Dynaflow, Turboglide, Roto Hydra-Matic, and Other Early GM Automatics

SIDEBAR: Torque Converters
As we noted in Part 1 of this article, the torque converter, like the fluid coupling, was invented by Hermann Föttinger, then an engineer for the shipbuilder Stettiner Maschinenbau AG Vulcan, and is described in Föttinger’s original 1905 German patent application for hydraulic couplings. Early torque converters were first used for large ships and were later applied to trains, buses, and some stationary industrial engines. The first production automotive application of which we’re aware was the 1946 Invicta Black Prince, which used a Brockhouse Turbo Torque Converter transmission. The Black Prince was produced in very limited numbers through about 1950, but Brockhouse (today known as BH Transmission Services Ltd.) continued to build torque converter transmissions for various other applications, including tractors.

A torque converter is a type of fluid coupling and operates in a similar manner. The fundamental distinction is that while a fluid coupling can only function as a clutch (hydraulically connecting an engine or prime mover to an output shaft), a torque converter can serve as both a clutch and a transmission, capable of multiplying input torque where a simple fluid coupling can only transmit it. In fact, in some bus and railway applications, torque converters are used solely as transmissions, with a separate clutch — friction or fluid — serving to connect the converter to the engine.

torque converter diagram © 2013 kamasko; dedicated to the public domain by the creator under a CC0 1.0, resized and modified Nov. 2015 by Aaron Severson
A diagram of a simple torque converter with three elements: an engine-driven impeller, a turbine connected to an output shaft, and a single stator on a one-way clutch. While most automotive torque converters have used one-way (overrunning) stator clutches, early torque converters did not. The use of one-way clutches for torque converter stators was first patented in the mid-1920s by Scottish inventor Allan Coats. (Image: “Menic bez spojky” © 2013 kamasko; dedicated to the public domain by the creator under a Creative Commons CC0 1.0 Universal Public Domain Dedication, resized and modified (cropped, added English-language captions, changed file format) 4–5 Nov. 2015 by Aaron Severson


In Part 1 of this article, we explained that a fluid coupling consists of two torus-shaped bladed discs facing one another in an oil-filled housing. The engine turns one torus, the impeller (or pump), whose motion is transmitted through the oil to the other, driven torus, called the turbine, which in turn drives the output shaft (or, in a typical automotive transmission, the transmission gears). Some of the engine’s power is lost to heat, or slip, within the moving oil.

That’s the simple version. The more complex explanation is that the rotation of the impeller and its blades sets up two distinct flow patterns within the coupling’s operating fluid, each in a different plane: rotary flow (sometimes called circumferential flow), which is in the direction of engine rotation, and toroidal flow (sometimes called vortex flow) between the impeller and turbine. As the name implies, toroidal flow follows the inner curvature of the impeller and turbine tori, moving outward from the impeller’s inner hub to its outer circumference, then from the outer circumference of the turbine to its inner hub and back again.

While rotary flow is a product of the impeller transferring its rotary motion to the coupling’s incompressible working fluid, toroidal flow is mostly a function of inertia and pressure. Because the impeller and turbine are not mechanically connected (unless the coupling has a lockup clutch), any change in the momentum of either torus must be transmitted through the oil. Doing so requires overcoming the inertia not only of the turbine, but also of the intervening oil, different areas of which may be moving at different speeds. That doesn’t happen instantly or all at once. Instead, oil propelled by the impeller blades sets up a high-pressure area as the fast-moving oil collides with and transfers its angular momentum to the slower-moving oil around it. This creates a ring of fast-rotating oil that expands both forward and outward into lower-pressure regions of the coupling. When that oil reaches the turbine inlet, the oil’s forward momentum carries it along the inner curvature of the turbine torus toward the hub and then back toward the inlet of the impeller torus, thereby setting up a toroidal motion.

The forward component of that oil’s velocity — and thus the strength of the toroidal flow — is proportional to the difference between the pressure in and around the impeller and the pressure in and around the turbine, just as the speed of a gust of wind is proportional to a difference in atmospheric pressures. The pressure differential is proportional to how fast the impeller and turbine are rotating relative to one another. If the turbine is turning significantly slower than the impeller — either because the impeller has accelerated or the turbine has slowed down for some reason — the pressure within and around the turbine will also be significantly lower, so the incoming oil will have a greater forward velocity. Conversely, if the respective speeds of the turbine and impeller are very close, the oil’s forward velocity will be low.

