As the reader will know, the distributor on a four- stroke engine has essentially two functions (see Fig. 1 for a diagrammatic general arrangement of an ignition circuit for a four cylinder engine):
(a) To distribute the high tension (HT, high voltage) electric charge, delivered by the ignition coil, between the sparking plugs in each cylinder in sequence.
(b) An equally important function is that of the contact breaker to switch the electric power supply to the distributor on and off to generate low voltage (typically 12 volt) pulses. These are transformed in the coil to HT pulses (around 17,000 volt) which are then distributed to each plug as above.
The first function is mechanically simple and is achieved by a rotor and a HT terminal located in the distributor cap for each plug. The second function is more complicated and the following actions must be accomplished during its operation:
1) The 12 volt low voltage on/off switching is traditionally achieved by hard faced contacts which “break” (pull apart) when the distributor timing requires to produce a spark at the plug. In modern engines the switching is done electronically but the principle is similar. When the motor is running this break is normally at a point which is several degrees before top dead centre (TDC), on the compression stroke, typically over a range of 5 to 35 crankshaft degrees (ignoring vacuum advance, see below).
On a typical four cylinder motor the firing sequence is usually, by cylinder number, 1-3-4-2. The slightly odd firing order is for reasons of optimum mechanical engine balance, to minimise vibration effects which become important at medium and high engine speeds.
2) The chemistry and physics of combustion of the fuel/air mixture after ignition in the cylinder dictate that the duration in degrees of crank rotation of the burning process varies with engine speed, revolutions per minute (RPM). Typically, as an average, 90% of a complete burn of the fuel / air mixture at 1500 RPM will take 45 degrees of crank rotation, and at 4000 RPM, 60 degrees. Note that the advance degrees do not increase proportionately as much as the RPM. This is because the time duration of combustion from ignition to peak cylinder pressure changes less than the interval between low and high RPM.
An accurate 100% burn duration is difficult to define exactly, as the burning rate tails off rapidly above 90%. The total burn is accomplished at about 40 degrees after TDC and is around 95 to 97%, the balance being incompletely burned hydrocarbons and carbon monoxide which exit the cylinder with the exhaust gases and a small amount (1 – 2%) which leaks past the piston rings and into the engine crankcase space (called blow by).
As a result of this variation of burning duration with RPM it is necessary to initiate combustion (the contact break or ignition point) earlier at high engine speed than at low. This of course is the reason for ignition advance on the distributor.
3) As the fuel/air mixture starts to burn, typically slightly before TDC, the pressure and temperature in the cylinder start to rise very quickly. At about 7 degrees before TDC the developing combustion is about 5% complete, the pressure typically 200 lb/ sq.in. (psi) from compression of the inlet mixture by the rising piston and the temperature several hundred degrees Centigrade (about 350°C). For the best engine efficiency and power the optimum ignition timing at all engine speeds is to ensure that the peak cylinder pressure and temperature is close to 16 degrees (crankshaft) after TDC. As this exact point cannot be easily determined directly on a running engine, a range of 15 to 20 degrees after TDC is used as a practical design target. The primary aim is to avoid an excessive burn in the period before TDC, so the higher figure is the more usual setting. A too-early combustion pressure development will push harder against the still- rising piston and can generate severely excessive cylinder pressures and temperatures and lead to loss of power, high exhaust valve, cylinder head, and piston temperatures, and a consequent lower reliability.
At the peak cylinder pressure point the combustion is about 75% complete, the pressure in the cylinder is around 600 psi and the temperature 2500°C. At 1500 RPM the corresponding combustion duration, from the ignition point to the 75% point and peak pressure, is about 30 degrees of crank rotation, and at 4000 RPM, 50 degrees. Thus, subtracting 20 degrees after TDC as the optimum peak pressure, the required ignition advance for the two engine speeds are, at 1500 RPM, (30 — 20) = 10 degrees before TDC, and at 4000 RPM, (50 — 20) = 30 degrees before TDC. Note that these advance figures are typical only, but are a reasonable starting point if an engine tuning exercise is contemplated. For a specific engine the optimum advance will be influenced by the engine design and operating conditions (e.g. compression ratio, cylinder bore and stroke, cylinder head design, camshaft timing and lift, mixture richness or leanness, tick-over speed requirements, etc).
