(first posted 9/23/2014) In this day of nine-speed automatics and CVTs, a two-speed automatic transmission may seem downright antediluvian. Not infrequently, comments here express shock, surprise or disdain for the fact that for decades, many of America’s most popular cars— even the mighty Corvette—were equipped with two-speed automatics well into the mid and late sixties. Part of that may stem from the inevitable comparison with the standard three-speed manual transmission of those times, as if the two-speed automatic was grossly handicapped by missing one of those precious three gears. Actually, in many key respects the two speed automatic was very much the equal to the three-speed manual, and then some. Unfortunately, that isn’t necessarily saying much.
The early development and design of automatic transmissions was all about convenience and smooth operation, which may well not be the key considerations for enthusiasts and lovers of stick shifts. So before we start comparing apples to oranges, let’s make one thing clear: a three-speed manual will generally (but not necessarily in every parameter) be faster accelerating and more economical in the hands of an engaged driver and proficient shifter then a two-speed automatic. Since the three-speed manuals of yore inevitably had no syncromesh on first, and were rather balky to shift, that excluded a large segment of the population, especially women. Yes, there were exceptions, and women who wanted or needed to drive made do, but not always happily.
Cars back in the 20s, 30s and 40s were also geared much lower (higher numerically), due to slow-turning engines and slower speeds. That allowed drivers to start in second gear from starts, depending on conditions and the car. On large and powerful luxury cars, that was essentially the norm, thanks to their huge engines with enormous flywheels which would even chug a away from a start in top gear without complaint. Luxury car automakers bragged about the ability of their cars to accelerate from extremely low speeds in top gear without any vibration or complaint. Shifting was something to be endured, but as little as possible, please.
One of the specific benefits touted by Borg Warner’s new 1934 overdrive was that it could eliminate much in-town shifting, by starting in second, with the unit up-shifting to second-OD automatically once a certain speed was attained.
We can’t do a complete historical review of all the early automatic transmissions here today, so we’ll focus primarily on two main decidedly different approaches: GM’s Hydramatic and the Powerglide/Dynaflow (Chevrolet/Buick). The Hydramatic utilized a fluid coupling with four gears, and the PG/DF had a torque converter with two gears, only one of which was normally used in their early years. Many who read that the Hydramatic had four speeds back in 1938 wonder what happened? Why did subsequent GM automatics revert to one or two geasr?
Unlike essentially all modern automatics, the Hydramatic didn’t have a torque converter, which multiplies torque output, roughly comparable to more lower gears in a transmission. It had a simple fluid coupling (above), which has only two elements and acts somewhat comparably to a mechanical clutch, or more like the centrifugal automatic clutches often used on small scooters and power equipment: when the engine is revved up a bit, the input turbine turns the output turbine via the oil, but there is no torque multiplication.
Strictly speaking, when a mechanical clutch is slipped there is some torque multiplication; with the engine revved up higher, there is even more, which is what allows one to start in second or even third gear on a manual, as long as one doesn’t mind the smell of the clutch frying. But that’s not the case in the fluid coupling, which takes up rather quickly even at modest rpm, necessitating those four gears in the Hydramatic. First gear ratio in the Hydramatic is 3.66:1, a whole step lower than the typical first gear in a three-speed manual (2.5-2.9:1, typically). Second gear in the Hydramatic is 2.53:1, which essentially equals the first gear on manuals. Third gear (1.45:1) approximates second gear, and fourth gear is direct, unlike modern four-speed overdrive automatics.
Effectively, the Hydramatic’s 2nd -4th gears replicate the three gears of a three-speed manual; first gear was necessary to get it off the line, for lack of any substantial clutch slippage or torque multiplication. The Hydramatic felt very different than modern automatics, with a very mechanical feel and abrupt shifts, with the shift from first into second typically occurring part-way through an intersection from a standing start.
A torque converter, which has three (or more) elements, does of course multiply torque, the degree dependent on its specific design. In the case of Buick’s Dynaflow and Chevy’s Powerglide, torque multiplication at stall speed (maximum engine speed with brakes on) was in the area of 2.25 – 2.6:1, or roughly comparable to first gear on a manual. That explains why the original versions of both the Dynaflow and Powerglide were set up to function as a one-speed transmission that started in High (direct) gear, unless Low was manually selected for extreme situations or those wanting a quicker get-away, at the risk of premature transmission life. Even with one effective gear, torque multiplication at take-off was roughly comparable to first gear in a manual.
Yes, the PG/DF were “slushy”, hence the origins of the term “slushbox”. Throttle response was poor and acceleration was slow, since the engine wasn’t able to utilize its full horsepower potential be revving up through the gears. But for plenty of Americans, that was just fine, if it eliminated the hassle of shifting. And compared to the distinctly jerky-shifting Hydramatic (“Hydrajerk”), the torque convertor transmissions were perfectly smooth, especially the early ones that never shifted at all.
The one-speed operation on the Powerglide turned out to be too pokey and inefficient, and in 1953, low gear was employed automatically at start, upgrading it a full-fledged two-speed. Low gear on the PG was 1.82 or 1.76, which corresponded very closely to second gear on manuals. With its torque converter adding substantial additional torque multiplication to that, the PG had much greater torque multiplication at start than a manual (almost 5:1), and in accelerometer measurements, a PG-equipped car could be faster off the line. So the lack of a “1st gear” really wasn’t really a significant handicap compared to a manual.
Where the PG and other two-speed automatics are generally faulted is the lack of an intermediate gear, especially when passing. The PG’s Low gear’s maximum speed was typically between 50 and 70 mph (or more), depending on engine and rear axle ratio; the lower end corresponds to the smaller engines; the top for the biggest V8s. This meant that 0-60 runs were often completed in Low gear only, and PG-equipped cars were typically a second or two slower in that test, compared to a three-speed manual.
But the oft-derided lack of an intermediate or “passing gear” is precisely the same shortcoming with the three-speed manual, whose second gear also peters out at the same speeds as the Powerglide’s first gear. They both suffer from the same primary shortcoming.
In truth, the three-speed manual is clunky: its column shifter is typically on the balky side, and first gear needs to be low enough to cope with starting on steep hills, which means second has to be fairly low too. The result is a big and painful hole between second and third gear. And at modertae speeds, the Powerglide’s torque converter can still provide some torque multiplication in high gear.
On a car with a fairly big and lazy V8, the manual’s gap between second and third is not much of an issue, as third can be used even in puttering down city streets at 25-30 mph and there’s enough power to chug up grades at speed with a load. But that doesn’t work so well with the six cylinder (or small V8 in a big car or truck), which is what was typically teamed up with the three speed manual. Third is often too low for around-town use, and second is painfully high. And in hauling a load up a freeway or highway grade, a truck or loaded wagon is really handicapped for the same reason.
In 1993 I used my ’66 F100 with the 240 six and three-speed manual hooked up to a trailer to move our vast horde of
crap accumulated treasures on a number of trips from the Bay Area to Eugene, via I-5. There’s several mountain passes on the way, and I had to use second gear both going up as well as down (to save the drum brakes), which meant about 35 mph or so, if I didn’t want to feel hear the engine screaming. That borders on being a hindrance on a freeway.
Fortunately, as it turns out, the Ford transmission crapped out on the very first trip (not catastrophically), so I replaced it with a HD Warner T89 three-speed and overdrive. The difference was huge; not just the ability to cruise at 60 @1900 rpm in third-OD, but more importantly, the missing gear between second and third was now there, in second-OD. It allowed me to scoot up the passes (and down) at a faster 45-50 with the engine running at reasonable rpm. Second-OD is the gear I use most around town; its perfect for the 25-30 mph zones.
My real point in this comparison? Three-speed manuals are intrinsically deficient, lacking a key intermediate gear, never mind the lack of an overdrive for highway speeds. A two-speed automatic also has that hole, but it’s actually masked by the torque converter to some extent.
Yes, an automatic’s intrinsic hydraulic losses means that PG-equipped cars were slower in broader acceleration measures (0-60; 1/4 mile) and somewhat less efficient, but again, that’s a moot point in the hands of a person that hates to shift. And the enduring sales success of the Powerglide seems to bear out that most typical drivers felt well-enough served by it. During the early-mid sixties, when both Ford and Chrysler were offering three-speed automatics, Chevrolet’s market share against them both surged, despite the missing gear.
The three-speed manual’s limitations were also evident in terms of performance, which helps explains the popularity of Overdrive among those in the know, like owners of tri-five Chevies: for one of these cars to be effective both at the strip and for daily driving, they would order a higher-numerical rear axle ratio (4.11 or more), which made the three gears usable at the drag strip, yet highway cruising was quite tolerable in OD. A four-speed in lieu of the real thing.
Ironically, it was Chevrolet that first made a four-speed manual widely available in passenger cars, on their 1959 models (not with the six). Teaming a wide-ratio four speed with a relatively long rear axle (low numerically) with a mid-level V8 that had a wide torque band did make for a very nice all-round power train, as fourth would provide almost-overdrive levels of highway engine speed.
About the same time as Chevy began offering four-speed manuals, Chrysler showed the way with its class-leading three-speed Torqueflite automatic (Ford’s Fordomatic had three gears going back to 1951, but started in second) until the MX/FX in 1958). With a 2.45:1 first gear and a 1.45:1 second gear, the Torqueflite (and the other modern three-speed automatics from Ford and GM) was the functional equivalent of a well-spaced four speed manual, and more, thanks to the additional multiplication of its torque converter.
