What are the symptoms of an ignition coil failure

Ignition coils


Information on ignition coils for HKZ ignition systems

Recommended action and summary:

- Preferably use the original, historical BOSCH ignition transformers 0221121001 (black cup housing, Made in Germany or Spain, batch / factory numbers deviating from 908). The coils are seldom defective because they are practically not subject to aging or thermal stress. The coils 02211210-02 to -10 can also be used (conditionally). The spark burn time is approx. 27µs

- BOSCH ignition transformers of type 022121001 can also be used more current Production. The bobbins are supplied with a silver cup housing, have a Made in Brazil sticker and the batch / factory number 908. The bobbins also have an embossed stamp on the underside (to the right of the 908) with the production date, which, according to our recommendations, is after Q3 / 2017 should, according to the stamp numbers 640 or higher. The spark burn time is approx. 30µs. Despite having identical article numbers, the coils are not completely identical to the historical coils, which in practice does not cause any problems.

- In our webshop you will find an alternative (real) ignition transformer with values ​​similar to the original ignition transformers. The spark burn time of approx. 40 µs is slightly higher than with the original ignition transformers, which helps to reduce the well-known partial load jolting in the case of a poor mixture in the partial load range. The ignition oscillogram shows the voltage curve of the original ignition transformer in white and our alternative in blue.

- The BERU ZS-109 ignition coil can be used, see the following explanations, even if, due to the principle, a longer ignition spark (120µs instead of 27µs) occurs with a lower combustion current. Engine test bench tests with overhauled original engines and moderately enhanced engines have shown no differences in the running behavior of the engines.

- The MSD 8222 model with almost the same properties as the BERU ZS-109 model and a spark duration of approx. 100µs can also be used.

- The models MSD 8203, 8202, 8200, which are electrically almost identical to the model 8222, can only be used to a limited extent, but as oil-filled coils may only be installed upright

- BERU has discontinued the ZS-109 model and named the BERU ZS-565 as its successor. The coils we had were filled with oil and are therefore not recommended for overhead installation. Furthermore, the measured values ​​deviate strongly from the model ZS-109 and tests in vehicles were not convincing for us either. For this reason, we advise against using the BERU ZS-565 in connection with HKZ ignition systems.
This is not a general quality assessment because the coil ZS-565, like the previous model ZS-109, was not intended for HKZ operation by the manufacturer. The intended use is transistor ignitions with starting boost.

- HKZ ignition coils and ignition transformers are always operated without a series resistor, regardless of what the manufacturer specifies. Series resistors are only relevant for the function of the start boost with transistor ignitions. With HKZ ignitions, series resistors are counterproductive and reduce the ignition performance



Special ignition coils are used for HKZ ignition systems. The term ignition transformer or ignition transformer would actually be technically correct, since it is a transformer with a transformation ratio of around 1: 100.

The ignition transformer boosts the 280V-480V ignition pulse of the HKZ to an ignition voltage of 28,000V to 48,000V (depending on the speed).

Ignition coils for direct interrupter ignition systems cannot be used, as they counteract the ignition pulse of the HKZ due to their construction with highly permeable laminated cores. The air gap that is often present in these ignition coils is also counterproductive. If there is any arcing at all, its burning time is too short or too weak to reliably ignite the gasoline-air mixture.

Another criterion for the continuously charging HKZ systems (Bosch 3-pole and 6-pole) is the risk of skipping a charging spark. The above-mentioned HKZ systems gradually charge their ignition capacitor (in the HKZ device) approx. 3000 to 4000 times per second. This charging current flows through the ignition coil and, if the coils are unsuitable, can trigger smaller charging sparks that lead to misfires.

The 8-pin Bosch devices (0227300003) and all the accessory systems I know have a charging diode that bypasses this problem. For this reason, these systems are somewhat more tolerant with regard to the selection of the ignition coil. Unfortunately, the charging diode suppresses the "spark tail" of the ignition spark, which leads to a slightly shorter and weaker spark. In the original, the ignition coil with the number 0221121010 (for single spark charging) was used for the 8-pole Bosch device, which as a rule should not be used for the 3-pole and 6-pole devices. Conversely, the coils of the 3- and 6-pole devices can also be operated with the 8-pole version without any problems.

