Radio interference caused by the ignition circuit on the ...



Radio interference caused by the ignition circuit on the MZ202 engine.

Author: Oskar Stielau

Email: oskar@hfpower.co.nz

Last update: 30 July 2007

Background

The MZ202 engine is a two cylinder, two stroke built by Compact Radial () in Canada. It produces 60hp and is a popular choice for microlights and single seat gyrocopters. It is also the standard engine used in the Mosquito ultralight helicopter ().

I built a GyroBee with a MZ202 engine. It’s a really nice engine, always starts first time and runs very nicely. I prefer it to the 503 I had on my previous gyro.

Since a radio is mandatory where I fly, I bought and installed a handheld ICOM A6 transceiver on the gyro. Using a good external antenna the radio works very well, as long as the engine is not running. The first time I tried to use the radio while the engine was running was a completely different story. Transmitting was fine, but the radio picked up so much interference from the engine that reception was hopeless. As soon as the engine stopped running, radio reception was fine again.

On a MZ202 engine the interference can come from one of 2 systems, namely the ignition system and the alternator (charging) system. The ignition system is the obvious source of interference, and was where most of the effort was focused on.

1 Magneto vs CDI

Here’s a brief explanation of the differences between a magneto system and two types of CDI systems.

Magnetos are electromechanical devices that are driven by the engine. They are self contained, the input is a mechanical driveshaft and the output is the spark plug lead. The electrical power for the spark is generated inside the magneto.

Capacitor discharge ignition (CDI) systems are electronic devices, there is no mechanical input from the engine. They are, however, not self contained because they need a source of electrical power.

A DC-CDI system needs a 12VDC power source, which typically comes the battery. The engine does not need to be turning for the ignition system to be fully functional.

An AC-CDI system gets its power from a coil in the alternator. This system does not need a battery, but the engine has to be turning for the coil to generate electrical power.

For more information on CDI systems check out .

The ignition system on the MZ202

The MZ202 uses a dual CDI (capacitor discharge) ignition system. It consists of an independent pickup coil, CDI ignition module, and coil for each system. The system is manufactured by IDM in Italy. ()

1 Pickup coils

The pickup coils are located on the flywheel and have two independent leads. Each lead has two wires, and these are wired directly to the CDI modules.

2 CDI modules

The CDI modules are manufactured by IDM part number 0560538. Physical dimensions are shown in Figure 1.

[pic]

Figure 1. Physical dimensions of the CDI module

The modules have two plugs with a total of 6 electrical connections. Figure 2 shows the plugs looking at the CDI module.

[pic]

Figure 2. Looking at the CDI module plugs.

The 2 pin plug (pins 3 and 6) is used for the pickup input, while the 4 pin plug (pins 1, 2, 4 and 5) is used for the 12V power supply and the high voltage output.

The datasheets for the CDI modules show the wiring diagram in Figure 3.

[pic]

Figure 3. Wiring diagram for the CDI module.

NOTE! The wiring harness delivered with the MZ202 is a bit different to the above wiring diagram. The actual coil used also has a different configuration and is explained below.

3 Coils

The coils are also made by IDM, part number 025 4047. The coils are not wired as shown in Figure 3, but have two galvanically isolated windings. The low voltage winding is driven by the CDI module, while the high voltage coil drives the two spark plugs which are electrically connected in series via the cylinder heads and engine block.

[pic]

Figure 4. Coil and sparkplug wiring.

4 Output voltage and current waveforms of the CDI modules

Voltage and current waveforms of the CDI output are shown in Photos 1 and 2. The peak output voltage of the CDI module is about 300V with a rise time of about 500ns. The peak output current is 22A, while the RMS current at 6000rpm is about 1.2A.

[pic][pic]

Photo1: VI waveforms, 10μs/div Photo 2: VI waveforms, 500ns/div

The wiring harness on the MZ202

The MZ202 comes with a very nicely made wiring harness which makes wiring the ignition system a breeze. Simply plug in the pickup coils, CDI modules and coils, wire in the ON/OFF switches, connect to a 12V battery, and you’re up and running.

The wiring harness has an interface plug for wiring of the master switch and ignition ON/OFF switches. The master switch is required to prevent the battery from going flat when the engine is not running, and needs to be closed when the engine is running. The ignition ON/OFF switches supply power to their respective CDI module, and need to be closed for the ignition circuit to work. A “mag check” is accomplished by opening one of the two ignition switches at a time, thus removing 12V power to that module. Figure 4 shows the interface plug detail.

[pic]

Fig 4. The interface plug.

