Capacitor Discharge car ignition systems

Dear Sirs

Re: CD ignition problems. (letter P.63/March 1980)

I was most interested to read the letter from Mr. D J Bruyns since I have been building and using CD systems for eight years. Last year I wrote to a well known manufacturer with some constructive criticisms of his design. A few days later I received the offer of a job!

The accompanying letter contains, in detail, my experience of the practical problems associated with the use of a CD system. It would be nice to think tliat you could find space to print it in full, but my conclusions are as follows:

1. The CD unit itself is not always to blame for spurious faults, although careful attention must be given to the choice of suitable components; in particular, the thyristor and the capacitor which must tolerate very fast risetimes, high voltage and high temperature.

2. The high output voltage from thie coil attached to a CD unit is susceptible to leakage and arcing caused by damp, dirty or cracked H.P. leads, faulty coil or suppressor caps.

3. CD ignition does not appear to cope with a weak mixture any better than standard Kettering ignition, possibly because although the voltage can be higher, the duration of spark is muoh less than that of the standard system.

4. The CD system does seem to cope with a rich mixture, perhaps because its fast risetime allows it to fire a fouled spark plug.

Note that points 2. and 3. can result in misfiring from a CD driven system but apparently acceptable performance from standard ignition.

I wrote to Ray Marston in March, care of another magazine, however he has not had time to reply as yet. Perhaps you could send a copy of this letter to him?

Mr. Bruyns "lathe test" of his CD unit was uninformative because the performance of the CD unit is dependent upon the load which the coil sees. In turn this load is determined by the electrical chain up to, and including, the spark plugs and by the engine conditions (load, speed, timing, mixture, compression pressure and temperature).

Consequently, the CD unit performance must be considered only in relation to the performance of the engine (system) as a whole. Perhaps you could observe your battery powered oscilloscope whilst being driven up a hill, but the normal solution is to use a rolling road dynamometer. The following information is derived from my practical experience:

Crossfiring is usually caused by open circuit suppressor cap resistors, open circuit H.T. leads or too large an air gap between rotor arm and cap electrodes. As a rough guide, there should be less than 20k‡. resistance between cap electrode and plug electrode but more than l0k‡ to limit interference.

Suppressor caps usually fail by corrosion of internal metal parts. I heard that one manufacturer stated that open circuit suppressor caps were quite acceptable because the spark would jump the gap! If I may state the obvious, several kV are lost across the gap; massive interference is produced and the heat generated quickly widens the gap. (One faulty cap of mine actually set fire to the H.T. lead!) I notice that, since the introduction of the trades descriptions act, suppressor caps no longer bear a resistance value.

There are numerous causes of misfiring. Some (most) of them are caused by faults external to the CD unit but appear only when that unit is in operation. CD ignition can damage coil insulation, especially when plug gaps are widened, since the voltage can rise so much higher, so much faster than with standard ignition.

Damp leads, cap or coil can produce a path to earth which the higher CD-produced voltage will take whilst the Kettering will not. Similarly, a loose H.T. lead can cause burning and carbonisation inside the coil nose; effectively shortening the path to earth. The coil output lead should be well secured to prevent internal arcing and the L.T. tabs should be well insulated and carefully bent away from the coil nose.

There is a lot of nonsense talked about spark plugs."Old Fred, up the road, has been using the same set of plugs for ten years and never a bit of bother. Mind you, the carburettor must be worn because he gets only 15mpg!" As far as I can determine, three things happen to spark plugs: The electrodes wear away by electro-erosion. The short-duration, bi-directional spark given by CD will reduce this erosion, in theory. I have never been able to prove it in practice. The problem would appear to be academic in view of the next factor:-

Apparently the various products of combustion, including lead salts, leach into the ceramic insulator until it becomes a conductor, instead. Useful life of a plug is limited to about 10,000 miles, according to manufacturers who, admittedly, have a vested interest! The third factor is the build up of carbon deposits around the electrode. CD ignition seems to cope with this type of fouling slightly better than standard ignition. Consequently CD can work better in a worn engine or one which is running too rich a mixture. Alternatively, in a good engine, a slightly "cooler" range of plug may be used to advantage. A "cooler" plug reduces pre-ignition (electrode too hot - not to be confused with "pinking" caused by incorrect timing or low grade fuel) and may last longer. CD combats the fouling of the cooler plug which can occur in traffic.

Inside the CD unit, the invertor circuit seldom gives trouble, provided that it is correctly designed and uses components of adequate rating. The trigger circuit, too, gives little trouble although I have found that the resistor (usually 39‡ ), which supplies wetting current to the points, can desolder itself from the board, consequently I have always used 150‡. If a transistor is driven from the points its base-emitter junction must be protected.

The capacitor is repeatedly charged to well over 300 volts and discharged at a rate of several volts per microsecond. Very few capacitors will tolerate that sort of treatment, although some will appear to whilst suffering gradual degradation. Metallised film types, for instance, undergo two distinct failure modes: fast voltage fall rate punches current through the film and, although the film is self-healing, a pulse is lost on each occasion. Naturally, there is a limit to the number of punches-through that the capacitor will tolerate: gradual decrease in capacitance is followed by short circuit. The high current density (typically ten amps) destroys the connection between the wire and the extended foil (the point of highest resistance) resulting in open circuit. I have observed both failure modes in the tubular capacitor normally used by the home constructor. The best solution I have found is the Siemens type B32562/400v/O.47µF which is rated at 7v/µS and happens to be smaller and cheaper than its tubular counterpart.

The thyristor can give rise to misfiring since it is subjected to the same voltage, current and dV/dT as the capacitor and is often allowed to reach too high a temperature. The 2N6240 is used by a renutable CD manufacturer and seems to cope well. Another manufacturer recommends the protection of the thyristor by the series connection of IM‡ and 10nF between anode and cathode and the addition of l5µH between anode and the original circuit. The thyristor should now withstand the high backswing voltage generated when an H.T. lead is disconnected. Nothing will protect the thyristor if the output from the coil is allowed to arc back to the terminal connected to the output from the CD unit.

Finally a few general notes: manufacturers recommend an increase in plug gap of about 20%. This idea seems reasonable since an increase of spark gap must also increase the probability of there being an ignitable mixture within the gap. However, a standard ignition coil may not cope with the resultant increase in voltage so a better coil should be considered.

Mixture, in my experience, can be set slightly rich to give extra power. However, even my latest "extended duration" system seems unable to cope with a weak mixture.

CD ignition will not cure faults such as incorrect timing or worn shock absorbers (the literature gives the impression that it is a universal palliative!) but it may ignore worn points, leaky condenser or even a partly blocked air filter. It is worth noting that, because of the fast risetime, a CD spark will usually occur about 6oµS in advance of that produced by standard ignition. At low engine speeds the effect may be unmeasurable but at 6000rpm the ignition is effectively advanced by about two degrees. This should not cause problems unless you drive a racing car, since individual distributors may vary more than this.

As a last suggestion, your car radio, tuned between stations, can be a most useful test instrument. If, during misfiring, interference is heard then the spark energy is obviously being lost outside the CD unit. If, however, the impulse were simply missing, no interference would be produced.

Send this page address - CLICK HERE - to a friend !

Back to Index