Left (or above) 3-pole and 6-pole with long spark burn time. Right (or below) 8-pin and accessories with a short burn time. The point in time of the spark flashover is also clearly visible as a small disturbance in both images.


The original historical Bosch ignition coils are very seldom defective, as only little energies are converted in them (in contrast to transistor ignition systems). Although the original bobbins were reissued by Bosch in the Brazilian plant, there are many reasons why customers often want to use an alternative bobbin.

Unfortunately, there is in principle only a small selection of suitable ignition transformers. You can find a model with similar characteristics in our webshop.

Experience has shown that the BERU ZS-109 and MSD 8222 models (approx. 60-70 euros) can be used as an inexpensive alternative. These coils are high-performance ignition coils with low primary resistance, almost identical to that of the original.

Unfortunately, both the primary and the secondary inductance are in principle much too high (air gap and highly permeable laminated core), which in principle contradicts the idea of ​​an HKZ ignition.

Voltage curve Original Bosch

BERU voltage curve

The yellow line in the above oscillograms represents the ignition pulse signal. The ignition point is the fall of the yellow signal. A short time later, 2 faults are visible on the yellow line. The 2nd fault is the flashover of the ignition spark. If you measure the differences more precisely, it becomes apparent that the spark flashes over 5 microseconds later with the BERU ignition coil, which is irrelevant (approx. 0.1 degree distributor shaft at 6000 RPM).

The later rollover fits well with the theory, since the high inductances counteract rapid changes.
The blue primary voltage curve on the ignition coil clearly shows that the spark on the BERU ignition coil burns longer (approx. 3x longer) because the stored energy is released over a longer period of time.

However, this actually undesirable behavior does not affect the maximum ignition voltage. This is roughly the same for both coils, as the voltage would theoretically increase to infinity if it did not find a way to flash over. With the BERU coil, however, it takes a little longer because the high inductances prevent the voltage from building up too quickly.

Nevertheless, the law of conservation of energy also applies here: The current flow during the sparkover is less with the BERU coil than with the Bosch coil, but the spark burns correspondingly longer, as the following oscillograms show.

Current curve Bosch (blue line)

BERU current curve (blue line)

The difference in the current curves is considerable and not surprising if you consider the high input inductance of the BERU coil.

Finally, of course, the question arises as to whether the BERU coil can withstand the voltages that occur up to 40,000V, which is not clear from the known technical data. In this case, however, experience helps and it says that I have not yet seen any defective BERU-ZS109 (from HKZ operation) on my laboratory bench. There are also customers who have been using this coil in vehicles for around 15 years. For this reason, one can assume (albeit somewhat unscientific) that the BERU coil can be used over the long term and will not be damaged by improper operation. The 8222 MSD coil is designed for voltages up to 45000V anyway, so that no long-term damage from HKZ operation is to be expected.

The comparisons between primary and secondary development are also interesting. In the attached oscillograms, the secondary curve was measured by means of a capacitive pick-up.

Bosch (blue primary / yellow secondary)

BERU (blue primary / yellow secondary)

The oscillograms confirm the previous findings (Bosch shorter and stronger / BERU longer but weaker spark).

It is interesting here that the ignition spark does not break off when the primary voltage crosses zero (Bosch approx. 15uS / BERU approx. 60us), but continues to burn (Bosch approx. 25-35us / BERU around 100 us).

These ignition patterns cannot be compared with those of simple coil or transistor ignitions. Many motor oscilloscopes are not even able to correctly display the ignition oscillograms of HKZ ignition systems. It is also wrong to assume that 12V (as with transistor ignitions) should be applied to terminal A (15). For the aforementioned reasons, misdiagnosis is common

Note: uS = microseconds

The last and most important question is: what does that mean in the vehicle?

Technical conclusion:

The "sparkover" of the ignition spark is not influenced, since the sparkover voltage reaches the same absolute values ​​in both types. This means that both coils are able to produce a spark even with sooty spark plugs, bad ignition cables or high mixture pressures.