4) The accuracy required to set an appropriate degree of ignition advance on a motor is not that great. A plot of engine torque, which is directly related to the combustion pressure developed in the cylinder, versus ignition advance will show a slight “hump”, or maximum, at the “best torque” (i.e. maximum cylinder pressure) advance figure. A variation of +/ – 4 crank degrees each side of the “best torque” advance position will reduce the torque by only 1 to 2 % from the maximum. Recall that ”torque” is the turning effect of the engine, normally measured in lb.ft. or lb.ins., which when multiplied by RPM gives the engine power. Thus the “best torque” at a particular RPM is also the ”best power” at that speed.
If an engine is to be examined on a dynamometer, it is preferable to determine the “best torque” ignition advance at each RPM tested. It is important under these conditions to run the engine for several minutes at each speed to allow it to stabilise and for the “best torque advance” recorded by the dynamometer to be electronically stored for each RPM. High RPM advance values can be determined by careful extrapolation of the numbers obtained at low and intermediate engine test speeds, say 1500, 2000, 2500, 3000, 3500 RPM. The advance amounts should be recorded electronically to an accuracy of +/ – 2 degrees or better. At each speed the ignition is advanced slowly from a slightly retarded value and the rise and subsequent fall of the indicated torque versus advance and RPM values carefully recorded to determine the peak torque position. Recall that a dynamometer measures torque directly. Power is calculated by multiplication of torque by the appropriate RPM.
Modern rolling road testing, I understand, is often performed by a simple RPM scan from the lowest to the highest engine speeds to be tested. This method is cheaper as of course the computer can pick off the intermediate values very easily. It is less accurate than the step-wise system as the temperature and other mechanical conditions of the motor cannot stabilise and the results are “blurred” over the test speed range and the exact “best torque” versus RPM curve is difficult to define.
5) Because of the small variation of power and torque with ignition advance around the “best torque” value, when tuning a distributor by feel or driving response on the road, there is a natural tendency to over-advance the ignition setting to obtain a “lively” throttle response. This generally leads to a small loss of power (normally not detectable) but can also disproportionately increase the combustion pressure and, more importantly, combustion temperature. The resultant signs of excessive temperature on the sparking plugs, and a concomitant tendency toward harsh running, is then often mistaken for the symptoms of a weak mixture or other carburation or fuel problems. If the remedy for the perceived problem is to richen the carburettor setting then the power is reduced still further and the fuel consumption rises. The extra fuel serves to lower the temperature by its cooling effect which thus appears to “solve” the problem, but at a decreased overall fuel efficiency. It is an old but valid axiom, “Ensure the ignition is set correctly before tuning carburettors”.
Overall it is better to under-estimate the advance required rather than over-estimate it. On a rolling road, adopt the principle, “minimum advance for best torque”; a more flexible and sweeter running engine will result, with no loss of power and a lower fuel consumption. The risk of an ignition “over advance” at high engine speeds, when temperature and pressure conditions are anyway severe, is also reduced.
6) In this account so far, no explanation has been attempted of the means by which the required ignition advance is accomplished. As the reader will be aware this is done traditionally by a system of rotating and pivoting bob weights and springs mounted on the distributor shaft which also drives the four-sided (for a four cylinder motor) contact breaker cam and the distributor HT rotor. This mechanism will be described in more detail later. On a modern distributor, of course, the advance curve is managed electronically.
