-driven planetary gearbox was marginally less complicated to use.

​

Preselector

Shortly after the First World War, a number of inventors, including the British Walter Gordon Wilson and the French Jean Cotal, began to explore the possibilities of a more sophisticated  'preselector' planetary transmission. With a preselector transmission, you select the desired gear ratio with a selector lever and engage it by pressing a selector pedal that took the place of the conventional clutch pedal. They were very easy. If they worked.  They were expensive, heavy, consumed a lot of power and too complicated for most mechanics. Wilson and Cotal preselectors were used in some British and French cars up to the early 1950s, but they never became a success.

Even these pre-selector transmissions were not automatic, although they did make shifting easy on their own. There have been experiments from 1904 with automatic shifting but all of them have had little success. Until the 1930s, hydraulic brakes were already advanced, let alone automatic shifting.

​

The Cord 810/812 was one of the few American cars equipped with a preselector transmission. A 4-speed Bendix 'Electric Hand' gearbox. It was used because it simplified the connection between the shift lever and the box placed in front of the engine.

EL Cord, the founder of the brand, was committed to technology, not selling the car. Despite the strange design, which many Americans could appreciate, the technique killed Cord.  Front-wheel drive with independent suspension, concealed headlights, missing running boards and more made the car look nothing like any other 1935 car. The production numbers were never reached and due to the difficult technique and the lack of interest of EL Cord  ensured that by the end of 1937 the story was over.

​

How it started

Of course the vending machine was not invented by one person and suddenly
miraculously introduced and an instant success. This is the first automatic that would still be recognized as such today. In France, Louis-Rene Panhard and Emile Levassor made a kind of automaton that unfortunately broke down immediately during a demonstration in 1894. The explanation was further completed with a demonstration drawn on a blackboard. In Boston, USA, the Sturtevant brothers worked on a 2-speed automatic transmission that had to work with centrifugal weights that would deactivate the clutch at a certain speed. A good idea in theory, but the execution was far from reliable. The Canadian Alfred Horner Munro had applied for a patent for a vending machine in 1923. His invention used compressed air (he was a steam engineer) but it wasn't really powerful or practical. In Brazil in 1932, José Braz Araripe and Fernando Lehly Lemos were working on a system that would use hydraulic pressure to handle gear shifting. Their ideas and prototype were eventually bought by GM for further development.  

​

Syncro-Mesh Silent-Shift​

Among the many people seeking a way to make shifting easier was a Portland, Oregon hydraulics engineer named Earl A. Thompson. Thompson's first step was to simplify the switching process. In 1918 he applied for a patent on what he called the 'automatic gear change mechanism'. It used conical synchronizers to equalize the speed of the gears before engaging them.  The patent for this invention was actually for a pre-selector transmission, but most importantly, it provided grinding-free shifting without 'double clutching'.

Thompson developed a working prototype and installed it in a new Cadillac donated by his younger brother Kirk who ran a Cadillac dealership in Spokane, Washington. In 1924, Thompson drove the Cadillac to Detroit to demonstrate it to automakers. Despite most automakers not wanting to hear from outsiders, Cadillac chief engineer Ernest W. Seaholm and president Lawrence Fisher believed Thompson's invention had potential. She forwarded it to GM's New-Devices Committee. Thompson was about to give up due to the general lack of interest. However, Cadillac was interested and took over the development itself and hired Thompson as a consultant.

Over the next few years, Thompson was developing his synchronizer mechanism. According to Ernest W. Seaholm, Cadillac built 25 prototypes that totaled 1.5 million miles at the GM Proving Grounds before production began in August 1928. The system debuted that fall on the 1929 Cadillac and LaSalle models.

From a cost point of view, these early synchromesh transmissions were only synchronized in 2nd and 3rd gears. To go to 1st gear generally meant stopping. Anyway, the system was a major improvement on the earlier 'crash' buckets and made driving a lot easier. Synchromesh soon became available on other GM brands and was subsequently licensed by many other automakers. By the mid-1930s, the system was standard on almost all American cars and many European cars.

When the first synchromesh cars went on sale, Fisher and Seaholm Thompson were promoted to assistant chief engineer. Thompson had taken the creaking out of the shifting process. The next step was to automate gear shifting.

Military transmission project

Since the mid-1920s, GM President Alfred P. Sloan has been trying to get development chief Charles Kettering to develop a working automatic transmission. This served both a commercial and a personal purpose, Sloan said he was not a good driver and could not handle a conventional transmission.