Either way, the incoming oil also has substantial angular momentum. As it passes through the turbine, the oil will therefore apply torque to the slower-moving turbine blades. In keeping with Newton’s Third Law of Motion, the turbine blades exert an opposing reaction torque on the oil. That torque is again proportional to the difference in angular velocities. If the turbine is stationary or turning very slowly, much or all of the fast-moving incoming oil will essentially bounce off the turbine blades, reversing the oil’s rotational direction. Consequently, if the impeller is rotating clockwise, some or all of the ‘backwash’ of toroidal flow from the turbine will have a counterclockwise rotation and vice versa.

Diagram of rotary and toroidal fluid flow in a two-element fluid coupling at stall (© 2016 Aaron Severson)
An admittedly crude illustration (definitely not to scale) of the fluid flow in a two-element fluid coupling at stall — that is, with the impeller (red) rotating and the turbine (blue) stationary. The circular arrow indicates the direction of toroidal flow while the vertical arrows indicate the direction of rotary flow. Fluid leaving the impeller outlet has a rotary flow in the direction of impeller rotation, but by the time the fluid leaves the turbine outlet, its rotary flow has been reversed. (Author diagram)

When toroidal flow is particularly strong, much of the impeller’s energy will be wasted overcoming the opposing momentum of the returning oil. This ‘backwash’ flow diminishes as the turbine accelerates, eventually reaching a minimum (although never quite zero) when turbine speed is as close to impeller speed as frictional forces within the coupling will permit — for example, when cruising at a steady speed and load. From then on, most of the force generated by the rotation of the tori is directed centrifugally (i.e., radially outward) and oil flow is predominantly rotary. In essence, the moving oil now forms a high-pressure cylinder, rotating at close to engine speed, that connects (or couples) the impeller and turbine near their respective outer circumferences. This is known in engineering terms as “coupling stage.”

If you’re still struggling with the above explanation (and we can’t say we blame you), we can shorthand it like this:

  • Toroidal flow –> energy-wasting ‘backwash’ from the turbine.
  • If turbine speed is a lot slower than impeller speed, this backwash is strong.
  • If turbine speed is close to impeller speed, the backwash is low or minimal.


These flow patterns are present in both fluid couplings and torque converters, but in a fluid coupling, toroidal flow can only decrease efficiency. A torque converter, on the other hand, uses strong toroidal flow to advantage by adding reaction blades or guide blades to manage oil flow between the turbine and the impeller. Each set of reaction blades, commonly called a stator, is mounted on a disc carried on a fixed shaft. In most automotive applications, the stator disc is rotatable, but a one-way clutch allows it to turn only in the same direction as the impeller and the engine.

When the turbine and impeller speeds are nearly equal and toroidal flow is minimal, the oil’s angular momentum is already in that direction, so the stator just spins freely. However, if the turbine speed is significantly slower than impeller speed and toroidal flow is strong, the ‘backwash’ of oil leaving the turbine will try to turn the stator backward. This negative (reverse) torque locks the stator against its one-way clutch. The now-stationary stator blades exert an opposing reaction torque on the oil, again reversing its angular direction. In simpler terms, the moving oil bounces off the stationary reaction blades before reaching the impeller.

Diagram of rotary and toroidal flow directions in a three-element torque converter (© 2016 Aaron Severson)
Another very crude fluid flow diagram, this time of a three-element torque converter. Again, the vertical arrows indicate the direction of rotary flow. Note that oil still leaves the turbine outlet with a negative angular velocity (i.e., spinning opposite the impeller), but passing through the stator reverses the oil’s direction a second time before the oil hits the impeller inlet. (Author diagram)

Since the angular velocities of the returning oil and the impeller are now in the same direction, the returning oil’s angular momentum actually serves to accelerate the rotary flow of oil through the impeller. This effectively multiplies the engine’s input torque, enabling the impeller to apply more torque to the oil than the engine actually exerts by itself.


Unlike mechanical reduction gears, the multiplication provided by the stator (or stators — a torque converter can have several, redirecting the returning oil more than once) is not a fixed ratio. Since it requires strong toroidal flow, it varies based on the relative speeds of the turbine and impeller.

The difference between those speeds is greatest at the instant just before the turbine first begins to turn, a point known as stall speed — technically defined as the instant engine torque multiplied by the converter precisely equals the load on the output shaft. Not coincidentally, that is also the point of maximum torque multiplication (the converter’s stall ratio).

(We should note that stall speed can vary considerably depending on, among other things, the shape and pitch of the impeller and turbine vanes, the engine’s input torque, and the total load on the drivetrain at any given moment. Whole treatises have been written on this complicated subject, so we won’t get into it any further here, but suffice it to say that rated or nominal stall speed is an approximate maximum value.)

