When most people think of GM triple-carburetor engines, they mostly think of the legendary Pontiac Tri-Power setup, but Oldsmobile actually beat Pontiac to the punch with its short-lived J-2 Rocket engine of the late 1950s. Here’s how it came about, and how the J-2 system worked.
In January 1957, Oldsmobile announced a new version of its 371 cu. in. (6,075 cc) Rocket V-8 engine: the J-2 Rocket, an $83 option that replaced the four-barrel carburetor with three two-barrel Rochester carburetors on a new intake manifold.
To the casual observer, this looked like Oldsmobile hastily imitating the new Pontiac Tri-Power setup, which had been announced weeks earlier, but wasn’t yet available to the public. In fact, it was likely the other way around: Oldsmobile had developed the triple-carburetor system first, but Oldsmobile assistant chief engineer Elliot M. (Pete) Estes had taken the idea with him when he became Pontiac chief engineer in September 1956.
There was nothing actionable about this sort of plagiarism — any intellectual property involved belonged to General Motors, not to the individual divisions, and corporate management usually did little to block inter-divisional thievery of this kind — but it did cause some resentment. “Our biggest competition was the rest of GM,” lamented Oldsmobile engine development engineer Gibson Butler. “As engineers, we were very proud of our ability to develop unique concepts.”
Multiple carburetor setups had been common aftermarket hop-up equipment for a long time, but automakers had mostly shied away. Such arrangements were finicky to tune, and if their linkages were synchronized to open all the throttles simultaneously, they could lead to overcarburetion in part-throttle driving, causing the engine to bog down. That was seldom a problem for racing, but it wasn’t ideal for normal street use.
By contrast, Oldsmobile proclaimed that the J-2 engine “provides flexible operation that gives plenty of power for all normal driving and keeps in reserve always available an abundance of extra power for unusual conditions.” The engine ran most of the time on the center Rochester 2GC carburetor, which Olds claimed gave “maximum economy” with good throttle response. For maximum power, a vacuum power control system then opened the front and rear carburetors, increasing total carburetor airflow from 280 CFM to 860 CFM.
Here’s an item from the June 1957 Motor Life on the new Olds system:
Editor Barney Navarro explained:
Though the triple carburetion of the Oldsmobile may resemble the type used for hot-rodding, it’s a far cry from being identical. Actually the only similarity lies in the fact that three dual throat carburetors are employed. Due to the great difference between this factory installation and the custom hot rod layouts, it will be necessary to compare certain points of each one. If this is not done, our coverage of the Olds assembly may serve as an endorsement for all custom triples causing many auto enthusiasts to be confused. Conversely, a prospective Olds purchaser may shy away from the factory triple because of past dissatisfaction with a custom layout.
… Being very much aware of the bad features that can be incorporated in a triple carburetion system, the Oldsmobile engineers took great pains to eliminate most of them. Until the advent of progressive linkage, synchronized throttles were common practice with all triple carburetion. But synchronizing throttles so that they all open simultaneously causes idle difficulties and poor throttle response at low rpm. A drag race enthusiast may not mind such faults but a passenger car produced for mass consumption just can’t be balky.
Where the Olds J-2 system (and the very similar Pontiac Tri-Power system) first parted ways from a typical aftermarket multiple carburetor setup was that the front and rear carburetors were greatly simplified. Neither had a choke, and they did without idle, part-throttle, and power circuits — they were designed to operate only at wide-open throttle. Only the center carburetor had a choke and the various circuits normally found on an automotive carburetor.
Navarro continued:
Unlike the custom linkages that can be purchased at speed shops, the Olds type is not completely mechanical in actuation. It has no direct connection between the center carburetor and the two end ones. The center carburetor is the only one that is connected to the throttle pedal. The end carburetors are opened by the action of a vacuum operated diaphragm after the center carburetor is opened 60 degrees or more. A vacuum switch is mounted on the center carburetor to effect this control.
Here’s how Motor Life illustrated the system. (I’ve color-coded the call-outs to make them a little easier to see.)
Actuation of the vacuum switch was mechanical: There was a tang attached to the accelerator pump lever of the center carburetor, which would swing upward as the throttle valves opened.