7) On an older system with a mechanically controlled (centrifugal) mechanism the ignition advance is set automatically relative to the engine speed alone. By design, the advance at each RPM is such that a high torque and power output is achieved with a wide open throttle (WOT). No arrangement is made for part throttle running, which a road car experiences for most of its operating life, where the combustion rate is significantly slower than that at full throttle. In effect, the advance value at each RPM, based on a WOT setting, is conservative (low) when applied at part throttle and the torque and power are less than could be accomplished by the use of more advance. Thus the ignition advance is less than the “best torque” value and a higher figure would give more power. The reason for this is that at part throttle the fresh fuel/air charge entering the cylinder on the induction stroke contains a higher proportion of residual burned exhaust gas, which enters the cylinder during the valve overlap phase near TDC when both inlet and exhaust valves are open, and the combustion temperature is lower than at WOT (lower developed power). At full throttle this contamination is around 5% of the cylinder charge; at part throttle it can be up to 15 to 25%. The residual exhaust gas proportion slows down the combustion speed and limits the power developed at the selected RPM. The mechanical advance distributor senses only the RPM so the advance is too late to achieve a “best torque” under part throttle conditions. The outcome is a lower thermal efficiency and increased fuel consumption.
8) This problem is solved by the use of a vacuum capsule device on the distributor to advance the ignition at part throttle operation. This capsule senses the inlet manifold vacuum which is controlled by the throttle position at a given RPM and advances the ignition closer to the “best torque” value. At WOT the manifold vacuum collapses and the vacuum capsule controlled advance becomes zero. The nett advance is then limited to the sum of the static and centrifugal values only. Recall that static advance is a fixed small amount of advance (typically 5 to 10 crankshaft degrees) to give a smooth tick over and easy starting.
A vacuum capsule normally carries its operating range stamped into the body in three figures, e.g. “5 – 13 – 10”. This information is interpreted as follows :
5 ins. mercury vacuum, ignition advance starts, (i.e. large throttle opening).
13 ins mercury vacuum, ignition advance finishes, (i.e. small throttle opening).
10 deg. camshaft (20 deg. crank), maximum advance.
Note that a 20 deg. crank vacuum advance is normally about the most that is used, particularly with a manifold sensing point, as opposed to the position on a carburettor (see below). A more modest experimental starting point could be 10 to 15 degrees, using a lower maximum advance capsule.
— At tick over the manifold depression is above 13 ins. mercury (typically depending on the engine, 15 to 22 ins, with the lower number on a more “sporting” engine set up). The vacuum ignition advance is at the maximum 20 degrees and is additive to the (low) centrifugal advance. On some cars the vacuum sensing port is on a carburettor and is masked by a closed or nearly closed throttle plate so the full vacuum is not applied to the capsule until the throttle is about 5 to 7% open. ln this way the initial vacuum is restricted during starting and at tick over to avoid the possibility of a back-fire.
— Once the car is under way, varying with the throttle opening and engine speed, the vacuum advance will automatically be regulated to a value between 0 and the maximum 20 degrees, which is normally achieved at some point between 1500 to 3000 RPM at part throttle. The high advance level at low RPM is designed to improve low speed acceleration and overall fuel consumption. The vacuum and centrifugal effects are additive (see Fig.2, above).
The maximum total advance occurs then typically around 2200 RPM and is composed of 20 degrees vacuum, plus about 20 degrees centrifugal, plus (typically) 5 to 10 degrees of static advance (i.e. at zero RPM, used to improve starting and tick over). Thus the total advance at 2200 RPM and part throttle is (20 + 20 + (say) 5) = 45 degrees, significantly above the 25 degrees static and centrifugal advance together. This increase shows up as more part-throttle power and liveliness, and a lower fuel consumption (often 10 to 15mpg better on a one litre engine). As the throttle opening is increased further the manifold vacuum drops and the degree of vacuum advance diminishes until just before WOT it becomes zero and the total advance becomes the sum of centrifugal and static values only, as noted above.
Typically this point is around 3500 to 4500 RPM and the resultant advance is then 35 degrees crank (centrifugal advance only).
10) For most older engines, eg. The MG XPAG, even with compression values above standard, say 9 to 1, the WOT advance will give very little performance gain above about 35 degrees crank. This applies even with present day non leaded fuel of octane rating 94 RON (Research Octane Number) or above, as power is developed from the heating effect (calorific value) of the fuel, not its octane rating.