Once the turbine is moving, toroidal flow drops off and the reaction torque the turbine exerts on the oil gradually decreases, which causes the ‘backwash’ of oil exiting the turbine to hit the stator blades at a progressively shallower angle. Torque multiplication fades until eventually the torque the ‘backwash’ exerts on the stator is more forward than backward (net positive rather than negative), which allows the stator to unlock and freewheel. If there are multiple stators, each may freewheel at a different point.

torque converter stator illustration showing one-way clutch © 2016 Aaron Severson
While transmission diagrams and schematics use various forms of visual shorthand to represent the stator, this is approximately what a typical stator actually looks like from head on. The fuchsia center ring represents the outer race of the stator’s one-way clutch; the black arrows indicate the direction of allowed rotation. The stator blades are not flat — each vane is essentially an airfoil, like an aircraft flap or the blade of a propeller. (Author diagram)

With the stator(s) unlocked and spinning freely, torque multiplication ceases and the converter’s effective ‘gear ratio’ is close to 1:1. However, any abrupt increase in either engine speed or load (for example, if the driver floors the accelerator or if the vehicle goes up a steep grade) will cause the speed of the turbine to momentarily lag behind that of the impeller, which again increases toroidal flow within the converter. If that flow is strong enough, it will lock the stator(s), which causes torque multiplication to resume. The effect is comparable to downshifting a conventional transmission.

While a torque converter’s ability to multiply torque in this way is useful — in fact, the impeller and turbine blades of most torque converters are curved in opposite directions to maximize the fluid reversals — it’s not without drawbacks. One limitation is slippage, which is naturally high during torque multiplication. Another disadvantage is that with a torque converter, as with any continuously variable transmission, the relationship of engine speed to road speed is no longer linear. That can result in brief but sometimes frustrating delays where engine rpm soars with little or no immediate effect on the vehicle’s momentum.


Add a Comment
  1. Hey,how come you can yack all day long about this ones gearset setup,or that ones turbine combination,but no illustrations???
    Just because you can picture the entire mechanical world with words doesn’t mean the rest of humanity can.
    Pictures Please!!!!

    1. Um, no “Thank you for an awesome article and site?”

      There is an illustration of a Turboglide and it’s hardly fair to expect Aaron to write an great article about the development of the automatic AND delve into all the technical details. He does to a degree, but that’s not the overwhelming emphasis of the site, as far as I understand it.

      How about Googling “Turboglide,” “Dynaflow” or “Powerglide?”

    2. There are now diagrams! I’m not a technical illustrator by any stretch of the imagination, but you can at least get a sense of how these things were laid out.

      1. There’s a site here that has a diagram of an overhaul of the controlled coupling hydra-matic. I can really see why GM wanted to get way from this design. Although today’s ZF 8 and 9 speeds are probably worse, but then half of the world industry is sharing the development costs for these.

        1. …And yet, they were damn near indestructible. We had a ’58 Pontiac that took a lot of punishment in the snow, yet worked without any issues, other than a small oil leak, until I had to sell it in late 1964.If I remember correctly, it was cast iron and weighed around 225 lbs.

          1. The ’58 edition weighed about 240 lb. GM was able to trim about 10-11 lb for 1960 by slimming down the case a bit.

  2. In the photo of the Hydra-Matic shift quadrant in the ’50 Olds 88, is that an aftermarket turn signal unit? If so, it’s a reminder of how times have changed! I understand that at that time, a heater was an option on many cars.

    1. I believe turn signals were standard on Oldsmobiles by 1949, at least on DeLuxe models. I’d need to find somebody with an Olds dealer book from that period to know for sure, but my information suggests they were standard fit.

      Pretty much everything [i]else[/i] was at least technically optional at that point, including oil filters, wheel covers, hood ornaments, windshield washers, and (at least until after the war) reversing lamps. Heaters didn’t become standard even on Cadillacs until almost the mid-fifties, and they weren’t standard on cheap cars for another decade after that. Very few cars were built without a lot of these items, but they weren’t included in the list price for many years.

  3. At least they did not charge extra for chrome after the war.

    I remember seeing a ’50s car ad that mentioned the [i]reverse[/i] gear was an optional extra. On the other hand many cars (particularly British) came with leather seats only because it was cheaper than vinal.

    1. I don’t know of any cars that late that didn’t come with a reverse [i]gear[/i], although reversing [i]lamps[/i] were still extra on many inexpensive cars at that point. Turn signals, as well.

    2. Just as well they didn’t charge extra for chrome.
      The ’58 Buicks & Olds would have cost a small country to buy.

      Back on topic, thank you once again for an
      entertaining read.