In the late 1920s, the Dynamics division began work on a friction transmission based on a patent acquired from Citroën. The design used  a set of toroidal bearings and gears to transfer the energy, combined with an automatic clutch. The whole functions just like a CVT. It was a collaboration between Buick and Cadillac, led by Owen Nacker, Cadillac, John Dolza, Buick and GM research engineer John Almen. 

Buick invested a lot of money in developing 'The Roller'. Although it worked well at the GM Proving Grounds, the system weighed twice as much as a conventional transmission and had a short life. Worse yet, Buick accountants had calculated that with the projected sales numbers, the Roller transmission would have to sell for at least $500, $8,000 in today's money. Harlow Curtice canceled the project in 1932.

At Cadillac, Earl Thompson had a completely different direction in mind. Ernest Seaholm said Thompson's inspiration came from a 1930s Daimler that Cadillac had bought for evaluation. This Daimler had a Wilson preselector transmission combined with a fluid clutch. The Daimler still had to be shifted manually, but Thompson saw the potential for an automatic transmission.  

At the beginning of 1932, Thompson started the project with developers Ralph Beck and Walter Herndon for its own automatic transmission. The following spring, Owen Nacker became assistant lead developer at Cadillac, allowing Thompson to devote all of his time to the project. The team was expanded with three additional members for testing, William Carnegie, Maurice Rosenberger and Oliver Kelley. The team was named 'Military Transmission Project' with a gruesome 'no trespassing' sign on the door. Although their work later had many military applications, it was only a code name, as during the development of the Cadillac V16, that project was disguised as a bus engine project. Cadillac also began buying up patents for other automatic shifting systems. Many were useless, but Thompson wanted to be sure and avoid lawsuits. So one of them was from the Brazilian inventors José Braz Araripe and Fernando Lehly de Lemos. Araripe claimed that the AST (see below) only got there because he traveled to Detroit and sold his invention for $10,000. It is unknown whether the patent of Oscar H. Banker for his machine was included. There is information to be found but no real proof.  

​

Maybe it was the Wilson transmission, but Thompson switched from the preselector to a planetary gear set. This system has been around for some time in the car industry. Henry Ford preferred it, as did Ford developer Howard Simpson, whose own later automatic transmission system would be used by Chrysler, Ford, GM and Mercedes. Cadillac had already used two- and three-speed transmissions with a planetary gear set, but had abandoned this in favor of the sliding-gear transmission.

The advantage of a planetary gear system is that it changes the proportions by alternately applying clutches and brake bands. This can be controlled by solenoids or hydraulic or vacuum servos. Even with such automation, the driver still had to select each gear himself. Thompson and his team devised a hydraulic system that automatically determines shift points based on engine speed, throttle position and other inputs.

A single planetary gear is normally capable of only two forward gears, so most of Thompson's concepts used two planetary gears in series, thus providing 4 forward gears. The earliest designs, patented in '34 and '35, operated one gear set manually and the other automatically using a multi-disc clutch and a brake band operated by a hydraulic brake band. These designs still used a conventional clutch pedal and featured a 'low' and 'high' range. Low range offered automatic shifting between one and two, while high range offered automatic shifting between three and four. The driver started in 'low' and manually switched to 'high' range. Depending on the engine speed, the third or fourth gear was then selected.

While the results were promising, the project quickly became unaffordable. The Depression left Cadillac's financial situation rather precarious. Around 1934, GM even considered disbanding the Cadillac brand. Service manager Nicholas Dreystadt convinced the board to give Cadillac a second chance, unfortunately cutting the development budget in half.  The development of the power steering, which took place around the same time, was canceled. The automatic transmission project was too valuable to scrap, Alfred P. Sloan said. In January 1935, Thompson and his team were transferred to the corporate development team, reporting to development chief Ormond E. Hunt.

​

The Automatic Safety Transmission by Olds

The automatic transmission project soon came to the attention of Oldsmobile manager Charles McCuen and head of development Harold T. Youngren. McCuen was familiar with Buick's earlier 'Roller' project. He was enthusiastic about developing an automatic transmission for Oldsmobile. He instructed Youngren's team to work closely with Thompson's Transmission Development Group to develop a production-ready version for Oldsmobile.  Thompson and his team were responsible for the conceptual design while the Oldsmobile staff concentrated on translating those ideas into manufacturable form.  