Which also explains why it was so effective on the drag strip; a big V8’s power at take-off could be more easily controlled through fluid torque multiplication than through a mechanical clutch, allowing for less wheel slip, which largely compensated for its internal losses (some 45 hp). In any case, Chrysler was late to the game with a four-speed manual, so the TF was certainly faster than its three-speed manual.
The 1960.5 Corvair Monza brought some hope for America’s compacts, which most badly needed a four speed, due to their small sixes that lacked torque. The optional four-speed in the Corvair transformed it from a sluggish thing, and it undoubtedly was by far the most popular American four-speed car of the early sixties. The other GM Y-Body compacts (Tempest, F-85, Special) also offered four speeds.
The Corvair’s four speed prompted Ford to offer a UK-sourced four speed on its Falcon starting in 1961. Now that did perk up the little Falcon six a wee bit, as the three-speed had that miserable hole between second and third. The four-speed should have been standard, as well as on the Corvair.
The biggest missed opportunity to do the right thing stick-shift-wise was Chrysler with its 1960 and up Valiant (and Dodge Lancer/Dart). Their very lively slant six, especially the little high-winding 170 inch version, really could have put a four speed manual to best advantage.
With the very powerful Hyper-Pak 170 six, which made 148 hp (and more) and would rev to 6500 rpm, the Valiants blew away the other new 1960 compacts on the race tracks. But oval track racing is hardly the same as trying to make good time on hilly and curvy roads, as I did in my father’s 170-equipped (not Hyper-Pak!) ’68 Dodge Dart with its three-speed manual. It handled deceptively well, and the little slant six gave it its all, but the hole between second and third was abysmal in spirited back-road driving. What a let-down to an otherwise excellent car.
Speaking of, the Dart’s three-speed still lacked a syncro on first gear, an unforgivable sin in 1968. The rest of Detroit was similarly late in adopting first-gear syncros; VWs had them since 1961.
The reality is that two-speed automatics didn’t really give much away to three-speed manuals, except a bit of acceleration in certain parameters as well as some fuel efficiency, a price readily paid by the majority of buyers. Yes, they stayed around for way too long, but then so did the three-speed manuals. They both were sub-par in a quickly changing world, and frankly, an embarrassment for Detroit; another factor in its decline in the face of imports sporting slick-shifting four, and soon five-speed manuals.
What American cars really lacked, especially starting with the compacts of 1960, was standard four speed manuals, along with optional three-speed automatics by all the makers, since the two roughly equaled each other. Or even better yet, five speed manuals and four speed automatics. But then Detroit was perpetually stingy when it came to more gears, syncros, disc brakes, anti-sway bars, radial tires, fuel injection and a few other key components, preferring to shower Americans with the really important stuff like vinyl tops, opera windows and loose-pillow seats. That strategy worked for surprisingly long; then it didn’t.
Powerglide: A GM’s Greatest Hit or Deadly Sin?
The Joys of Overdrive, The Planetary Kind
How GM Nickled and Dimed Americans (And Itself) To Death – With Undersized Tires
1967 Cadillac vs. 1967 Renault 10: A Unfair Comparison Thanks To A GM Deadly Sin
What I don’t understand is the lack of overdrive on most cars of the time. Even with automatic, I would’ve thought that an overdrive gear ratio would provide a better cruising speed when driving faster than 40 mph. I like both manual shifting and automatic.
Most non-performance cars had tall(low-numerical) final drive gearing, dropping engine speed on the highway even with a direct top gear. Big blocks were common in mildly upmarket cars, which made up for the missing gearing spread at the low end with massive torque at low rpm.
One reason was probably that where factory overdrives were available, they were not necessarily inexpensive. They cost less than automatic, but buyers were happy enough to spend the extra $75 or $80 to not have to shift. If overdrive had been offered with automatic, the combination would have cost $300 or more, which was a lot of money. ($300 in 1956 is equivalent to around $2,600 today.)
Obviously, there’s no technical reason you can’t have an automatic with an overdrive gear and most modern autoboxes do. One problem with the accessory approach, though, was that on a lot of the early torque converter automatics, the stall speed was fairly high, so if you added an overdrive on top of a relatively tall axle ratio, there would be a pretty good chance that dropping into overdrive would put you below the converter stall speed when cruising at lower speeds (particularly in the 45–55 mph range). In the absence of a lockup clutch, which only a few early automatics had, that would re-lock the stator and you’d get a lot of transmission slip, possibly overheating the converter after a while.
My dad’s 1980 Toyota Supra actually did have a separate overdrive box aft of the three speed automatic. It was engaged, electrically, with a pull out knob on the dash.
My second car was a three speed manual with a vacuum assist. The assist was out and using the car was a miserable experience. Over time with multiple examples of this type transmission I came to shift from 1st to 3rd, then up to second. It seemed to defeat most of the resistance to shifting probably because of non synchro first..
Have driven 3 speed manuals, 2 speed pg, and th350. Found the 350 to be IMO the best but it was backing up a 350 engine. It was old, ugly, and drove great. In my 77 Impala it gave me right of way that I never could have had with an attractive car.
I don’t expect I am going to own a 9 speed but my wife dragged me kicking and screaming into a car with CVT. It works well.
I’m in San Francisco on business this week and have a Nissan Altima (rental) with a CVT. It was very difficult to drive in SF. The engine ‘drums’ with a loud harmonic noise, and alternates between having no power to revving like crazy.
I tried using the ‘sport’ mode — that kept the noise down but the engine revs to 3k and above. It was a very difficult car to drive in the Nob hill area – actually anywhere north of Market ST. I just left it at the hotel and walked or used Uber.
I drove to Sacramento this afternoon, and it was like any other beige sedan on the interstate. Soulless transportation, which is good enough for the week.
I learnt to drive in a ’69 Falcon with the all-synchro box. Then I got into Dad’s ’67, and tried to shift to first for a slow corner, as I had in the ’69. Awful noises ensued.
I learned to drive in a 1964 Falcon with the 2-speed Ford-O-Matic. It would go from 0-60 in . . . well, I never timed it. Still haven’t after owning the car since 1989. I figure anyone — myself included — expecting something called ‘performance’ from a ’64 Falcon with a ‘170’ and a 2-speed auto tranny is expecting too much!
My 72 F100 with a 300 I-6 and a 3.70 rear gear is a joy to drive around town with it’s 3 on the tree. At one point, I replaced the 3.70 with a 3.00 because I was doing some interstate driving at the time. It would cruise like a champ on the highway, but was a real turd when trying to pull anything.
One of the great things about the 3 on the column shift today… theft deterrent!!! Last time I took it to a tire shop, I had to drive it on the rack.
My 1972 Citroen DS20 had a five-speed column shift – the guy who sold it to me left it unlocked in central London because he said nobody could ever figure out how to drive it.
Once I drove a Citroen C25 (which is a van) for about 400 km with a column mounted shifter.
It´s only at the end of the trip that I realized it was a 5 speed and not a 4 speed with a big gap 3rd and 4th.
I kept missing the 4th gear, going directly from 3rd to 5th gear !
Funny you mention that, Chris. The last time I went it for tires a few years ago on my ’64 Falcon I drove it up on the rack myself. Even though the car has an automatic transmission and is just a harmless old machine it nevertheless managed to scare the employees enough to where they just let me drive it up. (I suppose the fact it also has no power steering or power brakes had something to do with that as well. Plus, car was 50 and the workers at the tire shop were aged 25-30). And so it goes . . .
Whenever I can I’ll downshift manually my automatic transmission, usually from drive into second gear. I don’t know whether it’s ok for the transmission or not, but it helps to keep the speed down to a more manageable level when descending a steep hill.
That’s exactly what the “3” and “2” settings on an automatic are for–maintaining speed while descending a hill.
Thank you. I’ve been told by some people (I won’t name names), that downshifting an automatic transmission manually is bad for the transmission. I don’t know where they get the information from. I usually downshift according to the speed range I’m going. For example, if I’m going going between 15 and 25 mph, and I’m descending a steep hill, I’ll downshift from drive to 2nd gear. Once the road has leveled off, I may move the gear shift from 2 to D.
On some automatics, if you make a habit of doing that all the time — not just occasionally when going up a steep hill or the like, but trying to treat a normal automatic like a manual transmission — it will wear out the clutches pretty quickly.
Automatics with paddle shift controls and the like may be more robust, since they’re actually inviting you to shift that way. With your run-of-the-mill automatic, though, the design assumption is that the average driver will manually change gears for special occasions and mostly won’t bother.
You could compare it to a starter motor. Obviously, the starter motor is designed to start the engine — that’s the motor’s job. However, the assumption is that you’ll only start the engine a few times a day and that there will be a decent interval between starts. If you want to turn off and restart the engine at every stoplight like a stop-start system, you’ll wear out the motor more quickly; stop-start usually requires a heavy-duty starter motor that’s designed for that kind of severe use.
IMO, if that kind of operation will burn out an automatic, the transmission was under-designed!
If I have to stop my RAM2500 from high speed, I will often drop it down from overdrive to drive at 50mph to help with braking. On my way to work there’s a long hill with a stoplight at the bottom. if that light is red, I’ll sometimes shift it again, down to 2nd, when the speed gets down to 35mph. Been driving it for over 125,000 miles like that.