NOTE! The numbers molded on the plug do not correspond to these numbers, but are a mirror image.

Figure 5 shows how the wiring harness is made up. In the harness there are two solder connections, a ground connection joining 9 wires, and a +12V connection joining 3 wires.

Figure 5. The wiring harness as supplied with a MZ202 engine.

A discrepancy with the wiring diagram shown in Figure 3 is clear. Figure 3 shows pin 5 used as ground, while in the wiring harness pin 6 is being used as ground. A multimeter test indicates that the two pins are internally connected inside the CDI module.

Radio interference measurement

Before even starting it should be made clear that it is impossible to completely eliminate electrical interference from a running spark ignition engine. Radio interference in microlight aircraft is not a new issue, and if the topic is raised there are usually many people who can tell horror stories and attempts to cure it. It is not limited to a specific engine or ignition brand. It is, however, less of a problem in metal skinned aircraft for two reasons, enclosing the engine in a metal skin provides a certain amount of shielding, and metal sheet provides a very good antenna ground plane which an associated radiation pattern less susceptible to picking up the RF radiation caused by the engine.

Although the source of RF interference cannot be eliminated, it is, however, possible to reduce it to such low levels that it is not noticeable, and in some cases not even measurable. If the interference is not noticeable, it is generally not considered to be a problem.

The level of radio interference is subjective, and different people will have different views of what good, bad or terrible radio interference means. When trying to reduce the level of interference some form of quantification (measurement) of the level of interference is useful to determine how effective each measure is in reducing radio interference.

Measuring the level of radio interference can be done in various ways, some crude and others highly scientific. In this investigation two methods were used, the one giving an easily and accurately measurable result, and the other more subjective. The two methods are described below:

1 Squelch level method.

The ICOM A6 handheld transceiver used has an adjustable squelch level that can be adjusted from 0 (most sensitive reception) to 24 (least sensitive reception).

For the squelch method an unused radio frequency is selected. With the engine not running the squelch level can be turned all the way down without a signal being received, typically on the A6 the squelch level could be turned down to a value of 1.

With the engine running, the squelch level needs to be turned up to prevent the radio from picking up engine interference. The level at which the squelch needs to be set at so that the radio no longer picks up the interference gives an accurate measure of the level of radio interference (a value of 0 results in continuous reception).

The squelch level depends on various parameters, and these always need to be the same for all measurements. Some parameters are:

• Antenna gain: The squelch level is strongly dependant on antenna gain, the worse the antenna the lower the squelch level can be set. For the tests no attempt was made to decrease the antenna gain, the antenna used was designed for good performance throughout the airband.

• Engine speed: All measurements were taken at idle. It was found that the squelch level needed to be turned up by 1 or 2 at full throttle.

2 Ignition kill method.

This method is more subjective and consisted of selecting the ATIS frequency of a nearby airport (about 30 miles away). With the engine running the transmission is monitored and one of the two ignition systems switched off. The change in radio interference gives a good indication of how much radio interference is generated by that particular ignition system. If a difference cannot be detected anymore when one ignition is switched off the level of interference from that ignition system can be considered to be very low. Finally if the level of radio noise does not change when the engine is switched off, the interference can no longer be considered a problem.

To give an indication of how bad the radio interference was it should be noted that with the engine not running the squelch could be set at 1 without the radio picking up a signal. With the engine running and the squelch set at the maximum value of 24 the radio would still occasionally pick up interference from the engine.

Measures taken to reduce interference

A systematic approach was taken to try to reduce the radio interference. The ultimate aim was to be able to reduce the squelch level with the engine running to 5 or less.

After each step the level of interference was measured using the squelch level method.

The steps taken are described below:

Step 1: Shielding of coils and HV spark plug cables.

The coils were shielded using adhesive copper tape from RST Engineering (). Photo 3 shows the shielded coils.

[pic]

Photo 3. Coils shielded using adhesive copper tape.

The HV leads were also screened using ignition shielding braid from Aircraft Spruce (). The shielded leads are shown in Photo 4.

[pic]

Photo 4. Shielded HV coil leads.

Result of step 1: Squelch level dropped from 24 to 21.

Step 2: Shielding the wires between the wiring harness and the ignition ON/OFF switches.

Photo 5 shows the braid used (the same as that used for the spark plug leads). Only the wires between the interface plug and the ignition switches were shielded, the harness was left alone.

[pic]

Photo 5. Ignition switch wire shielding.

Result of step 2: Squelch level remained unchanged at 21.

Step 3: Shielding the wires between the CDI modules and the coils.