The first result of these efforts was a semi-automatic gearbox dubbed the Automatic Safety Transmission (AST) by GM. The safety aspect was that it was supposedly possible for the driver to keep both hands on the wheel. The AST had 4 forward gears while retaining a conventional friction clutch but the shifting is automatic. Keeping the clutch was a compromise, Thompson and his team still hoped to use a fluid clutch, but McCuen pressured him to develop something he could build and sell. As a result, there was too little time to develop a working fluid coupling.

The AST went into production in early 1937 and went on sale from June. Though essentially an Oldsmobile project, the transmission was built in a former Buick factory in Flint, Michigan. Buick engineers were not at all enthusiastic about the AST, they had their own ideas about an automatic transmission, which would culminate in the post-war Dynaflow. However, Oldsmobile did not have the capacity or equipment to build the AST. Buick spent $5 million on machinery that it recouped by billing Oldsmobile a hefty $180 for every transmission built.

The AST was initially an $80 option for the eight-cylinder Oldsmobiles, it became an option on the six-cylinder from the 1938 model year. Buick had no intention of using the AST, but under pressure from GM, Buicks director Harlow Curtis reluctantly introduced the ' Self Shifting Transmission' as an option on the 1938 Specials. Unlike Oldsmobile, Buick did not offer the transmission on the more expensive models and in the middle of 1938 delivery stopped due to low sales.  

The AST was by no means the first or the only semi-automatic transmission of the time. Reo had launched his Self-Shifter in May 1933 with much fanfare, but both the transmission and the brand had disappeared by the mid-1930s. By 1937 Chrysler was developing its own M4 semi-automatic transmission, which debuted in 1940.

​

Most semi-automatics were problem cases, the AST was no exception. There were many teething problems and transmissions so complex that Oldsmobile advised dealers not to repair them but send them to the factory for replacement. Buick offered little help, the developers in Flint considered it an Oldsmobile problem. The production  of the AST was very limited. Only 7% of 1937-1939 Oldsmobiles and probably 3000 Buicks used it. Since Oldsmobile paid more for each transmission than they asked for it, it cost them a lot of money. It appears that the purpose of the AST was to gain more real-world experience for the fully automatic transmission that Thompson and McCuen hoped one day to offer.

Hydra-matic

By 1939 Oldsmobile had built 5000 pre-production cars with a new version of the AST called Hydra-matic. Internally, the Hydra-matic was comparable to the AST. However, it uses a third planetary gear for reverse gear. Two oil pumps, one driven by the engine, the other by the output shaft, provided the working pressure of the valve housing. A fluid clutch replaced the conventional clutch. Oliver Kelley, the chief designer and the person to whom the patent was filed, called it the 'liquid turbo clutch'. This clutch had an unorthodox function that would become the Hydra-matic's trademark: to prevent the car from creeping forward at idle, the engine flywheel actuated the fluid clutch impeller through the front planetary gear train, rather than directly. As a result, the impeller always turned slower than the motor, reducing the creep tendency.  

As with the AST, 'low' range provided automatic switching between one and two. Drive functioned just like the AST's 'High' range, but the 'governor' has been modified to automatically shift between all four gears. The shift from two to three was still complex and tended to be jerky if the tires and links weren't adjusted perfectly. There was no clutch pedal and all you had to do was manually shift to neutral or reverse. There was no park position, but there was a parking catch that could be activated by putting the transmission in 'reverse'. First gear was very short because a fluid clutch is not very efficient. The first gear was so short that the car would shift up if you hadn't even started moving yet. That is why the Hydra-Matic also had 4 gears.

Unlike the AST, the Hydra-matic was not built by Buick but by GM's Detroit Transmission Division in Livonia, Michigan. Production started in May 1939. When the 1940 Oldsmobiles went into production in October, Hydra-matic became a regular option. Remarkably, the price of $57 was lower than the last AST ($73). Probably again a lot lower than the actual production cost.

When Alfred P. Sloan, GM's chief, took his first ride in an Oldsmobile equipped with Hydra-matic, he was impressed. 'For 15 years I have felt that a gear lever does not belong in a really modern car. It's only a matter of time until all cars will have this type of transmission." His statement was already very close to the truth. A manual transmission is rare in America. In recent years, we have also seen a significant shift in Europe towards automatic transmissions, some brands no longer even supply manual transmissions in the more luxurious models.

​

​

​