      1. Well, in essence, they did charge extra for the chrome, though fortunately not by the pound. On most cars of that era the amount of brightwork was tied to the trim level, and naturally the higher the trim level, the higher the price. Beyond that, there were often extra-cost dress-up packages (either factory- or dealer-installed) that primarily consisted of additional chrome trim. Such things didn’t really disappear from American options lists until the rise of Japanese-style tiered equipment packaging quite a few years later.

  4. Thanks for a great website and particularly for the GM transmissions articles. Every article I’ve read has been complete, accurate, and very interesting.

  5. Thank you for the automatic transmission article(s) on GM. Finally, someone has accurately chronicled the myriad development story for us.
    Your site is a valuable and entertaining resource – keep up the great work!

  6. This brought back some memories – I remember when I first got my license driving my Dad’s ’65 Olds F-85 with Jetaway and those 1-2 shifts at about 70mph if you held your foot in it. I have a question – I have an childhood memory of an early 50’s vehicle ( think it was a Chevy ) with a “Torque-Glide” logo on the trunk lid instead of “Power-Glide”, but that can’t be right, can it?

    1. Chrysler had a number of semi-automatics in that period with a variety of bizarre names: Gyro-Torque, Fluid Torque Drive, Fluid-matic, Fluid-Drive, and Plymouth’s Hy-Drive. Maybe it was one of those?

    2. Actually, from 1965 up, the F-85, Buick Skylark, and Pontiac Tempest all utilized the newly available Turbo-Hydramatic 300, which in essence was the same thing as a Powerglide, but with non-interchangeable parts. Early versions had variable pitch and a rear pump. It was with the advent of these new automatics that the shift indicators from that time forward would read P R N D L.

      1. The latter point is correct, but the rest is not. As the text explains, the two-speed transmission used on 1964-on B-O-P A-bodies is not Powerglide, although they’re similar in many respects. Although the two-speed (which Buick called Super Turbine 300) was manufactured by Hydra-Matic Division, it was not called Hydra-Matic. (I know the source you’re looking at, and it’s incorrect.) The three-speed Turbo Hydra-Matic became optional in 1967 with the big engines only and was later supplemented by the medium-duty TH350. The two-speed remained available on low-end models into the early seventies.

  7. anyone have a diagram of the dual path? It stopped shifting from low into second and I found a spring in the bottom of the pan. Where does it go?

    1. Sorry, I’m not qualified to give repair advice. You might try seeing if your local library has a service manual for it — I was able to find a copy of the Pontiac dual-coupling Hydra-Matic shop manual that way.

    2. Try this… as good an explanation of your problem as I’ve ever understood: https://www.youtube.com/watch?v=rLDgQg6bq7o

      1. He talks about your differential girdle spring at starting at ~1:10. It’s supposed to be hooked onto the upend of the gramys.

  8. [quote=steve dill]anyone have a diagram of the dual path? It stopped shifting from low into second and I found a spring in the bottom of the pan. Where does it go?[/quote]If you could provide a picture of the spring, I could look it up in my various manuals and give you an answer.

  9. I have a 62 Buick,Skylark,with the dualpath Tranny.the trans is in direct drive,only goes foward,no neautral,park orreverce,is thier a fix for this.

    1. Can some one HELP.
      I have a 1962 Olds Cutlass F 85, Auto Hydro Matic floor shift.
      I had the transmission rebuilt 3 times already.
      and the problem is that when the car warms to operating temp
      it starts to jerk and gos into neutral. it clears once i accelerate.
      RPMs Are normal. trans just dosnt stay in low gear when moving at 10mpg or at a stop. Thanks- Robbe California

      1. @Robert: I’m afraid I’m not at all qualified to offer repair or troubleshooting advice — sorry!

  10. this article was great. It answered my question as to why the 52 Super I just inherited doesn’t shift….that would be because it isn’t made to shift automatically….I read a blog online saying
    1952 Buick – the slowest car I ever loved….so true!

  11. anybody know where I can buy the flexible black fresh air vent tubes? Darn Mice

  12. Are the dyno-flow and power glides enter change able? With other motor?

    1. Well, there’s an old saying to the effect that you can make anything fit if you have a big enough hammer. I honestly don’t know how much trouble would be involved in interchanging them, but since they were never designed to be used behind the same engines or in the same cars, I imagine it would take some work.

      At one time, Buick Nailhead engines were popular with drag racers, so if you were asking this question in, say, 1964, there might have been aftermarket kits to mate an older Buick V-8 with a beefed-up Powerglide. (Some drag racers used Powerglide because it consumed relatively little power and they didn’t need a lower first gear.) Today, I suspect you’d have more luck finding some way to put in a Turbo Hydramatic. I’ve never looked, though.

      This is a question that would probably be best put to a performance transmission manufacturer or a shop that specializes in parts for older transmissions.