That makes sense. Not many people stop the engine of the car while stopped at a traffic light. They usually stop the car when they stop somewhere for 30 minutes or more before starting the engine again. Most of the time, I leave it drive throughout my time driving the car. It’s only when I ascend the steep hill or descend the steep hill that I downshift the automatic. That can’t hurt the transmission, can it?
Downshifting an automatic to lower gears at various speeds does not “help with the braking”. It puts more of the braking (RWD) on the rear wheels, which do less braking work than the front wheels, and uses up another and fairly stressful cycle of the transmission clutch’s life. Brakes are for braking, operating proportionally on all four wheels.
On long downhills of course putting the automatic in a lower gear will save brake lining and avoid brake overheating.
Driving around my very urban area with a 25 mph speed limit with a three speed (old) automatic, I often start in 2 not D and leave it in 2 mostly, since the transmission does not know there is a stop sign or red signal or traffic coming up in half a block and shifting to 3rd when I ease off on the gas pedal is pointless and only uses up a bit of transmission clutch and servo life for no reason, since slowing down or stopping is the next thing happening.
Most owners operation manuals tell you to downshift the auto descending steep hills whoever told you not to is a fool.
I usually drive my Mercedes auto manually, both up and down, so far
600 000 km so obviously it doesn’t harm the box.
I have no idea what kind of Mercedes you have, but Mercedes engineers for many years — being German in an era when automatics were relatively rare in Europe — assumed that drivers would want to shift gears manually when necessary. Conversely, the low gear on early torque converter automatics like Dynaflow and Ultramatic was described (for anyone who read the manual) as “emergency low” and using it regularly would wear out the brake band and/or clutch pack.
Brakes are less expensive to replace/repair than transmissions. Unless your manually shifting to a lower gear for safety reasons (like not overheating the brakes) I would suggest you keep this to a minimum. Also no good reason to downshift to lower gears as you decelerate with a manual transmission. It doubles the wear on the synchronizes.
I have to go down to 3rd before my 5-spd Sienna holds back on downhills. Maybe it’s simply too much momentum; I have less trouble with small-car automatics.
Intermediate speeds prevent transmission “hunting” while climbing long hills on Interstates, & a good reserve for passing slow semis on these.
I drove a new Yukon with the 6-speed automatic for a week this past spring. Even when I downshifted manually, it had very little engine braking until I downshifted a couple of times. The transmission had some other quirks too, which really made me hate driving it except on the highway.
I know that modern automatics use thinner fluid (eg:Dexron-VI) to reduce parasitic losses. They may also use some mechanical tricks as well, intended to improve fuel economy. I suspect that this makes them not provide as much engine braking when coasting versus older automatics.
The 1960 Daytona race where the 3-speed Hyper-Pak Valiants finished 1-7 was actually on the 3.8 mile road course. They won the following oval race too, but fewer finished well due to a crash.
Thanks for the explanation of how the Powerglides worked. I never understood why anyone would consider one against a Torqueflite, especially considering they were offered a dozen years after Mopar’s automatics were exclusively 3-speeds.
What’s always bothered me about manual shifting cars of the time was not so much the lack of more forward gears, but that the first gear was usually non-synchromeshed. Call me lazy, but who in their right mind would want to depress the clutch more than once between each gear? Except for reverse, all one wants to do is get into the car, press the clutch to the floor, put it in gear, and drive, only depressing the clutch once per forward gear.
What vehicle is that in the 10th picture? Looks pretty new to have 3-on-the-tree (70s or 80s?).
I’ll bet it’s a Ford F-100 or F-150. Other folks will pin down the year better, but you could get a 3-on-the-tree until at least 1986.
It’s definitely a “bullnose” (80-86) F-100/150, but I can’t say any more than that.
Fun fact: the leftover 3-on-the-tree trannies were used on a few F-150 Customs in 1987.
I’ve never seen a car or truck into the mid 70s or later with a manual shifting transmission with the column mounted gear shift control.
I’ve seen (and driven) single cab Nissan/Datsun utes with 5 speed column shifters from the early 90’s.
“Which also explains why it was so effective on the drag strip; a big V8’s power at take-off could be more easily controlled through fluid torque multiplication than through a mechanical clutch…”
Go to a Drag Strip and see all the ‘slushbox’ race cars going down the 1/4. Race prepped Poweglides are still available from Performance Parts shops.
Where manuals shine is the ‘twisties’ and autocross.
The other consideration with Powerglide for drag racing, which we’ve discussed here before, is that Powerglide’s internals, even with heavy-duty clutch packs, are light and don’t consume a lot of power.
Paul explains all. This certainly matches my experience learning to drive in a ’62 Bel Air with the 283 small-block/Powerglide and a ’65 C-10 with 230 Six and three on the tree. It is hard to describe adequately the feel of the small-block/Powerglide combination, accelerating from rest up to cruising speed in one steady rush of power, and then listening to it shift into a higher gear only when you LET UP on the gas. And yet, for the time, it worked fine. Perhaps an engine that didn’t breathe so well would have suffered by comparison. And there was indeed a big hole between second and third in the manual transmissions of the time. In town, geared as the truck was, you were fine, as second gear was just about right for 25-35 mph speed limits. But you paid a price on the open road, especially accelerating uphill (there was little traffic to speak of in the country at that time). You revved the engine up in second till it was as noisy as you could stand it, let up on the gas, pushed in the clutch, slid the shifter lever down, let out the clutch, stepped back on the gas and prayed. And buzzed.
Powerglides were fine for two-lane blacktops, where your real accelerative needs were from rest, or a low speed. But Interstates killed that. And the three-on-the-tree was already in over its head. Thanks, Paul, for reminding us that it’s not just the number of gears, it’s everything.
You have hit on the reason for the PG’s long success: it was usually mated to the excellent breathing Chevy smallblock V8, which was quite happy at higher revs compared with many other engines of the day such as the Ford Y block or the Mopar poly.
Wow, such a good analysis of an interesting issue. Personally, I didn’t notice much of a difference between a 3-speed and a 4-speed manual transmission coupled with essentially the same 2.5 L engine, other than the 3-speed having non-synchronous 1st gear which is of course a major penalty in modern traffic (may be this very fact attracted all of my attention ?..).
To engage a non-synchronous 1st gear on a standstill without clashing the gears (that wears them our really quickly), you have to switch to 2nd gear first, and, without releasing the clutch pedal, move the lever to 1st – which isn’t always possible to do from the first try, because sometimes the lever may “stuck” and refuse to move into 1st completely unless you rev the engine a little bit. Doing all that on an incline, holding the brake pedal all the time, is especially tricky.
To shift into 1st on the move, you have either to double-clutch, or to slow down to 5 km/h (3 mph), so the usual habit of the drivers “back then” was to avoid 1st gear completely, unless for starting from standstill.
While that wasn’t much of a problem on the usually half-empty streets of the Soviet cities of 1950s and 60s, I am truly surprised how the American drivers of 1940s and 1950s managed to drive cars with such transmissions in relatively dense traffic.
Surely, *any* automatic transmission was a BIG improvement over a 3-speed without synchromesh on the low gear ! This doesn’t apply to a 3-speed with synchromesh on all froward gears, of course.
Just as Paul describes it, in a 3-speed transmission there is a big gap in gear ratios between 1st and 2nd, as well as 2nd and 3rd gears. Of course, this is especially torturing with a non-synchronous 1st gear.
Surely, the same may be addressed to 2-speed automatic transmissions as well, but the torque converter is supposed to compensate for this somewhat, and large 6- or 8-cylinder engines of American cars provided more than enough low-end torque.
All in all, I’d say that for relaxed driving in relatively light traffic an all-synchromesh 3-speed manual transmission with a floor-shifter would be almost as good as a 4-speed; driving a car with a 4-speed manual and a reasonably torquey engine, I often skipped the 3rd gear, shifting right into direct-drive 4th if there was no need to accelerate very rapidly.
Combined with very short (4,55:1) rear axle ratio, GAZ-21 Volga’s 3rd (direct-drive) gear is good for any speed between 30 to 130 km/h (20-80 mph), and the 2nd – for anything between 5 and 50 (3-30) – of course, without much ability to quickly acceleration closer to the lower boundaries of these intervals. There is a penalty for such versatility, of course – a top speed of 80 mph should have been extremely sluggish for the American highways of the time, I guess. An external overdrive unit would’ve help much, though.
All in all, I always perceived “additional” gears as a result of the switch towards higher-revving, no-low-end torque engines.
Power glides; I remember sitting in the third, rear facing seat in the ’61 Chevy Parkwood, Someplace on a two lane road in Michigan or Canada and I’d hear the exhaust get louder and faster; Dad’s gonna pass another car. 283 two barrel. I’m still here so he must have judged the distance pretty well. But there were a couple of close ones…
’65 Corvair Monza, 140hp, 4 carbs, PG. Friend driving his mother’s ’67 Beetle was pretty much as fast as the Corvair. He could row between second and third and easily keep on my tail as the PG either over revved first or lugged high.
“’70 Camper Special, 4 speed with a granny first and then a regular 3 spread of gears. Up in the Sierra foothills with a 9.5′ camper in the bed and 4 dirt bikes on the trailer, either revved second to the moon, or lugged third. Finally decided to slow down and fart along in second and pull over to let the line of cars pass.