Once the accelerator pedal was depressed beyond about three-fourths of its travel (corresponding, according to the service manual, to about 68 degrees of throttle valve opening), the tang would move the switch and open a vacuum port to the diaphragm on the front carburetor. This would compress the diaphragm spring and flip the throttle valves open. Since the throttle linkages of the front and rear carburetors were synchronized by a mechanical linkage, opening the front carburetor would also open the rear throttle valves at the same time. If the throttle valves closed to less than 68 degrees, the vacuum switch would shut off the port and bleed air from the front carburetor air horn. The diaphragm spring would then return the diaphragm to its starting position, closing the throttle valves of the front and rear carburetors.
Since the front and rear carburetors were unchoked, there was a mechanical lockout lever that blocked their throttle valves from opening if the choke was even partially engaged.
Navarro continued:
The most amazing, we found, is the vacuum source that controls the throttles. Expecting to find a venturi vacuum system like the Ford products or manifold vacuum control as is common practice with four-barrel installations on General Motors products, learning that the vacuum source was the windshield wiper pump proved quite startling. Equally amazing is the fact that the vacuum operated throttles can’t be “feathered.” They have only two positions, wide open or closed.
I hadn’t grasped this point until I revisited this article and looked at the Oldsmobile service manual. I had previously assumed that the front and rear carburetor throttle linkage was held closed by intake manifold vacuum until the engine was at almost full throttle and there was little or no vacuum, in the manner of the cold-air intakes of the early ’70s. The J-2 system did the opposite, using vacuum to pull the throttle valves open as the accelerator was depressed.
This may seem counter-intuitive, since intake manifold vacuum decreases at wider throttle openings — if the J-2 setup relied on manifold vacuum, the other carburetors would only open at part-throttle, rather than with the accelerator floored. However, late ’50s Oldsmobile and Pontiac engines had a vacuum pump built into the top of the fuel pump to provide a vacuum source when manifold vacuum was low. Oldsmobile used two distinct types for air conditioned and non-A/C cars (illustrated below), but both worked the same way: The rocker arm that operated the fuel link would also compress the spring of the vacuum diaphragm (which I’ve highlighted in orange). As Navarro alluded, the main purpose of the vacuum pump was to keep the vacuum-operated windshield wipers from cutting out if the throttle was depressed too far. However, it also provided a vacuum source to engage the additional J-2 carburetors at wide open throttle.
Navarro continued:
This all-or-nothing arrangement seemed quite acceptable with the Oldsmobile that we drove despite preconceived prejudices. The test car was Hydra-Matic equipped so whenever the throttle pedal was floored at low speeds the transmission would downshift so there was no way of discovering whether a [carburetion] flat spot occurred. A stick shift model should prove rather annoying to a driver who makes a habit of flooring the throttle at low speeds. It would seem that a venturi vacuum operated throttle actuating diaphragm would be much more satisfactory for a stick shift.
Since 98.9 percent of 1957–1958 Oldsmobiles had Hydra-Matic, this was something of a moot point. Very few Olds buyers ordered, or wanted, manual transmission unless they were planning to go racing. (If you’re wondering, Oldsmobile did NOT yet offer a four-speed manual in this period.)
Even with automatic, Navaro felt the vacuum power control system demanded more careful driving technique:
Many drivers feel that progressive linkage completely eliminates the dead powerless feeling that results from too much carburetion at low rpm. They soon find out that a progressive arrangement doesn’t help if all of the carburetor throats are opened suddenly at a speed where two throats are sufficient. A progressive triple should receive enough throttle pedal depression at low rpm to open the center carburetor only. If a driver can accept this fact, he can live with progressive linkage and enjoy it; however, the drivers that can do this are in the minority.
Although Navarro was preoccupied with low-rpm overcarburetion, the flip side of the equation was that the additional carburetors cutting in could produce a sudden nonlinear increase in power that wasn’t always desirable in high-rpm driving situations, like powering out of a turn. Hydra-Matic was already troublesome in that regard because of its penchant for abrupt part-throttle downshifts, which could be a handful on a wet curving road. Suffice to say that a late ’50s Oldsmobile was an awful lot of car to get sideways, unexpectedly or not!
That neither Oldsmobile nor Pontiac was terribly concerned about that said a lot about how much U.S. automakers prioritized straight-line acceleration over all else. The J-2 system was intended at least in part for racing — at least until NASCAR abruptly banned multiple carburetors in February 1957 — but on the street, Oldsmobile obviously assumed the typical use case would be for an extra burst of passing power in combination with a Hydra-Matic kickdown. This engine was called the J-2 Rocket, after all, and no one expected rockets to go around corners.