    2. No the dynaflow and the powerglide are not interchangeable. the dynaflow is about three times heavier and will not fit up to any engine that was made for the powerglide. The powerglide came in two models first being the cast iron model that was used through 1954 then the aluminum powerglide after that. both very good transmission, and easily rebuildable.

      1. The earliest Powerglide is very similar to the early Dynaflow, although I doubt they’re easily interchangeable. As the revised text explains, Powerglide had several phases: the early dual-impeller variety, used through 1952; the later iron-case version with a three-element converter, used, with various evolutionary changes, from 1953 to 1962–1963; and the late aluminum-case version. The aluminum Powerglide (for RWD cars — all Corvair Powerglide units had an aluminum case) was introduced for some models in 1962 and for others in 1963.

  13. chevy had 2 auto transmissions in 61and62 1 was a turbo glide the other was –glide that changed by fluid. there was no gears in the trans. on the gear selector was P R D G G was for grade as going up a hill. what was the name of that trans?

    1. The two transmissions were Powerglide and Turboglide. Powerglide was the familiar two-speed-plus-torque-converter Chevrolet automatic, while the transmission you’re thinking of was Turboglide, which is described in the text.

      The G position was for Grade Retarder. It was intended not for climbing hills, but for descending them; it was supposed to mimic the effect of engine braking, of which the Turboglide otherwise didn’t allow very much. The Grade Retarder was not useful for acceleration or hill climbing, although some people had problems because they assumed it worked like the Low position on Powerglide, which was definitely not the case!

  14. Re read this as a refresher on the development of the automatic. Thank you again. Your site is an invaluable resource and I cannot thank you enough for doing what you do.

  15. Thank you for your clear and concise explanation of Dynaflow, and how it differs from the other two GM automatics. As we were a “Buick family,” the innate superiority of Dynaflow was never a question; it was an article of faith. I remember the feelings of incredulity and betrayal I felt when I was told for the first time that Dynaflow was “Just Powerglide with a different name,” and that Hydramatic was obviously better, because Olds and Cadillac used it. You have restored my faith in Dynaflow.

  16. We have recently inherited a 53 Roadmaster. I think it is an early model serial #26854377 because the 322 nailhead has a weighted pully instead of a rubber loaded harmonic balancer. The Dynaflow is now in the transmission shop and we are finding puzzles. According to the shop manuals the 53 should be the new twin turbine with only 1 pump and one stator. This trans has the words “twin turbine” cast into the bellhousing. But inside it has 2 pumps and 2 stators. Do we have a transitional factory job or a trans shop hybrid? Was the change made to save money (fewer parts) or to improve performance? Will our new Roady rise and fly?

  17. Fascinating info.

  18. Hi can any body help me
    I have a 1958 Buick Road Master fitted with a Dynaflow Flight Pitch
    gear box can any one tell me where i can get spares for the gear box
    and will ship them to England

    1. I’m not able to help with technical issues or buying parts — sorry!

  19. Just wanted to say this is a great article. I started out looking to find the difference between the hydra-matic dual range and the strato-flight and wound up learning a lot more.

  20. The article refers to the Hydramatic’s jerkiness. Actually, many Hydramatics were so smooth that you could not even feel the shift; you could just hear the drop in engine speed. I remember in 1959 riding in a 1949 Lincoln with Hydramatic; it accelerated quickly and so smoothly that I could not feel the shifts. The same was true with some other cars with Hydramatic in which I rode, including a 1950 Pontiac, and those were all before GM introduced the Hydramatic with the second (controlled) fluid clutch in 1956. On the other hand, I rode in a 1953 Cadillac with had very firm shifts.

    The downshift resulting from flooring the accelerator were another matter; they were always accompanied by a mechanical clunk.

    1. The issue with the original Hydra-Matic was that because its shifts were mechanically complex (particularly between second and third, which was the most complicated sequence), its smoothness depended a great deal on how well the bands were adjusted, the condition of the transmission fluid, and other maintenance- and condition-related factors. If everything was perfectly adjusted, it would be quite acceptably smooth (particularly by the fifties, by which time GM had made a lot of minor refinements). If not, it would throw off the shift timing just enough to make the shift jerky, albeit not necessarily enough to really impair the transmission’s function. I suspect a lot of owners who complained to their dealers or mechanics were told, “Ehh, they all do that.”

      Even some of the engineers who originally designed the Hydra-Matic thought it was too complicated for its own good, which is why they subsequently got into the torque converter automatics, which didn’t shift at all. The original Dynaflow was very much the antithesis of the Hydra-Matic in a lot of these respects.