On a high performance motor with a hemispherical OHV cylinder head and a high compression ratio (e.g. XK series Jaguar) a maximum WOT advance of 40 to 42 degrees is required at high RPM as this type of head has a slower combustion than the more usual “bath tub” or “wedge” pattern. Otherwise the lower figure (around 35 degrees) is specified to give better flexibility and smoother running. The guiding design principle in all cases is To achieve a maximum cylinder pressure at 20 degrees after TDC, as noted above.
11) To revert to the mechanical details of the centrifugal advance mechanism….
The above sketch (Fig.3) illustrates a typical design. Several methods are used by manufacturers to achieve the task of varying the ignition advance over a range of typically 500 to 4000 RPM and above, tick over to full throttle.
The basic principle, common to all, is to use a rotating bob weight, pivoted at one end and allowed to move over a small arc under the control of a light spring. The arcuate movement of the weight is linked to a rotor carrying the contact breaker cam so that as the weight pivots under the influence of centrifugal force generated by the distributor rotation, driven by the camshaft, the angular displacement acts to advance the ignition point by making the contact points open early. Typically two weights are used, each with its own spring to control the amount of cam displacement and thus advance. The two springs are normally of slightly different strengths and lengths to give an approximation of the upward convex curve of advance versus RPM which is theoretically required (see Fig.2).
On some cars (e. g. some versions of the MGB) one spring is very weak compared with the other so the single stronger spring alone dictates the shape of the advance versus RPM curve. This is a design option which effectively turns the upward convex curve into a straight line from tick over to maximum advance, in the MGB case at about 3,500 RPM.
The distributor on the standard MG TC for comparison has a slower advance versus RPM curve. Maximum advance is 30 to 32 degrees (centrifugal only with no static) which is not reached until 4,400 RPM, with an intermediate 16 degrees advance at 2,200 RPM.
12) As discussed above the action of the vacuum advance capsule is designed to be additive to the centrifugal system. In contrast with the centrifugal advance mechanism, which acts to advance and retard the contact breaker cam, the vacuum capsule rotates the entire contact breaker plate on which the points are mounted. This displacement in total, in our example, is 10 degrees which equates to 20 degrees of crankshaft rotation, (see Fig.1). In the absence of a vacuum from the engine the system reverts to a fully retarded condition (i.e. zero vacuum advance) under the influence of a spring effect built into the capsule.
13) The active range of advance of both the centrifugal and vacuum capsule systems is, within limits, adjustable. For instance, a faster centrifugal advance can be arranged by using weaker springs, but careful experimentation is necessary to determine by how much. Trial and error methods are very complicated and could lead to a temporary over-advanced condition which could at least degrade the engine reliability or driveability. A properly conducted trial on a rolling road could be a solution but is expensive, and not always conclusive.
Adjustment of a vacuum capsule is even more limited. The active range can be reduced but not increased, except by replacement of the entire unit.
In the case of the XPAG engine, if a vacuum advance ability is required, the entire distributor can be replaced with an MGB unit, but the installation is not simple and modifications are necessary to the standard distributor.
14) In this study I have attempted to address as simply as possible the main elements of the functions of a typical ignition distributor. I have drawn upon my own experience and reading when I have quoted figures but these are indeed mostly just typical and are not necessarily of direct application to a particular engine. Where I have quoted MG TC or XPAG data I believe these to be accurate.
To quote directly from my own experience, I have modified my MG TC to accept a distributor from a MGB with a 5-13-10 vacuum capsule as in Paragraph 8 above. The resultant advance curves are very similar to those of Fig.2, with an almost straight line from the static advance at 5 degrees and a maximum centrifugal advance of 28 degrees (total 33 degrees) at 3,400 RPM. The vacuum pick up point is on the manifold so the full vacuum advance (20 degrees) is similar to that of Fig.2.
The car starts easily with a smooth tick over and it will pull strongly from 1,500 RPM. The performance is lively and I have seen speeds in excess of 70 mph, with good acceleration in the mid range (3,000 to 4,000 RPM). I cannot make it pink on a compression ratio of 8.8 to 1. As a result of these and other changes I get 45 to 50 mpg on a long run.
Finally, I would like to express my thanks to David Heath for his many suggestions and corrections during his review of this study.
John Saunders 17th February, 2011