A rocket it was too, at least by the standards of the time. At Daytona Beach in February 1957, Lee Petty used a J-2 powered Olds 88 to set a speed record of 144.928 mph in the flying mile …
… to the consternation of NASCAR co-founder Bill France Sr.:
The J-2 engine normal buyers could order wasn’t THAT fast, but it was rated at 300 gross horsepower (up 23 hp from the four-barrel engine) and 415 lb-ft of torque. Not all of that extra power was due to the carburetion — J-2 engines also had a thinner head gasket that increased compression from 9.25 to 10.5 to 1 — but it was nevertheless formidable, especially in the lighter Olds Golden Rocket 88, which was capable of 0 to 60 mph in well under 9 seconds.
The Automobile Manufacturers Association anti-racing resolution of June 6, 1957 brought an abrupt end to Oldsmobile’s participation in organized competition, but a 1957 Olds Golden Rocket 88 Holiday Coupe with the J-2 engine served as the pace car of the Pikes Peak Hill Climb on July 4. Pikes Peak president Lloyd Faddis then used a well-equipped J-2 88 to set a new record, completing the climb in 19 minutes, 1 second.
Oldsmobile offered the J-2 engine again for 1958, now rated at 312 gross horsepower. That might have been a conservative rating: The March 1958 Motor Trend tested a fully loaded J-2 Olds Ninety-Eight convertible, with a hefty curb weight of 4,953 lb, and still managed 0 to 60 in 8.6 seconds and the standing quarter mile in 17.1 seconds at 88 mph.
Barney Navarro felt that the J-2 setup had “great possibilities,” but the AMA racing ban and the recession seemed to sap whatever enthusiasm Oldsmobile originally had for the system. Also, non-racers who ordered the J-2 option weren’t necessarily very happy with it. In principle, the J-2 Rocket was fairly easy to tune, and with a light foot on the accelerator, it could return good fuel economy. However, the setup was intolerant of disuse: If the front and rear carburetors weren’t engaged very often, residual fuel deposits and grime could cause the throttle valves and linkage to stick or seize; the vacuum lines could also clog or crack with age. Another headache, not strictly related to the carburetion, was that the compression ratio was rather ambitious for contemporary premium gasoline. Motor Life noted significant auto-ignition in hot, dry conditions.
Although Pontiac continued to market Tri-Power through 1966, Oldsmobile dropped its triple carburetors after 1958, only returning briefly to multiple carburetion with the 1966 Tri-Carb option for the Olds 4-4-2. After that, GM corporate policy prohibited multiple carburetors except for the Corvette.
Oldsmobile doesn’t seem to have ever compiled any official J-2 production totals, but I’ve seen guesstimates as high as 2,500. That seems high to me, but as performance options went, the J-2 was relatively inexpensive, and an effective way to add a couple of extra stages to your Rocket Olds.
Related Reading
Car Show Classic: 1957 Oldsmobile Super 88 – Pressed In A Book (by Aaron65)
Curbside Classic: 1957 Oldsmobile Golden Rocket 88 Holiday Sedan – Silence is Golden (by Jerome Solberg)
1958 Oldsmobile Dynamic 88 – OLDSmobility In The Rocket Age (by J P Cavanaugh)
Curbside Classic: 1958 Oldsmobile Dynamic Eighty Eight – In Defense of The Baroque Beast (by Laurence Jones)
1958 Oldsmobile Dynamic 88: So Much Incredible Styling All In One Car! (by Stephen Pellegrino)
Love these articles that look at things from an engineering perspective, don’t give a toss about the company executives frequently mentioned on this site or the marketing aspect.
From admittedly limited reading, I get the impression that US tuners just added bigger carbs for more petrol and even more petrol, good to see the engineers were concerned about economy and driveability.
I do not see much about gas flowing the heads and intakes that dominated European thinking in the days of carburettors, though exhaust gas flow seems to be well catered for with header design.
What an excellent article yet again, Aaron. I have learned something new today!
Triple carbs must have been a service nightmare. All those linkages and vacuum servos would be difficult to set up and only provided a small performance gain. Also, can an engine of this size really make use of 850 CFM?
It wasn’t that much of a headache for normal service, which was the whole point: Only the center carb really needed adjustment or calibration, with the principal extra step being periodically making sure the little lever that actuated the vacuum switch was set correctly. Troubleshooting or repairing an actual problem was more complicated because there were a couple more linkages to check and because of the vacuum lines, which could be a hassle the way vacuum lines can usually be a hassle.