    2. My experience with Hydromatic cars was that they were fairly smooth in shifting. PowerGlide cars had a very pronounced jerk when shifting. When my city purchased GM buses in the sixties, the Hydromatic was very rough when shifting with an easily heard lowering in engine sound as speed increased.

      1. The difficulty with making blanket statements in this area is that each of these transmissions was around for a long time in several quite distinct versions, not all of which felt or acted the same.

        As the text explains, early Powerglide cars did not provide any automatic shifting in Drive, relying on torque converter multiplication exclusively. Powerglide was revised in 1953 to start in first and shift automatically to second. So, early Powerglides (or Dynaflow) were smoother than even a well-adjusted early Hydra-Matic, albeit not especially quick or efficient. After that, there were early (iron-case) and later (aluminum-case) Powerglide transmissions, tuned in different ways for different engines.

        Similarly, the early (1940 to 1955) and late (1956-1964 dual-coupling) Hydra-Matics were significantly different mechanically — albeit still related — and felt quite different.

        So, while it may sound pedantic, it’s important to qualify statements like, “X was smoother/rougher than Y.”

      2. Those GM buses had a 1 speed automatic Allison transmission. Great roaring noises as the variable torque converter changed pitch and allowed the bus to gradually accelerate to 25 mph, then an almighty clonk as the torque converter was locked-up with a mmm-uhh-mmm vibration that gradually settled down as the engine bounced up and down on its mounts. Crude or what! Engine note and speed decreased at point of lockup.

        I blame those buses, their braying, outlandishly noisy two-stroke GM diesels and the pathetic transmission for ruining the quiet of our city at night when introduced. Went to London for grad work in 1969, and it was obvious that a AEC 4 stroke diesel packing all of 120 hp and four speed preselector gearbox not only got a double-decker bus going from stop much quicker than a GM bus, it was at least 10 times quieter doing it.

        Speaking from my point-of-view as a mechanical engineer. In those days as a student I had to ride buses and had a keen interest as to why the GM was so unrefined and the engine so noisy. No domestic competition would be my guess.

        1. Noisy or not, I loved those old roaring GM buses, when in “hydraulic drive” mode. That mode would seem to be not very fuel-efficient; a 4-speed pre-selector as you mention, should indeed have been more fuel-efficient (as well as quicker, as you mention). I have read that a later version of this Allison transmission arrangement actually had a second gear, making for a true two-speed, plus lockup in high. I cannot confirm that, though.

  21. I’ve heard a story about the Hydra-Matic, as follows:

    Supposedly Rolls-Royce acquired a Hydra-Matic for evaluation. They liked it but thought one particular part had too rough a finish. When they fabricated a smoother-finished version of the part and incorporated it into the reassembled Hydra-Matic, the transmission didn’t work. True, or urban legend?

    1. I’ve heard that story in regard to the Turbo Hydramatic (not the original), which Rolls-Royce also built. The way I’ve heard it is more that they tightened up the tolerances, which didn’t necessarily work out well. I don’t know if it’s true or not, but it’s not implausible. There’s an analogy to be made with pistols, where getting everything “tuned” to tight tolerances improves accuracy, but makes the action less tolerant of dirt or debris. (This is why police and military sidearms are not built like target pistols.)

      1. I am reasonably certain that while Rolls Royce licensed & built in England the original HydraMatic, it imported the Turbo HydraMatic 400 from GM in the states.

        1. You’re correct; my previous comment was based on a point I was only half-remembering. They did import them, but asked for higher-than-standard tolerances.

  22. Thank you for this very complete summary. I have been curious about these transmissions for quite some time, and this is quite helpful. Your research is impressive, as is the writing.

  23. The main problem with reliability of the Slim Jim was the weakness of the front oil pump cover; they cracked. An improved pump with webbing on the cover was designed to replace failed units. RHM 375 Model 10’s made at Willow Run ceased in 1962. The THM 350 signalled the beginning of a long slide toward mediocrity by GM.

    1. I have to wonder if the Roto Hydra-Matic’s various weaknesses, including the propensity for leaks and the issue you describe, were exacerbated by the very high operating pressures. As mentioned, the RHM’s operating pressures were substantially higher than the earlier dual-coupling HM’s, which is a lot of added stress to put on what was still fundamentally an adaptation of the earlier transmission.

      I’m not sure how your last statement follows. The THM350, which didn’t arrive until five years or so after the RHM expired, was effectively a replacement for the Powerglide and Super Turbine two-speed automatics, and in that sense were an improvement in most respects. (There have been some harsh criticisms of the later TH200, but that’s a different story.) Since most rivals had long since offered three-speed automatics for most engines, the TH350 was also arguably overdue. It wasn’t quite as heavy-duty as the TH400, but it wasn’t designed to be, trading off some torque capacity for lighter internals and lower power consumption.