“What American cars really lacked, especially starting with the compacts of 1960, was standard four speed manuals, along with optional three-speed automatics by all the makers, since the two roughly equaled each other. Or even better yet, five speed manuals and four speed automatics. But then Detroit was perpetually stingy when it came to more gears, syncros, disc brakes, anti-sway bars, radial tires, fuel injection and a few other key components, preferring to shower Americans with the really important stuff like vinyl tops, opera windows and loose-pillow seats. That strategy worked for surprisingly long; then it didn’t.”
It was amazing how backward the Detroit iron was in the ’60s and ’70s. The transmission thing was obvious to me as a teenager in the late ’60s, you had to wonder why the genius’s at the Big 3 couldn’t….
Love the article, Paul. Very informative and interesting.
So there were no divisional separations between the transmissions after the PG/Dynaflow came out? Or was the Hydramatic always reserved for Cadillac and Olds in the beginning?
I really want to experience both of these transmissions, specially the original Hydramatic.
Was there a mechanical difference between the Dynaflow and Powerglide?
I too always wondered why GM remained stagnant with the 2 speeds, but it seems like they remained fairly competitive, at least compared to the Ford and Chrysler.
Interesting that Ford/Chrysler having more gears than GM didn’t usher in the Transmission Wars some 50-60 years early.
Now everyone who’s anyone needs at least 6 speeds. But I think Corollas, at least until recently, packed 4 speeds. And my ’08 PT Cruiser also has a 4 speed. I could use the extra gear for fuel economy purposes (though 3k at ~80 isn’t bad), but I don’t see the need for much more than that. I don’t have to fish gears for power, just go from D to 3 if I want to pass and hold the gear, otherwise just press the pedal a bit more, and she’ll know what to do.
Pontiac also used the “old school” Hydramatic from the beginning. It has been a long time since I have driven a four speed Hydramatic but as I remember them they would accelerate rapidly due to low gearing in 1st and then shift to 2nd at about 8-10 mph. I’m sure the shifting was clunky but when all of your experience is in older cars, they all tended to be clunky. The Dynaflow was a hoot, you would push down on the gas pedal, there would be a delay, and then the car would start to move as all the parts spooled up. The current CVT remind me of being in a 1950’s Buick in that way, some delay between applying the gas and actually moving.
In my 63 Cadillac, the 1-2 shift and the 3-4 shift were quite smooth, but the 2-3 was fairly abrupt in comparison. Paul is right that the 1-2 shift came very quickly from standstill, unless you were really standing on the gas.
The 2-3 and 3-2 shifts require two planetary gear sets to shift simultaneously (fat chance). I sort of understand that the first generation hydramatic probably started with the safety transmission, so the design is not a surprise. But the hydramatic had 3 planetary gear sets, so the updated redesigned hydramatic could have used all three for first gear (with 2 1.6:1 ratios and third gear around 1.5:1). This would have allowed on gear set to shift at a time. First gear would have been about 3.8:1, then second 2.4:1, then 1.5:1.
The two-three shift was the bête noire of all the old fluid-coupling four-speed Hydra-Matics. The reason is that executing the shift involves several separate internal operations that need to happen simultaneously. If the timing of any of those operations is a little off, which is pretty much inevitable with a complicated mechanical object that’s under frequent but inconsistent stressed, then you get a jerk. The later dual-coupling transmissions took pains to reduce that, but the two-three was still more mechanically complex than on a Turbo Hydramatic.
I recommend you read ateupwithmotor’s two articles on early GM automatics for all of the more detailed issues that I didn’t have time to get into here:
I think that the first Dynaflow and Powerglide were the same design, but the Buick used castings for the turbines and Chevy used stamped parts. While most torque converters are basic 3 element designs, both the dynaflow and powerglide had 5 elements. There was one turbine, but the pump had two turbines, one could overrun (at higher speed?), and the stator was in two parts, one at a higher angle. When Buick went to the twin turbine design (one pump, two turbines, one stator), Chevy went to a simple (I think) torque converter and converted the gear box to an automatic two speed unit.
Buick then proceeded to refine the twin turbine with variable pitch stators and finally added a stator between the two turbines. This final refinement brought the torque ratio to 3.5:1.
What you got with the early GM automatics was a reflection of the fact that much of the basic development work was done by the central Engineering staff rather than by the divisions. The corporate engineering team would come up with the concepts and then if the divisions were interested, the division engineering department would develop the idea for production.
So, both Powerglide and Dynaflow were based on a common set of ideas developed by the corporate transmission engineering team, led at that time by Oliver Kelly (whose name is on many of the related patents), but since Buick and Chevrolet were each going to manufacture their own transmissions, they went their own ways in terms of the execution and the two transmissions eventually diverged quite a bit.
In some cases, several of the divisions would agree to collaborate on a common production version so they wouldn’t have to eat all the development and tooling costs. The two-speed torque converter automatic that Buick, Oldsmobile, and Pontiac used from 1964 until the early ’70s — typically called the Super Turbine 300, which was Buick’s name for it — was developed like that.
In the case of Hydra-Matic, production responsibility was actually handed to a new, separate division pretty early on, after which the divisions that wanted to use Hydra-Matic would buy transmissions from the Hydra-Matic Division.
Excuse me, Oliver Kelley (who later became Buick’s chief engineer, although he wasn’t in the ’40s or early ’50s).
Hydra-matic division. I was (and am still) amused that the. Five- speed manual transmission in my 1991 S10 Blazer was built by Hydra-matic, the name is cast right into the case. Cracked me up the first time I was under the truck.
“In some cases, several of the divisions would agree to collaborate on a common production version so they wouldn’t have to eat all the development and tooling costs. The two-speed torque converter automatic that Buick, Oldsmobile, and Pontiac used from 1964 until the early ’70s — typically called the Super Turbine 300, which was Buick’s name for it — was developed like that.”
I often wonder why they did that. Why not buy Powerglides (modified to fit each engine) from Chevrolet, and save the developmental expense and that for a new production line?
Or was that one of the last vestiges of the Old GM, with its largely independent divisions?
In part the latter, but the logic will make more sense if you look at the comparative production of Chevrolet and the three mid-price divisions.
In 1964, the first year the ST-300 was used, Chevrolet built something approaching 2.4 million cars, probably at least 85 percent of those with Powerglide. Buick, Oldsmobile, and Pontiac’s combined volume was on the order of 1.8 million that year. I don’t have the time or the patience to calculate how many of those were intermediates and the Jetstar 88 and LeSabre, but at a very rough guess, it was around 600,000 units, probably 85 to 90 percent of which had the two-speed automatic.
GM operated (and probably still operates) on a profit center model where each division has its own P&L, so if one division sells components to another, it’s usually at a markup. If the division has to built a new plant or expand existing ones to supply those components — which I strongly suspect would have been the case here — that investment is going to be reflected in the price.
So, Buick, Oldsmobile, and Pontiac had three choices: 1) convincing Chevrolet to supply them at what would likely have been a substantial premium, 2) tooling up to build Powerglide at their own plants, or 3) developing their own transmission and splitting the cost three ways. My assumption is that option 3 was still less than option 1 and not dramatically more expensive than option 2, since by the early sixties two-speed torque converter automatics were not exactly startling new technology. It was probably also a better solution than trying to get Chevrolet to make changes to an existing and popular transmission just to suit the other divisions’ needs.
The other factor in that case was that ST-300 was not just Powerglide under a different name. It looked very similar and the actual gear layout was about the same (Ravigneaux gearset with short and long planets, two sun gears, one ring gear, and a single planet carrier), but the control system was not. The ST-300 was vacuum-controlled, like Turbo Hydra-Matic, so it adapted its shift points and engagement pressures based on manifold vacuum. Powerglide could modulate engagement pressure based on engine vacuum, but shift point was varied with a mechanical throttle valve, so it wasn’t as “smart” as the ST-300/TH400 in varying its shift behavior.
Some versions of the ST-300 also had a variable-pitch stator, which Powerglide never did.
So, the ST-300 was a somewhat superior transmission — not hugely superior, but a bit more modern — and sharing it across the BOP A-bodies and some low-end B-body models meant the unit cost was probably competitive with buying Powerglide from Chevrolet (especially if that involved trying to convince Chevrolet to make these engineering changes).
Also, on the technical front, yes, Chevrolet eventually dumped the five-element converter for a straightforward three-element unit.
The four- and five-element torque converters are a very complicated subject because there were several variations — dual impeller, dual turbine, dual stator, triple turbine — that worked differently. Some incorporated epicyclic gearsets for additional reduction that varied based on the rotation speeds of the various converter elements, separate from (and in the case of the triple-turbine, instead of) the normal planetary gearbox.
Everyone had to make due with Dynaflows for a little while in 1953 when the Hydramatic Plant in Livonia Michigan burned down, though GM was able to get another plant up and running in a few months, there are a few Dynaflow equipped Cadillacs and Oldsmobiles.
Most GM transmissions today are made by the Hydramatic Division, though I think its part of GM-Powertrain group today, they just launched an 8 speed Hydramatic for the Corvette, large SUV’s and Cadillacs.
GM also buys transmissions from Aisin, notably in the Cruze.