Again the same braid as used for the spark plugs was used. The wires were shielded from the coil to where they entered the big join in the wiring harness, shielding all the way to the CDI module can only be done if the wiring harness is cut apart.

Result of step 3: Squelch level dropped from 21 to 19. Grounding the shield at one or both ends did not make a difference.

Step 4: Shielding the CDI modules with copper tape.

The same copper tape used to shield the coils was used to shield the CDI modules.

Result of step 4: Squelch level dropped from 19 to 17.

Step 5: Removing the original wiring harness and rewiring completely with short wire runs and twisted wires where possible.

By dismantling the original harness completely, it was possible to tidy up the wiring substantially. At the same time the ground pins on the CDI modules were changed as follows:

In the original wiring harness pin 6 of the CDI module is used as the only ground connection. The large (22A) current pulse to the coil flows through this pin and also through the soldered ground connection in the harness.

When rewiring pin 5 was used to exclusively conduct the coil current, while pin 6 was used as the ground for the 12V supply. This simplifies the wiring between the CDI module and the coil, as two dedicated pins of the CDI module are now used to drive the coil, and the full length of the wires between the CDI modules and the coils can be shielded as shown in Photo 6.

[pic]

Photo 6. Shielded wires from CDI module to coil.

Figure 5 shows how the system was wired.

[pic]

Figure 5. Final wiring configuration.

Result of step 5: Squelch level dropped from 17 to 15.

Step 6: Placing CDI modules in an aluminium box.

An aluminium box was made and the two CDI mounted inside. It was hoped that this would improve the shielding provided by the copper tape in step 4. Photo 7 shows the box with the CDI modules inside, some of the copper shielding from step 4 is visible.

[pic]

Photo 7. Aluminium box with CDI modules inside.

Result of step 6: Squelch level remained unchanged at 15.

Step 7: Changing the wiring between the CDI modules and the coils from shielded wires to co-axial cable.

The shielded wires between the CDI modules and coils were replaced using shielded MIL-C-27500 wire from Aircraft Spruce (). The outer shield was used as the ground wire (pin 5 of the CDI module), while the inside conductor was used as the high voltage wire (pin 4 of the CDI module). Photo 8 shows the two shielded cables running up to the coils.

[pic]

Photo 8. CDI module to coil wiring.

Result of step 7: Squelch level dropped from 15 to 12.

During the first 6 steps there was always a noticeable difference when one of the two ignition systems was switched off. The difference between the two ignition system was, however, seldom the same (one ignition system would always generate more noise than the other).

After step 7 there was no longer a noticeable difference between the two ignition systems. With one ignition system switched off the squelch level with the engine running could be decreased to 8, but there was no noticeable increase in interference when the ignition was switched back on again.

When switching off the engine, however, there was still a noticeable change in radio noise. Radio performance is not “perfect” yet, but can now be classified as somewhere between “good” and “very good”.

There was no single measure that eliminated or even caused a huge reduction in radio interference. Most steps resulted in small but measurable reductions in interference, and the end result was the cumulative effect of all the small improvements.

The Ducati CDI system used by Rotax

Newer Rotax engines also use CDI ignition systems, their systems manufactured by Ducati. Figure 6 shows a drawing of a typical 2 stroke, 2 cylinder, dual ignition Rotax system. The four stroke Rotax (912 family) uses a different system which will be briefly touched on a bit later.

[pic]

Figure 8. The Ducati CDI ignition system used on 2 stroke Rotax engines.

There are two main differences between the IDM system and the Ducati system as described below.

1 Power supply

The Ducati system derives its power directly from the alternator, while the IDM system derives its power from the 12V battery. The Ducati system can operate without a 12V battery, but not the IDM system.

On a side note, this is the reason why a MZ202 starts quicker than a Rotax 503, I think most MZ202 owners are impressed at how quickly their engine starts.

With the IDM system a lot more care needs to be taken when designing and wiring up the 12V system, as a loss of 12V power to the ignition modules will cause the engine to stop.

A loss of 12V power could be caused by any one of the following scenarios:

1 Broken wire

A broken wire anywhere in the ignition 12V supply or ground return loop will cause the engine to stop. The broken wire could be at the battery, master switch, kill switches, ignition modules, or anywhere in-between. This loop is the most critical 12V supply loop, and special attention should be paid to it.

2 Short to ground

A short to ground in the ignition circuit will cause a voltage drop dependant on the severity of the short circuit. If the short is severe the engine will stop.

