Self-Starting Devices for Explosion Engines
by Herbert L. Towle

The reasons for employing self-starting devices with explosion engines are threefold. The first mat be said to be simply the desire for perfection--in other words, the desire to start these engines with approximately the same ease as a steam engine is started by opening the throttle. In the last analysis this amounts simply to a desire to save the muscular effort of turning the crank against the compression. The second reason for desiring a self-starting device is found in the difficulty or actual impossibility of starting large engines by hand. The third reason lies in the fact that these engines do not always start at once, and may have to be turned a number of times before the first impulse takes place. In reality, this reason, like the first, is a matter of physical effort, since the self-starting device presumably does nothing to improve the action of the carbureter and ignition, but simply relieves the operator of the necessity for exhausting physical effort.

From any logical standpoint, the last reason above mentioned is entirely irrational, and would lose its weight at once if the carbureter and ignition always functioned properly. Again, if these organs do not perform their duty at once, it is a plain waste of time to continue to turn the engine in the hope that something may occur to make them work. The only real excuse for this latter procedure is found in the fact that a carbureter will sometimes fail to give a suitable mixture when primed or when the suction on it is negligible, as is the case when cranking by hand, whereas the same carbureter will give a good mixture when the engine turns fast enough to create an appreciable suction. In cases of this class the function of the self-starter is not simply to turn the crank past the ignition point against compression, but to impart to the engine a certain velocity at which it is found that the carbureter takes up its work. Aside from this particular consideration, it is an obvious futility to continue running the engine by a self-starting device, and if the motive power is compressed air or anything else whose capacity is limited (and this includes every class of self-starting device in practical use), it is evident that the proper solution is more gray matter and not a self-starting device per se.

Owing to the physical necessities of the case, self-starting devices have been used for many years with large stationary engines. The earlier forms of these for the most part took the shape of explosive cartridges which were introduced into the cylinder through a special plug, provided for that purpose, and were discharged usually by striking a percussion pin. Before introducing the cartridge the engine was turned to a position slightly after the beginning of the impulse stroke, and the explosion of the cartridge furnished the energy to turn the fly-wheel through the exhaust, suction and compression strokes immediately following. As these devices provided only for a single cycle of strokes, so that if the engine failed to start at once the process had to be repeated, their utility has been very limited. For marine use they may be left out of consideration altogether.

A substantially equivalent device sometimes used with large stationary and marine engines is simply to stop the engine by interrupting the ignition while the carbureter or mixer continues to function, and to start by turning the engine with the inlet valve open to about the position above described and then producing a spark by manually snapping the igniter. This method is properly applicable only to gas engines, since in a gasolene engine the gasolene vapor is liable to condense in the cylinder when the engine gets cold, thereby weakening the mixture below the ignition point.

Automobilists have discovered that under any conditions of temperature which do not cause condensation of gasolene in the cylinder, a 4-cylinder four-stroke engine or a 2-cylinder two-stroke engine may be started by simply retarding the spark timing lever beyond the position corresponding to the "top center" of the cranks. This is because such an engine will almost invariably stop with the cranks about mid stroke owing to the influence of the compression, and if the spark lever be retarded sufficiently, a spark will occur in the cylinder whose piston is in the impulse stroke position. An engine may frequently be started in this manner after standing overnight or even two or three days, and simply as a matter of convenience this trick is deserving of mention, although its limited applicability bars it from classification as a regular self-starting device.

Contrivances have frequently been proposed and constructed for starting engines by means of coil springs, which, like clock springs, may be wound up by the engine when the latter is running, and which, by means of ratchets and other devices, engage the crank-shaft when it is desired to start the engine. Devices of this class, however, are practically suited only for small engines, and therefore have only the least important of the three possible justifications for their existence. In the present state of carbureter and igniter it is hard to believe that the added expense and complication which they involve will generally be considered justifiable.