    2. I would disagree; I had very good luck with the THM350 in my 1973 Nova 350; it reached 185,000 miles, with no issues other than some fluid leakage. Shifting was still quick and firm. I have not heard of a lot of issues with this tranny.

      1. The lighter TH200 has gotten a pretty bad rep, but I’ve never heard anything particularly bad about the TH350.

  24. I had a 1949 buick super with dynaflow, four door. It averaged about 8 mpg. It took everything I earned as a super market clerk to keep the transmission running, most repairs were $300 to $400.

  25. Studebaker developed their own automatic and introduced it in 1950. Ford wanted to license it, but Studebaker turned them down. Studebaker started using the Borg Warner later, when manufacturing costs of theirs got too expensive. If I recall, a European manufacturer bought the tooling, and used it in their own cars?

    1. I believe the Studebaker automatic became the basis of the Borg-Warner DG, which was used on a number of British and European cars of the ’50s.

  26. Thanks so much for the great overview.

  27. Great job like the article ? would you have any info on the olds roto hydromatic . I have a 62 any m having some small issues
    Thank you Mike

    1. I’m not able to help with any kind of troubleshooting or repairs, sorry!

  28. Thanks again for a great resource. I find myself returning to it for a periodic refresher when a relevant vehicle appears. (Today’s is a 1961 Buick.)

    1. Thanks, Ed! I’m actually in the process of updating this article as I recently did with the Hydra-Matic story, to fix some minor factual glitches, clarify the technical details (which is a major project, let me tell you), and add some new info.

  29. All this effort and expense just so drivers don’t have to clutch and shift? Turns out major beneficiaries of automatic transmissions are texters. Who cause many of the accidents on the road now!

    1. Given the timeframes of the respective inventions, I would said that definitely constitutes an unanticipated side benefit…

  30. I believe that the first automotive use of planetary gears was in the Model T. As I recall, you would press down on one pedal to get the car going (1st gear), then move the gear lever and let the pedal up for high gear. It wouldn’t have taken much to use a servo to make these motions and a combination speed and throttle position sensor to determine when to make them. That could have been an early two speed automatic. The original Hydra-Matic is just a more sophisticated, four-speed version with a fluid coupling, isn’t it?

    1. That is how a Model T transmission worked, although it was not the first automotive application for epicyclic transmissions; a number of other cars, including Cadillac, used planetary gears before the Model T was introduced. (I’m always leery of pointing to anything as The First just because it’s often wrong unless you add a lot of qualifiers — a surprising number of innovations were tried or at least considered decades earlier than you might expect, even if manufacturing or machining technology wasn’t up to making it work.)

      It is certainly true that Henry Ford remained a stubborn proponent of planetary gears, which he continued developing for tractor use even after he was persuaded to allow a conventional gearbox in the Model A. (One of the engineers who worked closely with him in that, Howard Simpson, went on to design and patent the “Simpson gearset,” licensed by many other manufacturers including GM and Mercedes-Benz.) However, the Model T certainly wasn’t automatic and it would have needed some other control mechanism to execute shifts without driver intervention.

      As Part 1 of the Hydra-Matic article touches on, there were various efforts to do that, many of which used planetary gears because the brakes and clutches could be controlled hydraulically, electromagnetically, or by some other remote mechanism. So, there is a parallel, but it only goes so far and there were a lot of steps in between.

  31. Minor glitches: The TH 400 was used by Buick AND CADILLAC in 1964. The variable-pitch stator was not used on the TH 400 in ’64, but was available on some Olds, Buick, and I guess Cadillac vehicles from ’65–’67. Ironically, the variable-stator design was used on the “big” engines in the more-expensive cars; the small-blocks and six-poppers needed the torque boost more than the big-blocks.

    For the record, the ’64 TH 400 uses a substantially-different valve body and in-case fluid channels than the ’65-newer TH 400. The valve body of the front-wheel-drive version (the TH 425) uses the ’64-style system. Therefore, a “shift kit” for a 65-newer TH 400 won’t fit a ’64 TH 400 or the TH 425, but a shift kit for a TH 425 will work in a ’64 TH 400.

    The TH 350 was actually a joint development of Chevrolet and Buick engineers, both divisions looking for replacement of the two-speed transmissions they were currently using (Powerglide and Super-Turbine 300) with the resulting “350” produced by the Hydra-Matic Division.

    1. Thanks for the notes — I’m aware of both of the errors you note and they’ll be fixed in the extensive revamp of this article on which I’m currently working. (See the most recent post for details.) I won’t be getting into a detailed discussion of Turbo Hydra-Matic in the revised version, which is already monstrously long and has been eating my brain for months.