Wrong-The Cruze uses a GM produced transmission
Excellent piece. I’m going to have to read it again before the quiz! I like the Dynaflow’s simple explanation. This kind of text didn’t exist in the ’80s. I recall kids in high school shifting into park while rolling at low speeds. If the knuckleheads understood that a steel bar was dropping into a gear, (and could break things) maybe they’d have been a little smarter, or at least understood better where their $400 dollars was going for the rebuild.
Maufacturers have lost the mindset of that Dynaflow explanation: Making things simple so you can attend to the job of driving. Complexity for complexity’s sake in so many controls and systems. Hopefully this trend will reverse itself at some point.
Studebaker, was always an odd-ball. They sold most of their fleet with three-speed Over-Drive manual units, but also had their own, fully functional three speed automatic unit starting in 1951, made by Borg-Warner. Sadly, since their production numbers did nothing but decline from 1950 onwards, they couldn’t maintain the cost of production and were forced in 1956 to switch to the units BW was already selling to AMC — the Flash-o-matic.
Oddly, With their 3-speed Automatic Transmissions, Studebaker only allowed first gear start in cars with the six-cylinder engines. Their V8s, even the Avanti with the PowerShift Automatic (a floor-console mounted shifter that let you easily shift gears at will), all started in 2nd, with first gear only a ‘selectable’ option from the column shifter. This may have been done to lessen the jerk into first gear on slowdown at a stop, but it just killed the performance if your main interest was to just get away on the green! And with their column-shift setup of PNDLR, you had to be careful not to pull it into ‘R’ if you wanted to gain that performance at a stop light! (And then you also had to remember not to pull the shifter all the way down for reverse if you were driving someone else’s car).
The early Ford-O-Matic/Merc-O-Matic (also developed with Borg-Warner — Harold Youngren, Ford’s chief engineer at the time, had been at Borg-Warner before going to Ford) worked the same way.
The important thing to understand was that when those transmissions were developed, the three-speed-plus-torque-converter automatic didn’t yet exist except on paper. The three prevailing paradigms were the ‘pure’ torque converter transmission, which had a low gear that could only be engaged manually; the fluid coupling four-speed Hydra-Matic; and the semiautomatic transmission, many of which had a clutch pedal as well as a fluid coupling or torque converter.
At that time, the idea of a transmission that would give you a torque converter and two automatically shifted planetary gears plus an emergency low was an appealing compromise, and even as some of the torque converter automatics started adopting automatic shifting from low to high, the Studebaker/Ford/Borg-Warner approach was really no worse. It wasn’t until the introduction of the ‘true’ three-speed torque converter automatics — TorqueFlite and Cruise-O-Matic, initially — that they older variety started looking crude. Ford’s solution, interestingly, was to develop a new, simpler two-speed Fordomatic, which they offered as an alternative to the three-speed transmissions into the mid-sixties.
Part of the reason it took a while before the three-speed torque converter transmissions appeared was that an ex-Ford engineer named Howard Simpson had managed to sort of blanket the field in terms of patents. That created some delays while the various manufacturers tried to see if they could find a way to do avoid having to license Simpson’s patents, although all of Detroit and some non-U.S. manufacturers eventually conceded that Simpson’s patents were worth paying for.
Hey, what about the push-button-shifters on the Torque-flights, Packards and Edsels?!?
That’s what really made the interiors of the Exner-era cars feel like they were into the space-age!
They made it so easy to rock my ’64 Dart out of a snow drift when stuck. That all changed after GM’s evil work to change the Federal Regulation in 1965 that all Driver-Education Cars to require column mounted shift linkages, not buttons. But I guess that’s another column…..
My driver’s ed car in the summer of 65 was a 65 Plymouth Fury III with the automatic on the column. Ironically the shift indicator did not align properly with the gear engaged so it was probably less “safe” than a 64 with pushbuttons. I’m not familiar with the regs but it did make sense to standardize the confusing array of shift quadrants and mechanisms used in the 50’s and 60’s. It seems we are going back to that situation today with pushbuttons (Lincoln, Acura, etc.), dials (Jag, Chrysler), column shifts (M-B), floor shifts, etc. I assume the governing regs changed at some point…
I’ve never driven a Mopar where the PRNDL needle lines-up perfectly with the letters.
The gov had nothing to do with Mopar dumping pushbuttons. It was a decision by Lynn Townsend. The main reasons were that Townsend wanted to purge Chrysler products of weirdness in general. And yes, he did take the Drivers Ed market into account, believing that pushbuttons were hurting sales in that department. The proof is in the pudding. Look at any ’65 model Chrysler product compared to just a couple of years before. No more pushbuttons, no more rotating door locks (as opposed to the plunger type), no more square steering wheels, no more dash mounted mirrors and turn signal switches, no more lots of nutty stuff.
Wasn’t there something about GSA specifications for 1965 government car buying that specified a lever operated transmission? Or is that an apocryphal story?
I recall chasing that down once. Wish I had kept the citation. In 1964 you still had lots of GM cars (and Studes) with PNDSLR patterns and buttons in every single Mopar. In 1965 it was only the Stude which, as a canadian import, was probably not on the menu for US government purchase.
The buttons were a big turn off for non-Mopar buyers, but were very popular with existing customers. If it were purely market driven, how come the new 64 Imperial didn’t get a lever? Also, why spend money on the final 65 B body to retrofit the lever? The new 63 Valiant got a lever for the 3 speed but kept the TF buttons? The switch to a lever had the earmarks of a rush job in that it operated the same cable mechanism that the pushbuttons did. I don’t doubt that Townsend wanted the buttons gone, but he was also an accountant. He could have saved a lot of money going to levers tied to model cycles (every car in the line turned over between 63 and 66) instead of across the line in 1965.
The explanation at the time was that Chrysler product owners liked the push buttons but a substantial number of others thought they wouldn’t and wouldn’t consider Chrysler products for that reason. So, conquest sales was the reason. And probably a few driver’s ed cars.
The first Chrysler products to drop the buttons were those with bucket seats and the newly fashionable floor shifter in 1964. Then in 1965 the rest went to the column lever.
The ’65 Imperial was switched to a lever.
I don’t recall GM having any particular action that caused the universal adoption of PRNDL. Can you cite some sources?
What I do recall is that the Federal Government tried to pass safety standards requiring standardized shift patterns. Owing to Big3 lobbyists, the laws did not get passed.
Government regulators made an end-run around this by telling manufacturers that the government would no longer buy any fleet cars from the Big3 that did not have that shift pattern.
My Grandmother’s 1950 Buick had the PNDLR pattern. Pushing the car to start it (in exceeding cold weather) would require care in moving from N to L after getting it up to 30-40 MPH.
I should add (now that I found it), the gear box on the Buick dynaflow (not the triple turbine) was Ravigneaux gearset. I am sure that the powerglide used this too. The torqueflite used a simpson gear set. Wiki has info on both. The turbohydramatic was a simpson.
My Dad’s 57 Olds 88 Super also had the PNDLR pattern. One time while going down a steep mountain road he slipped past low into reverse. The tires skidded and the rear axle was bouncing like crazy. He quickly shifted back in to low, the car smoothed out and was fine. I was amazed nothing got damaged.
That brings back memories! I remember the antics my dad went to getting cars started in extremely cold weather when the power had failed, meaning no block heater. That is just a memory now. You jump in, turn they key and at any temperature, the car starts. No more cold start voodoo rituals.
I found this on PRNDL:
I think SAE & the insurance industry could indirectly encourage many of these things instead. More good stuff => lower insurance premiums.
Your recollection is more or less correct. In fact, GM was not the instigator of the rule but the target.
Around 1960, an independent engineer and transmission inventor named Oscar Banker contended loudly that the GM PNDLR pattern was unsafe. He wrote a pointed letter to the president of GM about it and spoke out about it at SAE meetings on that subject. It was probably Banker’s personal campaign that led to the General Accounting Office (GAO) regulation you mention.
That became one of the points that Ralph Nader picked up for Unsafe at Any Speed, which annoyed Banker to no end, but may have been part of why the GAO rule later became one of the first batch of federal safety standards.
Again, the issue was the pattern used by some GM transmissions that put reverse next to Low. I’ve never seen the text of the GAO regulation, but assuming it’s similar to 49 CFR 102, there’s no particular reason the Chrysler pushbuttons wouldn’t pass those rules. In fact, since the concern was the potential for accidentally selecting reverse instead of low, the Chrysler system was arguably safer since the buttons were unambiguous.
In all Chrysler push button controlled automatics starting with the first 1956 Powerflites pushing the R button at over maybe 5 mph resulted in the transmission going into neutral.
Ford actually was the first user of the PRNDL layout, GM for the most part, liked to used the PNDLR pattern on most cars until the Turbo 400 came out in 1964. Corvairs didn’t even have a Park position through their entire 10 year run, just NDLR.
It depended on the transmission. Some GM transmissions, like the Flight Pitch Dynaflow/Triple Turbine, had PRND patterns while Dynaflow and the Hydra-Matic family had reverse below low.
There were common sense reasons for GM to have made the change. When one talks about jerky shift changes my mind went back to about 1960 when driving my Dads 55 chevy. I shifted into reverse at perhaps 40-50 mph when pulling out to pass. Stopped the car, of course, but it started right up again and drove the same.