It was pointed out above that a gasolene engine, even if it contained a good mixture on stopping, could not be left standing indefinitely with the assurance that the gasolene would not condense. To overcome this difficulty is the object in certain devices somewhat used abroad. In this class of apparatus a pump is used to force a supply of fresh mixture to the cylinder in the working stroke position, and this fresh mixture is ignited by manually producing a spark. in one of these devices, the Mors, a special surface carbureter is employed in connection with an air pump. This pump draws in a mixture of vapor and air from the surface carbureter and forces it to the cylinders on the down stroke, after which a spark is produced in the cylinder in the impulse stroke position. In another device, the Daimler, a special spraying carbureter is mounted on the engine and a distributing valve is set by the operator to deliver the mixture to the cylinders on the down stroke. An indicator attached to the spark timer shows in which cylinder the spark is to be produced. With the above two forms of apparatus, provision is made for two impulse strokes instead of one, since the second cylinder receiving the pump charge compresses and ignites it in its turn.

In one American car, the Harrison, acetylene gas is substituted for gasolene, with the result of much more energetic explosions.

From a broad point of view it can hardly be questioned that the best agent for starting explosion engines is compressed air, since it represents the actual storage of a certain, though limited, amount of energy, which is available to turn the engine in case the first attempt to start is not successful. It is true that to use compressed air one must have an air pump, but this is a far simpler mechanism and far preferable from almost any point of view to the trappy spring engines, special carbureters and hand pumps, and other makeshift substitutes. There are three principal ways of applying compressed air. One, which is extensively used in large multi-cylinder stationary engines, is to employ one cylinder only as an air engine and use it to start the condenser. After the other cylinders have taken up their cycle of operations, the compressed air is shut off from the third cylinder and it begins to work with the others. To accomplish this result, it is necessary to provide a special inlet valve for the compressed air, and to modify the action of the regular inlet and exhaust valves. It is possible to admit the compressed air to the cylinder on every down stroke and open the exhaust valve on every up stroke. If the regular inlet valve is automatic, this is done with perfect ease b y simply mounting the exhaust valve cam of that cylinder on a sleeve on the cam-shaft and shifting the sleeve, or by making all the cams wide enough so that the whole shaft may be shifted to bring special exhaust cams into play for the starting cylinder. The same shift of the exhaust cam-shaft will naturally bring into play a cam for opening the compressed air inlet valve. In case the inlet valves are mechanically operated by a separate cam-shaft and it is not desired to shift both shafts, compressed air may be admitted on the impulse stroke alone by arranging the special exhaust cam to open on the compression stroke, thereby relieving the cylinder of the pressure which would otherwise occur.

A modified arrangement is that applied to the four and six-cylinder Fiat engines. In these engines the exhaust valves are not distributed by the starting arrangements, but the inlet valves, which are also mechanically operated, are put out of action. The compressed air is admitted by special valves over the inlet cam-shaft and the same lever, which opens the main valve between the compressed air tank and the engine, acts likewise to shift the inlet cam-shaft to bring the compressed air valves into action. These compressed air valves are opened during the impulse stroke and the exhaust takes place in the regular manner. When the engine has attained speed sufficient to cause regular suction through the carbureter, the lever is returned to its normal position and the engine takes up its cycle of operations before it has lost its momentum. The air is compressed by a small water-jacketed compressor mounted at the front end of the engine where it is run by an eccentric from the main shaft.

The third of the commoner methods of using compressed air is to employ a small air engine to turn the crank-shaft. This dispenses with the special valves and cam-shaft arrangements above described, but at the very considerable cost of an extra piece of machinery. Several devices of this class have been tried. one type having a small engine of conventional design geared to the crank-shaft so that several turns of the engine shaft produced one turn of the crank-shaft. Another type of engine has a piston connected to a rack which turns a spur gear attached to the crank-shaft through a ratchet. The piston naturally has a long stroke and is worked by automatic valves. The arrangement, although operative, seems about the least desirable of the possible compressed air devices.

(Excerpts transcribed from The Rudder, July 1907, pp. 627-629. )

[In five years time, Charles Kettering, of the Cadillac Motor Car Co., developed and patented the Delco electric ignition and self starter, which really became the "standard of the world." In short order all gasoline engines were liberated of their extra cams, and plumbing and other devices. - GWC]

[Thanks to Greg Calkins for help in preparing this page. —LF]

Hydroplane History Home Page
This page was last revised Thursday, April 01, 2010 .
Your comments and suggestions are appreciated. Email us at wildturnip@gmail.com
© Leslie Field, 2000