auto

Auto trans consume more engine power than manual trans. They have a pump that pumps around a lot more fluid than circulates in a manual trans. The power consumption of an auto trans might be 20% of crank hp, whereas a manual trans might consume 10%. These estimates may include other frictional losses and accessory pull; it is the difference between the two that is important. Auto trans also usually weigh more than manual trans, but not dramatically more and the auto trans version of a flywheel, the flexplate, is lighter than a stock flywheel. To be fair, you would have to weigh the entire assemblies.

The torque converter in an auto trans is sort of like the clutch in a manual trans. However, it can do a pretty neat trick, which is to multiply torque. All geared trans, manual and automatic, multiply torque in the lower gears through gear multiplication (gearing is lower than direct drive of 1:1, which usually appears around 3rd in an auto if a 4 speed and 4th in a manual if a 5 speed). The torque converter multiplies torque by slipping. This is sort of like slipping a clutch on a manual, except that it occurs without damage to the clutch and flywheel. It is designed to slip. Torque converters usually have a multiplication rate in the 1.5 to 2.2 range. The multiplication depends on stall speed. Stall speed can be measured and rated in a lot of different ways that can be misleading. The only fair way to rate it is to see what rpm you get
when you put your foot on the brake and bring the vacuum up to zero; in other words, no boost.

What is happening is that the engine is loading up against the trans. This makes heat, which the auto trans absorbs. There is lots of fluid in there that gets circulated and usually gets routed through at least one cooler. The stall speed may be anywhere from 1200 rpm to 5000 rpm, but on street cars it is usually in the 1200 to 2500 rpm range. Again, this rpm would be what is achieved when the brake is held and vacuum is brought up to the zero boost point. Obviously, your engine makes more torque at 2000 rpm than it does at 1000 rpm. This is the advantage of the auto trans. Can it be achieved with a manual trans? Yes, but it is more difficult to modulate the desired rpm precisely and causes a lot of wear on the trans. So when you leave the line with the auto, you are not only taking advantage of the gear multiplication just as on a manual, but you are also using rpm to your advantage.

Some people who drag race put very high stall speed converters on their cars. They usually do this because they have a race motor that makes very little low end torque. The idea is to use the stall speed of the converter to get the motor into the beginning of its torque peak. 4000
rpm stall speed converters are not good for street driving. They make a lot of heat due to all the slipping. And on the highway, if you are going 60 mph and give the throttle a bit of gas to accelerate, the rpm will rise up to around the stall speed of 4000. So there will be a
noticeable fuel economy loss also. My car had a stock stall speed of about 1800 rpm. Unfortunately, my turbo was not coming into boost at 1800 rpm, so I was nowhere near my torque peak when the converter stopped slipping. The result was a bit of a bog and a bit of lag. I changed the converter to one with a rated stall speed of about 2600-2700 rpm. Using the zero boost method, the stall speed was really about 2400 rpm. This worked great, because even my larger turbo was making a pound or two of boost at this rpm. So when the converter stopped slipping, I was making boost and the boost just keeps coming. I still don't have V8 like torque coming off the line, but it is a lot closer to that feel than it used to be. With real small turbo cars, the effect is not as noticeable, since they make boost at much lower rpm.

Torque converters are very interesting. They are not constant. The stall speed depends on the amount of torque fed to it. Put your converter behind a V8 and it might have a stall speed of 5000 rpm, not 2000 rpm. Measure your stall speed during circumstances other than the
zero boost test, and it will vary. Measure it by the coming off the line test and it will usually be a few hundred rpm higher than the zero boost test. That is because the turbo is making some boost and that boost makes torque and that torque raises the stall speed. Remove the kickdown cable of the automatic and drive the car and you will get a different stall speed in the upper gears. For example, in 3rd or 4th, your stall speed might be 3500 rpm, not 2500 rpm. This is because the engine is much more loaded up and is producing more torque; it can make 15 psi of boost at 2500 rpm on the highway, but not at 2500 rpm in 1st gear coming off the line. There are charts and formulas that you can use for calculating your stall speed under all sorts of conditions, for sizing converters, and so forth.

The torque converter slippage makes dyno runs more difficult. Usually the dyno wants 3rd gear for a consistent power curve. In an auto, it is hard to hold a gear. If you pull the kickdown cable, the stall speed will rise even further, making even less of a measurable power curve. Even if you measured off the line, you cannot generate a power curve for below the stall speed. If this is 2500 rpm, that is where your power curve will start. But since a good dyno reading requires a higher gear than 1st, the stall speed rises when 2nd is locked in, or when 3rd is locked in, making the power curve that is measurable start at even higher rpm.

Finally, some auto trans have lock up torque converters. Depending on certain inputs like rpm, throttle position, etc., the converter will lock up and stop slipping altogether. This promotes fuel efficiency. It can also be irritating on a small displacement motor with fairly low off
boost torque. As the turbo goes in and out of boost under some conditions, the converter locks and unlocks. Sort of like gear hunting on some autos on small motors. Normal slippage might be 5 to 10%. This is part of the reason why auto trans generate heat and have lower fuel
economy.

There are a couple of ways to make a higher stall speed converter. A stock one can be cut open and "loosened up" inside by trimming the vanes and then rewelded, or a new, smaller converter can be fitted. The first method should be used only for very modest increases in stall speed, like a few hundred rpm.

Auto trans are luxury items first. They are designed for soft shifts. Soft shifts are slower shifts. Performance automatics or automatics fitted with performance valve bodies have firmer shifts. Real firm shifts are pretty quick. The tires can even chirp. So non performance autos are almost always slower than manual trans. Performance autos, even though they usually have one less gear (autos are now going to 5 speeds, but manuals are going to 6 speeds), are sometimes as fast as manuals. Under extreme conditions, they are often as fast or even faster. Autos absorb drivetrain shock that comes from clutch dumping and tend to handle high power with fewer problems. Typically, drag racing autos are quicker than manuals in the quarter mile. The power consumption of the automatics, however, typically means that they have slightly lower trap speeds. These are generalizations; there may be some autos that lose nothing in trap speed just like there are a lot of autos that will not be quicker than a manual. A lot depends on how idealized the combinations of engine, power curve, final drive, trans gearing, and
torque converter is.