I guess it was locked out of reverse but I still think it was one tough transmission and that the jerk in the shift was driving the car.
The pushbuttons really didn’t have anything to do with the design of the transmission. Essentially, with non-electronic automatics, there’s a Bowden cable that connects the selector controls to the transmission controller; moving the lever changes the effective length of the cable. The Chrysler pushbuttons did exactly the same thing in a slightly different way — there’s still a cable, just like a column or console shifter, which is operated by pushing the button. The Edsel and Packard electrical systems did it with a solenoid that was triggered by pressing the button.
Great read, with some good information explaining the idea well. Thanks!
As regards manual gearing, the irony was that the Corvair four-speed had notoriously awkward gear spacing. The Corvair box, which was sort of contrived out of the three-speed unit, was 3.65/2.35/1.44/1.00, so you had big gaps between first and second and between second and third — better than the three-speed, but still far from ideal.
The Falcon wasn’t vastly better. I think the Falcon four-speed was the gearbox from the six-cylinder Zephyr and Zodiac, which had 3.162.21/1.41/1.00 ratios. That wasn’t bad from first to second or third to fourth, but still left you with a big gap between second and third. (The smaller English Ford all-synchro four-speed had ratios about like the Corvair’s.)
That got largely fixed after a couple of years. The 1964-1965 Corvair had a 3.20:1 first gear, and the ’66-up had a 3.11:1 first gear (along with different ratios for the other gears). Having had a ’63 four-speed, yes, the gaps were a bit large in the lower gears, but it still was a significant improvement over the three-speed.
The rpm drop wasn’t all that bad (about 1200rpm from first into second and 1300rpm from second into third), considering that the Corvair engine was hardly a revver, and had a pretty fat torque curve.
It’s surprising that the Dagenham 4-speed wasn’t included and mandatory with the Futura. Having seen a couple of early Falcon Futuras, they’re an odd duck indeed with the buckets and console and a column shift.
Well, Ford undoubtedly knew that a lot of Futura buyers were going to order Fordomatic anyway. Also, I imagine supply might have been an issue. I don’t think the Zephyr/Zodiac line had a four-speed until early 1962 (there was a smaller four-speed for the Anglia and Classic, but it didn’t a synchronized low and probably didn’t have the torque capacity for the Falcon sixes), and sending Dagenham’s entire production abroad might have been a problem.
My 1970 MGB had a 4 speed with ‘electric’ overdrive. A solenoid engaged a small planetary gearbox bolted to a modified transmission tail casting (iirc). It was cool that it could be engaged in either 3rd or 4th gear too. Did Detroit manuals with OD do something similar?
Essentially the same, but with some operational differences. They generally worked on third gear only, unless one disconnected the governor and ran them manually, like I did with mine in my truck.
Full OD story here: https://www.curbsideclassic.com/blog/the-joys-of-overdrive-the-planetary-kind-not-interstellar-or-otherwise/
Most American factory overdrive units were similar in principle (though not operating detail, as Paul says) to the Laycock de Normanville overdrives used in many British cars, but there were also some two-speed rear axles. I don’t know of any postwar factory applications of those — although there were a few before WW2 — but there were some aftermarket units.
There is a video on YouTube by 4spdBernie of his 1959 Chevy Bel-air with 283 powerpack with 3 speed manual with overdrive…..He drives the car and demonstrates the overdrive operation and kickdown for passing….He also shows 2nd gear overdrive operation which overcomes the shorter direct drive 2nd gear limitations. ..much as what is described in tbis article on the overdrive equipped 1966 F-100.
My 2005 Impala with 3.8 V6 and 4 spd auto overdrive is set up for fuel economy….I can drop the car manually down into 2nd gear going down a steep grade and it will build up speed to well over 50 mph while the engine is at only a fast idle….The engine speed in braking mode is not equivalent to the speed in acceleration mode in same gear. ..50 mph with foot on gas in 2nd gear is well over 3,000 rpm…..but coasting down a hill in 2nd gear braking mode, the tach is lucky to touch 2,000…..
Ive never driven a PG equipped car, so cant really comment there. My ’05 Ram has the 5 spd automatic. Essentially its a 4 spd with 2 alternate 2nd gears. In theory, its a major upgrade from a normal 4spd…but being a lifelong manual fan, NO automatic seems to respond exactly how I would cane the gears myself…
I do know 3 speed manuals on the street though. My dad’s ’84 Dodge Power Ram has a 4 speed with granny low first…essentially drives like a 3 speed unless you need to dip into that super low. Useful for dragging Godzilla back out to sea, climbing the steepest grades where you can still get traction or gouging a canyon into the earth. My ’78 CJ-7 Renegade had a T-150 3 spd manual…either of those had their limitations on the street…always made you want just ONE more gear.
Installing the 4 speed in the 70 C10 replacing the 3 on tree shift probably saved my life when my trailer brakes failed going down the Sisku’s. A must for towing. The truck had a 3.73 rear gear, and 4th was direct drive, the same as the 3 speed. So it really didn’t help freeway MPG. I found driving it in town, unless you were starting up a steep hill, 1st 3rd 4th worked best. It seemed redundant to shift in and out of 2nd rather then just rev it a little in 1st and go right to third. Of course towing I used all four and was glad to have them all. I just got back from a road trip from PDX area to Boise Id. The Titan has a 5 speed auto and 2.96 (non tow pkg) rear end. I set the cruise at 65. The rpm’s were about 1600 rpm. I got 20 mpg. The Chevy would get around 12 mpg, around town or freeway. Once I got 15 on flat interstate holding 55 with my right mirror folded in and tailgate down. Nissan says 1 2 3 gears are the same as having a 3.73 rear, and 4th and 5th are overdrive, If you get tow package gearing (3.23 rear) 1 2 3 are the same as a 4.11 rear. I never really used the cruise control before, but now I am a total believer in it’s usefulness. I was not expecting a big 2 wheel drive truck to do that well. And there were a couple of pretty steep mountain passes in that trip. Around town it’s only good for 14, 15 mpg tops. Considering this truck weighs 4600 lbs, and the Chevy 3900 lbs, that real progress. Nice write up.
Interesting perspective. I’m old enough that I’ve had plenty of experience with Powerglides as well as 3 speed manuals. By the late ’70s when I first started driving both were pretty outdated but the cheap heaps my buddies and I could afford were too. I always hated Powerglides and swapped out quite a few for TH350s over the years. They were indeed a great transmission for drag racing, and maybe OK around town if all you needed was forward movement, but they were a misery out on the highway here in BC’s mountainous terrain. When I got into drag racing I was a 4 speed guy, but I usually ended up getting beat if we were bracket racing by someone running a Powerglide as they were more consistent and easier to launch.
3 on the tree was almost always found on low line sedans or pickups with 6 cylinder engines in my experience. I had several for winter vehicles and beater pickups and they did the job but wrestling with balky linkage and no synchro on first didn’t make for enjoyable driving either. My first car had a TH400 and I didn’t know how good I had it until I parted it out and had to make do with a ’65 Pontiac with a 230 six 3 speed for a daily driver. Great winter car though.
I haven’t had the opportunity to try the newer 8 or 9 speed autos, but I’m not sure I see the need for that many gears. Makes the 4L60E in my pickup look like a modern day Powerglide!
I kind of think when you get up to eight or nine gears, all the extra weight and complexity would make a CVT a better choice, since the desired effect is exactly the same.
Automatic transmission with 6 speeds have combinations of fewer planetary gears and may be able to have 8 or more speeds with no additional planetary gears. More complex yes, but not a lot of extra weight.
After googling GM’s new 8 speed, the 8L90, the press release says that it is about the same size and weight as the 6L80 transmissions.
I did find that the 6L90 has 3 planetary gear sets, while the 8L90 has 4.
Unfortunately, I’ve not had the opportunity to drive a car with a manual shifting gear shift control on the steering column. I would think that it’d be better than a floor mounted controller, since it’s closer to the driver, and since it’s on the steering column, rather than on the floor, it’d be easier to reach, since the driver doesn’t have to move his/her hand very far from the steering wheel to shift gears.
It’s not better. Maybe when new but age gives slop to the linkage. Think about the route it has to take to get from column to trannie. Forward, down, then back. I remember times under the hood unsticking that linkage in all sorts of weather. I can think of at least four cars where I installed a floor shifter.
The floor shifter takes the linkage out of the picture. In a perfect world you would be right but it isn’t a perfect world. As someone mentioned a column shifter is good theft prevention. At least once, the floor kit shifted backwards and we couldn’t figure out how to defeat that. That was good theft prevention too. Best ever solution would be to pull it and install a four speed.
Gee, I liked them all — I was driving!
Second overdrive was a hoot in my Hudsons, and gave plenty of power. My ’59 Dodge with Powerflite had plenty of getaway from a stop, and was almost as good as a Torqueflite in some ways. And yes, I also drove the early Hydramatics, in which you definitely felt the shifts.
Good article and subject.
Excellent piece Paul. 🙂
Sort of a tangent, but I wonder how much of the “shortcomings” of American cars were a result of geography? If the US auto industry had developed somewhere other than the Midwest (with mostly flat, mostly straight roads) would there have been less emphasis on cruising comfort and straight line acceleration, and more emphasis on handling and braking?