I was thinking the same thing about 850 CFM, maybe it was to a theoretical maximum more for marketing bragging rights.
Had a ‘70 Benz W114 250 “/8 Series” and it’s just dual Zenith 2 barrel carb set up was it’s Achilles Heal, especially with it abruptly stumbling when you most needed the power and running too rich giving mpgs in the low teens in town. It was old and tired by my ownership though.
Oldsmobile didn’t advertise the CFM capacity of the carburetors, that isn’t listed in the service manual, and the specifications just list the carburetor bore diameter in inches, so there were no “marketing bragging rights” involved. That’s the approximate rating of that size Rochester 2GC at 3 inches Hg, which I included here because that’s how carburetor sizes are more commonly understood today.
Two-barrel carbs were/are rated at 3 inches of mercury vacuum.
Four-barrel carbs were/are rated at 1.5 inches of mercury vacuum. Less vacuum reduces the CFM number; and is appropriate because the larger carburetors reduced intake manifold vacuum at WFO and higher RPM.
Rating a Tri-Power system based on the two-barrel rating protocol is silly, since the added flow capacity more closely resembles a four-barrel than a two-barrel carb.
850 cfm rated as a two-barrel is equal to 600 cfm rated as a four-barrel. And a 371 cid engine running a 600 cfm four-barrel carb is likely under-carbureted (depending on how high the engine is revved.)
The Corvette and Mopar Tri-Powers/Six Packs had ~350 cfm center carbs, with 500 cfm end carbs (rated as two-barrels.) This adds up to “1350” cfm…but correct that to the four-barrel rating, and you end up with about 950 cfm.
Why do race-cars have enormous carburetors? Because intake manifold vacuum kills power, so the carbs are sized to reduce intake vacuum at maximum revs to the lowest amount practical–you’d see two, 1100 cfm or larger four-barrel carbs on certain classes of racing engines.
It’s always been my understanding that GM usurped the need for tri-power from within, with the QuadraJet. Both systems had small primaries and relatively larger secondaries. I don’t think the Q-Jet got quite as big as 860cfm, but the secondary circuit was plenty big enough for most applications. The Q-Jet was orders of magnitude simpler than Tri-Power and therefore it was simpler to keep in tune.
Sort of, but dropping the triple-carburetor engines from everything but the Corvette was a policy thing, and DeLorean then canceled the Corvette LJ2 (3×2 454) for 1970 (allegedly) in the interests of clearing out expensive-to-produce limited-run options.
A very interesting article. I had heard of the J-2 option but assumed it was just another tri carb setup. Using the vacuum pump was a clever way to control the secondary carbs. Somewhere in all my junk I have a Chevy fuel pump with a vacuum pump on it and several Trico rebuild kits for the old vacuum windshield wiper actuators.
Multiple carb engines do need a little exercise to stay in shape. I don’t beat on my 4 carb Corvair but I do like to make sure the fuel in the fuel in the secondary carbs is always fresh!
Some beautiful rides here. The “blue/white”, tutone is the scene stealer.
This is one of the many cars I thought about buying at some point. I saw a ’57 Olds with triple two barrels for sale at the local antique festival at least 25 years ago. It was a sedan with a $3,900 price tag, but I couldn’t get over the fact that it was a sedan, and I didn’t inspect it closely enough to determine if it was a real J-2 anyway (or if there was any way to tell).
Even back then, I knew it was something fairly rare, so I sat thinking about it for quite a while.
Excellent article, interesting to see the differences between the Olds and Pontiac’s setup. While multiple carburetors were certainly considered advanced at the time, I wonder why Olds didn’t really push the envelope with Rochester Ramjet fuel injection. Seems it would have been a great fit in an Oldsmobile considering Olds’ reputation for technical advances.
Gibson Butler said Oldsmobile looked at the Rochester injection system, but “we felt we could do as much with carburetors as fuel injection. We didn’t have the precision machining capability to produce fuel injection components for a mass-production passenger car at that time.”
The ’57 is hands down better than the ’58 for styling.
Great article. Did not know about the Olds tri-power option.
Very much agree; the 3 1957 convertibles pictured are stunning!
10.5:1 compression ratio could have been a problem in 1957, before super premium gasoline was commonly available, if at all, in many parts of the country.