If the ignition circuit is fused and the fuse blows, the engine will stop and it will be impossible to restart the engine before the fuse is reset/replaced. The ignition circuit is not fused it must be wired with extreme care to prevent any chance of the 12V wires shorting to ground.

All circuits not related to the ignition circuit must, however, be fused. This is to prevent a short in a non critical circuit from pulling down the 12V supply and stopping the engine.

3 Master switch accidentally opened

The Master switch should be mounted in such a manner that it is virtually impossible to accidentally switch it off in flight.

4 Master switch failing open circuit

If the master switch fails mechanically (due to vibration) or electrically (due to overloading). If the standard MZ202 wiring harness is used the switch must be rated to carry the maximum charging current of about 12A.

5 Battery going flat

If the charging system stops working, it is possible that the battery could be charged enough to start the engine, but then loose charge and cause the engine to stop in flight. A low voltage warning system would detect this fault and give enough time to make a safe landing.

The charging system on the MZ202 engine can deliver about 13A. The ignition system requires about 1A. The total constant electrical load of all other loads on the aircraft should thus never exceed about 10A (around 120W).

2 Kill switches

With the Ducati system the ignition is killed by shorting a lead to ground, while for the IDM system the ignition is killed by opening the 12V supply switch.

3 Integrated coil

The Ducati system combines the CDI module and coil into one unit. There are thus no external wires that run between the CDI module and the coil. It has been shown that this link is a dominant source of RF interference, and the Ducati system in this respect can be expected to generate less RFI. It is, of course, possible to mount the IDM CDI modules and the coils next to each other and have very short leads between them. This is in effect is what Ducati have done.

4 Rotax 912 system

The Rotax 912 ignition system is similar to the IDM system, with separate CDI modules and coils. The modules are, however, mounted very close to the coils with the wires between them not more than about 7cm (~3 inches) long.

My opinion…

These are my thoughts on the issue, and I’m open to discussion.

1 Source of RF interference

It should be realized that in this whole investigation no changes were made to the engine. Also no changes were made to the CDI modules or coils. The physical location of the coils and CDI modules was not changed. All that was changed was the wiring between the various components, and the addition of external shielding in the form of either copper braid, copper tape or aluminium sheet.

One can thus not say that a specific engine, or specific ignition system generates more or less radio interference than another. It is how the different components are connected to each other that affects the level of interference.

The wires between the CDI modules and the coils seem to be the main culprits when it comes to RF interference. The measured voltage and current waveforms confirm this, with a very high rate of rise in the voltage, and a large associated current peak. Using a shielded wiring structure with the shield used as the ground conductor greatly reduces both electric and magnetic field strengths around these wires.

2 My recommendations

The MZ202 engine runs beautifully with the supplied ignition system and wiring harness. If I could fly without a radio I would use the wiring harness.

If I was to build another flying machine using an MZ202 and knew that a radio was mandatory to fly (like my Mosquito helicopter that’s on its way) I will do the following:

1. Throw away the wiring harness supplied with the MZ202 engine.

2. Make a aluminium box for the CDI modules, and mount them inside as close as practically possible to the coils.

3. Wire the coils to the CDI modules using co-axial cable.

4. Complete the wiring as per Fig 5.

5. Do not fuse the ignition system 12V supply.

6. Fuse all other electrical loads.

Icom A6/A24 headset transmit issue

The ICOM A6/A24 handheld transceiver works really well as a handheld, both transmitting and receiving.

It is possible to plug in a headset using the ICOM OPC-499 headset adaptor, or alternatively 3.5mm (for speaker) and 2.5mm (for microphone and PTT) plugs. When doing this there are no known problems when receiving, but there is a known feedback issue when transmitting. This feedback issue is not well documented; here are a few links that refer to it.







I did some experimentation using a headset on an ICOM A6 and found the following:

Using the supplied rubber antenna (FA-B02AR) there was always feedback when transmitting. It was generally so bad that transmission was not possible.

Using an external antenna, the feedback was dependant on the SWR of the antenna. With a good antenna (SWR < 1.5) there is no feedback. As the SWR gets worse, feedback goes from barely audible to so bad that the radio cannot be used for transmitting.

-----------------------

(-)

(-)

(+)

(+)

6

1

5

3

2

CDI2

CDI1

Master

To voltage regulator ground

To engine ground

To battery negative terminal

To battery positive terminal

G G C +B

Alternator

4

2

6

3

4

2

6

3

Interface plug

Voltage

regulator

Coil 2

CDI2

PU2

+12V solder

connection

Coil 1

Ground solder

connection

CDI1

PU1

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download