      TH400 wasn’t offered on all 1964 Cadillacs, incidentally; it was initially available only on De Ville, Eldorado, and Fleetwood Series Sixty. I wasn’t aware that the TH425 used the original valve body pattern, though. (I know generally how the TH425 is laid out, but I can’t say I’ve ever looked at its hydraulic control layout.)

      1. Okay, the revision is now complete and those corrections are now reflected in the text.

  32. Great, great job Aaron! That was awesome, and I was glad to help

  33. I think I can appreciate how big an undertaking revising this article has been. Hats off to you Aaron, for possibly the best explanation of early GM automatics expressed in laymans terms.
    GM didn’t swallow its pride and licence the Simpson system and tried to develop practical cost effective alternatives in its various divisions until the ’60s. Seems a classic case of corporate wilful blindness until we remember hindsight is the only exact science.
    In 1966 “Motoring Which?” the UK’s equivalent to “Consumer Reports” published a test of three 1.5 liter automatic British sedans, a Ford, a Hillman, and a Vauxhall. Vauxhall is the UK subsidiary of GM. The Vauxhall had a GM two speed transmission, the others both used a Borg Warner 35 three speed. They noted that they all had slightly worse performance and fuel economy than their stick versions, but the Vauxhall also had a big gap in its performance between 35-50 mph just when it was most needed. It was likened to driving a stick four speed using only second and top gears. The article also mentioned “Consumer Reports” had harsh words for GM cars using two speed transmissions, I’m guessing Ford, Chrysler, and AMC had all switched to three speed transmissions by then?.

    1. By 1966, I think Ford’s two-speed Fordomatic may still have been available for the cheapest U.S. Falcon models — I would have to double-check, as it may have been dropped after 1965 — but otherwise the other U.S. automakers all had smaller three-speed units for their low-end cars by then. (The light-duty TorqueFlite was one of the big pluses of Chrysler’s compact Plymouth Valiant and Dodge Dart, in my view.)

      The general attitude of GM engineers in this era was that a two-speed torque converter automatic was a perfectly reasonable substitute for a three-speed manual transmission while being simpler, lighter, and cheaper than a three- or four-speed automatic. The latter was of course perfectly true and the former was at least a supportable position. I also suspect some of the transmission engineers were soured a bit by experience with the small three-speed Hydra-Matic, which was little better than a decent two-speed automatic. (The transition from the smaller three-speed unit in the 1961–1963 Y-body Oldsmobile F-85 to the two-speed Super Turbine 300/Jetaway in the 1964+ A-body equivalent was certainly no great loss and probably an improvement in some respects.) On the other hand, by the mid-sixties, very, very few Americans still bought three-speed manual transmissions and it was certainly clear that a good three-speed torque converter automatic was considerably better than the best two-speed. It was also a bigger deal for non-U.S. cars and the later U.S. ventures into the “subcompact” [sic] realm, since having 3 or more liters’ displacement to fall back on masks an assortment of deficiencies.

      I don’t think GM was willfully blind so much as having a fair bit of (understandable) inertia. As this article should hopefully make very clear, GM had invested an absolutely staggering amount of money in automatic transmission development and engineering, accumulating a towering stack of basic patents. The tooling alone was a king’s ransom — in the early fifties, Detroit Transmission built more Hydra-Matics each year than the entire contemporary British auto industry built cars, and that wasn’t even GM’s only automatic! So, a reluctance to completely reinvent the wheel or to unnecessarily license outside technology isn’t difficult to understand. (To be clear, what GM licensed from Simpson and Simpson’s estate was a specific arrangement of planetary gears, not a complete transmission. Part of the reason that arrangement ended up being so widely licensed was that Simpson, like Pol Ravigneaux a decade or so before, had patented many different variations that there was no getting around them.)

  34. Great job Aaron, you’ve outdone yourself. I enjoy coming to this site to expand my knowledge. It’s a fantastic resource indeed. I also enjoy your clarifications on “Curb Side Classics” and can faithfully know that any input you offer will be well reasoned and researched. You offer a great service to like minded Auto Industry nuts.

Leave a Reply

Your email address will not be published. Required fields are marked *

Click here to read our comment policy. You must be at least 18 to comment. PLEASE DON'T POST COPYRIGHTED CONTENT YOU AREN'T AUTHORIZED TO USE!
Except as otherwise noted, all text and images are copyright © Aaron Severson dba Ate Up With Motor. (Terms of Use – Reprint/Reuse Policy) Trademarks referenced herein are the property of their respective owners and are used here for informational/nominative purposes.