The total spark energy is naturally almost identical (law of conservation of energy). The Bosch coil releases this energy faster with a stronger current flow, the BERU ignition coil with a weaker current flow, but over a longer period of time.

There is no danger of ignitable charging sparks with the BERU coil, as the charging voltages are only around 2-3kV.

Now one could make further theoretical considerations about Poisson equations and the ignitability of real gas mixtures, which would certainly be interesting but more of an academic nature.

It is easier to test both coils on an engine test bench or in the vehicle under the same conditions, which many customers have done and could not find any difference in engine performance.

Advantages of Bosch coil: Very powerful spark (large current) / Disadvantage of Bosch coil: Very short spark

Advantages of BERU / MSD coils: Very long spark burning time / Disadvantage of BERU coil: Ignition spark is less powerful (low current)


The different ignition behavior of the Bosch coil and that of the BERU / MSD coils can be clearly determined both theoretically and by measurement. In real use in the original ignition systems of the Porsche 911, however, there are no significant differences in engine performance. Both ignition patterns show advantages and disadvantages that cancel each other out and are equally suitable for the desired purpose, provided the engine and its components are in good condition. If the condition is not optimal or if the engine is tuned, the situation may look different and the optimal coil may have to be found in a practical test.

Note on the series resistor:

The BERU ZS-109 ignition coils indicate that they should only be operated with a series resistor. This note only relates to operation in simple transistor ignition systems (Pertronix Ignitor I, 123-Ignition, Fulmax), as the charging currents would be too high. Note: The aforementioned systems should not be used in the Porsche 911 without an HKZ, as the ignition pattern for the engine is not optimal. In connection with an HKZ, however, it is possible and also makes sense.

With Bosch HKZ ignitions, no series resistor must be used in conjunction with the BERU ZS-109 ignition coil!

Other alternatives.

Since BERU has discontinued the ZS-109 ignition coil, the MSD coils and the ignition transformer available from our web shop remain as alternatives.

On the accessory market (mainly USA) there are a number of other coils that are offered explicitly for HKZ systems. Presumably these coils work well, but all my measurements so far clearly showed that these coils are high-performance ignition coils and not ignition transformers. Furthermore, these coils are often filled with oil instead of asphalt (bitumen) (you can hear it when you shake it). I doubt whether these coils are suitable for the usual overhead assembly in the long term. The risk of leaking oil is relatively high.

An exception is the MSD Blaster High Vibration Ignition Coil (MSD-8222), which is filled with epoxy and is available in German stores for around 70 euros. The original sticker is partially removed from this reel and the reel is sold as a special HKZ reel. Basically, however, it is also a high-performance ignition coil.

Characteristic values ​​for comparison:

The original Bosch coils have the following (measured) characteristics (primary resistance: about 0.6 Ohm / primary inductance around 0.2mH / secondary resistance: about 1.5kOhm / secondary inductance: about 1H / transmission ratio about 1: 100).
The high-performance ignition coils are easily recognizable by their significantly higher inductances (primary from 5mH / secondary: from approx. 18H) and the high secondary resistance (from 4kOhm).

Likelihood of confusion:

Unfortunately, ignition coils and HKZ ignition transformers are almost identical in appearance. Please use the item number of the ignition coil to check whether it is suitable for use on the HKZ. In the case of Bosch HKZ in Porsche vehicles, to the best of our knowledge, these are at least the following Bosch article numbers: 0221121001/0221121006/0221121009 (without guarantee). For the 8-pole HKZ device: 0221121010/0221121007/0221121004/0221121002/0221121005), the latter can also be used for 3- and 6-pole HKZ, or vice versa.

Defective ignition coils:

Defects in ignition coils are difficult to determine because simple measuring methods fail or are life-threatening due to the high voltage that occurs.

The classic ignition coil defect is an insulation fault between the primary winding and the secondary winding or a short-circuit in the secondary winding

Often this error only occurs when the vehicle is at operating temperature, so that the vehicle stops after a few kilometers with misfiring, but starts again without any problems after a cooling phase.
Unfortunately, a defective ICC often shows the same symptoms, so that it cannot be clearly clarified whether the ICC or ignition coil (or both components) are defective.