I have always felt this was the case. The cars of the 50’s and 60’s were, for the most part, downright dangerous in places like British Columbia. I take that back-they were dangerous and the death toll was prodigious. Combined with poor roads and it being just hunky-dory to drive around plastered, British Columbia had carnage on the roads in the era.
Flatlanders were and still are, a big problem. In the old days, they didn’t know how to drive in the mountains. I remember as a kid going down the Salmo-Creston with my parents. Dad had the 1979 Impala in 2 and was stabbing the brakes to extra slowing power, but as sparingly as possible. Then, we were passed, on a double yellow, by a 1960’s Chevy half ton with an 8′ slide in camper, totally overloaded. There was smoke pouring from the brakes and he was headed for the run-off lane. He made it, and the camper separated and blew into matchsticks. We stopped and both the driver and passenger, although scared witless, were fine-and exceptionally lucky.
These days, flatlanders still don’t know how to drive in the mountains but are saved by modern brakes and stability control, and the roads are better. However, the don’t understand how to keep the rhythm of mountain driving, so they are scared all the time and way too slow. Fortunately more and more passing lanes are being added all the time.
I am sure the cars of the 1950’s and 60’s would have been just fine on flat land, though.
Point taken, but apparently some locals had issues keeping it within the line too.
I remember that story well. I’ve driven/ridden past that exact spot thousands of times since 1972 and before. It’s not the first time this has happened either. A car and driver missing since the early ’60s turned up at the bottom of a lake here about a decade ago. Some people just vanished and haven’t been found yet. Back then it was very possible miss a curve, go over the edge and leave no trace.
A moment of inattention was all it took.
I love a good “car in a lake with skeltons” mystery story, this is a super common thing down here in Miami, we have lots and lots of lakes and canals, at least once a year one of these “missing person found” stories is in the news, last one was an old man that was missing since 1994, he was found in his MN12 Thunderbird in a lake in his own neighborhood.
The one I remember best was a missing county water dept. worker found in his county issued 1972 Bel-Air sedan in a flood control canal, he had been missing since 1979.
Good question. Keep in mind that FWIW, in the 50s and 60s, the mountain states/areas were substantially less populated than now, with CA being an exception, although the overwhelming majority didn’t live in the mountainous areas either.
I think it was just expected that folks who lived in the mountains were going to adapt, and drive appropriately, especially using engine braking by downshifting.
Frankly, the hole in two-speed automatics and three-speed manuals was a real limitation for that too, as second in a three speed auto, and third in a four-speed manual are ideal gears for mountain highways, both up and down.
In Europe, some cars in the 50s and early 60s also came with three-speeds, but they were called “flat-lander cars” and avoided in the Alpine regions for just that reason.
Very informative and interesting article Paul. As someone who grew up spending more time in “Old Look” GM busses than cars, I’d like to note the Allison “V Drive” 2 speed hydraulic automatic transmission – talk about lurching through the gears………
Very familiar with it too 🙂
That transmission was essentially a one speed:, it just used the torque converter to get up to speed (25-30) and then shifted (abruptly) into direct mechanical drive.
In the fleet of buses I drove, each one shifted slightly differently, at a somewhat different speed. It was one of the main things that gave each bus its distinctive character.
I remember that two speed one big shift thing in city buses back in the day. I didn’t know it wasn’t actually two gears. On modern city buses anywhere I don’t notice any shifts. Time for a CC article on bus and truck transmissions!
I am late to comment on this excellent piece. I plead guilty as one of the more vocal detractors of the 2 speed auto, at least in the 60s cars.
i agree completely that the Glide and other 2 speeds equate pretty well to the manual 3 speed that had been the main alternative through the 1950s. This is undoubtedly why Chrysler settled on the same design for its first auto, the Powerflite.
My gripe is that by the early 60s, the world had moved on (except at GM). GM certainly sold a lot of cars, but many of those people could hardly have cared less about the mechanical attributes. GM could have used trannys from city buses and they would have sold as well. Nobody else in the industry was selling a 2 speed auto by 1965. Sadly, until the THM, the Glide and the ancient HM were the only decent trannys they made.
GM really went into a kind of Transmission Wilderness from the mid 50s to the mid 60s. Every attempt they made to advance the state of the automatic was just not very good. It was only after they decided to give up and follow the template set by the Torqueflite and the Cruise O Matic that they built a winner in the THM400.
As I mentioned above, I think part of GM’s reluctance was that Howard Simpson had basically the field with patents on three-speed torque converter layouts. (We sometimes talk about the Simpson gearset, but he had something in excess of 40 patents.) Detroit was not keen on having to license that technology, which with the number of units involved probably ended up being some pretty serious money, although in the end even Mercedes-Benz licensed a Simpson patent(s) for its three-speed torque converter transmission.
I don’t know anything about his patents, but I would think that a torque converter with plain planetary gear sets were not covered. I have thought the Simpson’s basic idea was a certain combination of how the planetary gears were combined in simplified set of gears. There were variations on this idea.
I drove my parents 63 Impala with powerglide to college for a few years. I can say that it was much better than the Buick dynaflow, but with a 327 vs the straight 8 it would have been anyway. I can say that without a passing gear the higher speed performance was not what one would really like.
GM’s hydramatic was a good transmission. Buick’s twin turbine was not bad, and with the upper stator had good performance from a standing start. With the low range the twin turbine was OK, but the passing performance is still the drawback. The triple turbine was the holy grail for Buick in that it totally eliminated shifts for forward motion, except for the grade retard.
At the end of the 50’s and early 60’s there was a need for automatics for the smaller cars, so there were a few more transmission developed. The dynaflows and hydramatics were expensive transmissions. GM could have developed a 4 speed overdrive instead of the turbohydramatic, but it would have been expensive and while it would have eliminated the need for the big V8’s that were all the rage in the mid to late 60’s, the competitions advertising ever bigger engines would have been hard to refute.
The Chrysler Powerflite made up for having only two speeds with a four element torque converter with a maximum 2.7 ratio. It had fewer parts and was lighter than the competition when it came out.
All this talk about PRNDL! No wonder my ears were burning yesterday. 😉
Sorry I’m late for the party. Another first-rate article by Paul, backed up by our ever excellent commentariat. It’s an honor & privilege to read here!
“Conversely, the low gear on early torque converter automatics like Dynaflow and Ultramatic was described (for anyone who read the manual) as “emergency low” and using it regularly would wear out the brake band and/or clutch pack.”
Sure enough. I bought a ’63 Buick LeSabre in 1970. There was no manual, and I did wear it out, in a few months.
A friend of mine had the identical car. One day, we were going up the road when I pulled alongside. He took it as a challenge and pulled the shifter all the way down going about 30. That was the last time his reverse gear worked.
One thing I notice in this article is that the Corvette automatic floor shifters were all quite attractive! By the 70s and 80s, very few floor shifted automatic levers were attractive, excepting Mercedes. I wonder why so many GM floor shifted autos were so boring looking…
I can testify to the huge hole between 2nd and 3rd, as I had two cars with the 3 speed stick. One was a 250 equipped ’68 el Camino, the other was a ’68 Mustang with a 170. 2nd was good on both up to around 40, then you had a huge torqueless band until after 50-55 mph. Since I lived in a rural location with lots of steep hills, I either spent a lot of time lugging the sixes uphill at full throttle and them downshifting them into 2nd and enduring full rpms until crossing the peaks of hills. I would have loved a 4 speed on either.
The Powerglide in my ’69 Camaro wasn’t much better, as the 327 with a 2bbl was a slug. It could only squeal the tires when the roads were wet or when doing a sharp turn from a dead stop. Max speed with the PG in 1st was 50, which made passing difficult if there was any oncoming traffic at all.
The comments about non-synchro first gear three speed manual transmissions need an update. Beginning in 1964, Ford introduced the model 3.03 top-loader 3 and 4 speed manual transmissions. A common feature of both these 3 and 4 speed manual transmissions was the synchromesh 1st gear. The 3 speed all synchromesh manual was first used in the larger Ford models and then eventually migrated into the second generation Mustangs beginning in 1967. The top loader all synchro 3 speed manual remained in production till 1973.
As an interesting anecdote, while I was on a trip to France several years ago, when in Sarlat, I attended a local auto show where I saw a beautiful, well maintained, original well optioned red 1968 Mustang convertible with a 2 bbl 302 V8 which the non-English speaking French owner proudly told me had an all synchromesh 3 speed manual. He enthusiastically told me that with the torque of the Windsor V8 that the all synchromesh 3 speed was a delight to drive, and more than adequate in the South of France. It was enjoyable to see true Mustang Love in France, and enthusiasm for the all synchro 3 speed manual. Car enthusiasm is a worldwide phenomenon.
An interesting sidenote is that in the later 1960’s GM purchased the all synchro three speed Ford top-loader from Ford for installation in Chevies and Pontiacs. So these Ford manual 3 speed transmissions were found in GM cars without any alteration of the Ford markings on the transmission cases giving GM buyers an all synchro 3 speed manual transmission. I suspect that most of those Chevy and Pontiac owners appreciated the non clash downshifts into the synchromesh first gear, AND additionally were completely unaware that they were shifting FORD gearboxes in their GM cars.
History has a way of repeating itself, now with Ford and GM having recently collaborated on the Ford lead design work of the 10 speed automatics being used in both GM and Ford trucks and on the GM lead 9 speed transaxle for future front drive cars. Nothing new under the sun, I guess.
As a final note, in the late sixties Hot Rod Magazine published an article championing the need for 5 speed manual transmissions for US muscle cars, to the effect that “what the US needs is a good 5 speed transmission”. Likely the current Porsche and Corvette 7 speed manuals are an endpoint of the manual transmissions developmental arms race.
Automatics have ultimately triumphed, and the reliable Powerglide was an initial stepping stone success along the path of automatic domination, not a deadly sin.
Cheers, Yr Lyl Fthfl Srvnt.
I was mulling recently about the development of the automatic transmission in the US, and I think the same could be said for the 3 speed manual. Our engines here (especially pre-1950) had a really wide usable torque band, making a simple 3 speed stick or a 2 speed automatic something of a workable solution. They may not have been optimal, but they were tolerably decent. The multiplicity of speeds (early 4 speeds and 5 speeds in Europe and Japan) overseas as well as the slow acceptance of automatics there were mostly a function of small displacement engines with very narrow torque bands, often way up in the rev range. Remember that there had been a batch of 4 speeds offered in the US in the early 1930s, but in the era of in line L head engines never really caught on here.
The American designs were not satisfactory at all behind those small foreign powerplants which simply needed more help from gearing. Any improvements we in the US got in numbers of gears tended to come from overseas where they had been absolutely necessary. Over here those transmissions with more gears took cars that were OK and made them really good.
I have often wondered how the automatic transmission might have developed if they had germinated in Europe instead of in the US. There would have undoubtedly been more gears and higher shift points instead of the fewer gears and lower shiftpoints that we got.
Yes, I’m quite familiar with the Ford top-loader 3 speed, as that’s what my truck had until it failed and I replaced it with the ultra-tough BW T-87 and OD. I gave up syncro first gear, but it didn’t matter much, as driving with OD freewheeling makes clashless downshifts into first possible.
The issue with 3 and 4 speed manuals boils down to this: terrain. In flat country, 3 speeds are quite adequate, if not optimal. Many European cars from flatter areas came with three speed transmissions. But in the Alpine regions, 4 speeds became almost ubiquitous early on, as the gaps between gears on a 3 speed are very obnoxious in the mountains or hilly areas.
I’m not surprised the happy Mustang owner lives in the south of France. And who knows how much happier he’d have been if his Mustang had the 4 speed?
The other very big issue is simply that traffic was profoundly lighter and slower back in the 1930s and 40s, and even the early 50s. Most highways were narrow and typical speeds were 45-50 mph. That all worked with cars that had 3 speeds and low (high numerical) axle ratios.
The problem with 3 speeds is circumstances changed in the 60s and 70s. Freeways, much higher average speeds, and a general speeding up of traffic conspired to make the 3 speed, especially in a weakly-powered car, increasingly uncompetitive and unpleasant. And this issue was of course greatly exacerbated in hilly terrain.
A 4 speed Corvair was profoundly more pleasant to drive than the 3 speed. The compacts of 1960 should have come with 4 speeds. At least they were available on the Corvair and Falcon, although the take rate was low.
Regarding automatics from Europe: the early attempts at automated or semi-automatic transmissions pretty much all had higher number of gears. The automatic available in the Mini/Austin 1100/America was a 4 speed. I’m pretty sure Renault’s automatic in the Dauphine was too. And Mercedes’ only built 4 speed automatics, except for a 3 speed for the larger V8s.
The original Mercedes-Benz 4-speed automatic introduced in 1961 was really a “1+3” gearbox. The transmission shifted automatically up and down through 2 ,3 and 4, but 1st could only be engaged manually by one of two methods: either quickly and fully depressing the accelerator all the way to the floor, or by moving the gear lever to position “2” which limited the transmission to speeds 1 and 2. In this mode, the transmission would start off in 1st and then automatically switch to 2nd (and back to first if one slowed significantly.) Even some of their later boxes worked in this manner as far as the early models of the W201 190 of the 1980s.
To move off both briskly and smoothly, it was best to push the lever into 2, start off and then hand control back to the transmission by selecting 3 or 4.
My Dad bought his first car in the early ’70s when I was about 7 yrs old, a ’67 Ford Fairlane with a 289 V8 and a 3-spd manual with a column shift, the proverbial 3-on-the-tree nonsense. Although he became pretty skilled at driving it I still remember it as not being very smooth with jerky shift motions and that awful mechanical whininig sound. I was soooo glad when he traded up 2-3 yrs later to a ’69 Galaxie with a full automatic – man that was a luxury car in comparison!
I have never been a fan of manual transmissions and am so glad that when I started driving automatics were so good that you never needed to bother with one. My attempt in the mid-90s to learn to drive a stick (persuaded by a friend who was a real manual Honda fanatic and swore up and down that I’d love the experience) ended in nothing but frustration. I was glad to tell him where he could stick his manual-shift cars, and go buy another automatic!
I had no problem driving a Mazda 3 5-speed in San Francisco, my only personally owned stick shift car. But really, the constant clutching and rowing were pretty much pointless and it’s been automatics before and since.
I’ve written about it elsewhere, but the convincing moment was when I backed the Mazda into a space it just fit into on the very steep Masonic Ave above Frederick St. I opened the door to the distinctive smell of clutch and thought I got into the space at the cost of ten thousand miles of normal clutch life.
Fortunately the Mazda was later totaled when an oncoming Buick turned left in front of me like I wasn’t there on San Pablo Avenue in Berkeley. The Buick had a dent in the door and the Mazda was a foot or two shorter than it was before.
I once drove a friend’s Powerglide-equipped Chevy. I rather liked it and marveled at the low gear’s long legs and steady feeling of acceleration to reach 50-60 mph before it shifted to high gear and cruise effortlessly at speed.
Didn’t realize how good the Powerglide was when properly mated to an engine with a good broad power and torque range. Didn’t think it needed that intermediate gear.
On the other hand, my mother’s 1982 Buick Regal with the 3.8 V-6 and 3-sp auto was a gutless wonder. Under moderate acceleration, it spent more time in 1st, shifted quickly from 2nd to 3rd. On the freeway when accelerating to pass, the transmission would downshift to 2nd briefly, then spent most of its time in 1st for its acceleration run. Wonder why it had a 2nd gear in the first place.
I cut my teeth on my dad’s 1968 GMC pickup. The pickup had a 292 inline 6, a 3:54 rear axle and wide-ratio 4-speed manual with the low-low nonsynchro 1st gear. 1st gear was rarely used except when starting with a heavy load; otherwise start out in 2nd (low) gear and work it through 3rd (middle) through 4th (high) gear. In other words it operated like a 3-speed manual.
I remember 2nd gear was good to about 20 mph. 3rd gear had long legs and was good for 20 to 45 mph….just about right for around town, although at 35-45 the engine rpms were running pretty high. Also recall acceleration in 3rd gear between 20 and 45 mph was brisk and lively.
Back in the early 1990s, I had a brief experience with a friend’s Toyota equipped with a close-ratio 4-speed manual transmission. Very different feel compared to the GMC’s wide-ratio 4-speed. Around town, I had to shift more frequently through the gears to keep the rpms up where the engine had power, whereas in the GMC I could leave it in 3rd gear most of the time if I wanted to.
On second thought, I think the GMC’s 3rd gear was probably more comfortable around 20-35mph, but have pushed it around 40 mph.
I ran a hydramatic equipped 63 Holden sedan for several years in OZ I replaced the hydramatic with a more modern trimatic from a torana THM 180-200 because the hydramatioc could not cope with roundabouts and used to hunt gears other than that it was ok the trimatic did transform the car though with only 115 hp on tap the smoother shifting box was much nicer to drive the car only weighed 1090 kgs so performed ok for a 40 year old car I sold it in 03 when we moved countries.
My biggest issue with the 3-speed Nova I had was the column shifter always hitting my knee in 3rth. Never understood why they could not turn around the pattern so 3rth gear would be up. Or was I the only one with long legs driving 78 Nova?
Some years ago, I saw a comment: “If the automatic transmission had been invented first, would anyone have bothered with a manual?” Pretty well sums it up, I’d say.
My 1966 Jeep Wagoneer, built in August 1965, with AMC 327 V8 had a Turbohydramatic 400. It was also available with the AMC 6-cylinger engine.
It had an angle adaptor for the vacuum modulator to avoid hitting the front driveshaft. The transmission was also clocked about 15 degrees CCW to avoid the driveshaft.
One interesting exception to the “no passing gear” problem was the short-lived Buick Dual Path Turbine Drive, used only on the 1961–1963 Special and Skylark. This was a very unusual transmission, bearing no resemblance to Powerglide or the ST-300 that replaced it: The torque converter was “backwards” compared to most automotive torque converters, and in reverse, the turbine was held in place and the stator became the turbine!
The Dual Path transmission had a single planetary gearset with dual sun gears (and a split torque high gear). Its Low ratio was only 1.58:1, and the converter ratio was about 2.5:1, so it had a step-off of close to 4:1, but with the standard axle ratio, you could hold Low up to almost 60 mph, and had kickdown up to about 55 mph. It also shifted faster and more decisively than the three-speed Roto Hydra-Matic (whose second-gear ratio was similar), so it had better performance.
Its downfall was that its torque capacity was fairly limited, and its design wasn’t easy to scale up for larger engines.