In many cases, a defective ignition coil also causes the HKZ device to be defective due to an internal flashover, so that only a repair of the HKZ with simultaneous replacement of the ignition coil leads to a successful repair.

Backward rollover:

Reverse flashovers are one of the most common causes of defective HKZ units.
In this case, the ignition voltage (up to 40,000 V) cannot flow through the spark plug due to a further defect in the ignition system, but looks for another route to the vehicle ground. Since the spark can be up to 40cm long at 40,000V (no typing error), the possible spark gap courses are diverse.

Unfortunately, one of the shortest routes is the one from the ignition coil dome, over the mostly heavily soiled ignition coil cap, directly to terminal A and then into the HKZ. The following photos illustrate the effect:

If the ignition voltage flashes back into the HKZ, as in the picture above, it will be severely damaged after just a few flashovers.

Where the ignition voltage strikes back depends on which is the shortest and most conductive path to the vehicle ground. For this reason, we recommend cleaning the ignition coil and the connector on the ignition coil thoroughly and ensuring that the connector is firmly and smoothly seated.

It should also be ensured that the connection cables at terminal A are further away from the flange of the ignition coil than the connections at terminal 1. If the ignition voltage flashes over to terminal 1 in the event of a reverse flashover, the HKZ will not be damaged.

In the following picture, a soldering lug was screwed onto terminal 1, which is significantly closer to the flange of the ignition coil than terminal A. The rearward flashover is therefore relatively safe on terminal 1.

The following picture illustrates the assembly of the solder lug on terminal 1:

If you want even more security, an earth cable around the spark plug flange is recommended, which is connected to terminal 1 of the ignition coil.


Causes of the back rollover:

The backward flashover always occurs when the spark cannot flash over on one of the spark plugs. The voltage then rises from approx. 15,000V with a normal flashover at the spark plug to over 40,000V and can thus bridge an air gap of up to 40cm.


  • Spark plug connector loose or fallen off
  • Plug on the ignition coil or on the distributor cap loose or fallen off
  • Burned interference suppression resistors in the spark plug connectors (normally 3kOhm or 5kOhm) or on the distributor (normally 1kOhm)
  • Burned interference suppression resistor in the distributor finger (normal 5kOhm)
  • Burned contacts in the distributor cap, on the distributor finger or on the head contact of the distributor finger
  • Burned out spark plugs
  • Broken spark plugs (break in the ceramic)
  • Broken ignition cable (break of the copper wire inside)

Testing options:

Faults in the ignition cables, spark plug connectors and spark plugs can easily be checked with a commercially available ohmmeter. The component must be completely removed and the resistance measured. Components that are not suppressed have a resistance of a few ohms (approx. 5 ohms). Interference-suppressed components have resistance values ​​of 1kOhm (connector on the distributor), 3kOhm (spark plug connector) or 5kOhm (distributor finger). The resistance values ​​can vary depending on the model and not all vehicles have extensively suppressed ignition systems. Defective components usually have resistances of several 10 kOhm up to the megohm range, so that a fault can be detected with relative certainty.

If this does not lead to success, only a motor oscilloscope can help, which can display all 6 ignition patterns of the individual cylinders at the same time. The comparison of the ignition patterns usually provides at least the ignition path that is responsible for the flashover very quickly. If none of the cylinders ignites, the fault is to be found in or in front of the ignition distributor.

If a motor oscilloscope is not available, an ignition stroboscope (with contactless signal pick-up) can be used in each ignition path and between the coil and distributor to check whether ignition is taking place.
If you take a closer look at the sequence of flashes while idling or when the ignition timing marker is flashed, individual misfires can also be detected.

Warning notice and disclaimer:

The above information is given to the exclusion of any guarantee or liability. There is no claim to correctness. Please note that any change to the vehicle represents a design change and is the sole responsibility of the person carrying out the work.

We generally advise against working on live ignition systems. Please keep a distance of at least 1m from all live parts of the ignition system and do not touch the ground of the vehicle.

Laypeople are generally advised not to work on motor vehicles and, in particular, on ignition systems. For this reason, the above information is only intended for master vehicle mechatronics who are familiar with the functionality and dangers of HKZ ignition systems.