AutoStart Engine Controls

Shopping For An Engine Controller?

back

Here's Some Points To Ponder
By: Kevin Chisholm and Daniel Watkins

In principle, the meaning of the term engine controller ranges from as simple a device as a speed-switch, that engages, and then releases the starter motor (once the engine "starts"); to something as complex as an engine management system, that regulates engine speed and combustion conditions.

This article deals with the mid-range: controllers whose functions are starting, fault monitoring and stopping of an engine, sometimes referred to as ‘Auto-Starts’.

A typical instance in this category might fit in the palm of your hand, and have, at the very least, the following input terminals:

• Engine-speed sensing
• Start (Run) contact
• Oil pressure switch
• Temperature switch

And two output terminals:

• Starter
• Fuel/Ignition

Also, it will feature various Status Indication lights, as well as means for adjusting:

• Starter Disengagement Speed
• Over-Speed Trip
• Maximum Cranking Time
• Max. no. of crank attempts.

Briefly:
Upon closing of the Start contact, the controller goes from a Stopped state to Cranking, and then to Running. As soon as the Start contact opens, it goes back to Stopped state. Additionally, activation of the Oil Pressure or Temperature contacts, and other conditions, can cause a Failure state to be asserted.

Need for Speed Feedback

Although the type of engine controller we are focusing on does not control engine speed, it does need to keep track of the engine’s speed to decide when to disengage the starter motor, for over-speed protection, and other reasons.

One of the first things to establish, when putting together a specification, is the type and frequency range of speed signal that the controller must be able to accept. Voltage, frequency and other signal parameters depend on the sensing method used, as well as on an engine’s speed range.

Some of the most common speed-sensing methods are:

1. Magnetic Pickup
2. Flywheel Alternator
3. Engine Alternator
4. Generator Output

Let’s Look At Them One By One:

1. Magnetic Pickup
   A Magnetic Pickup is a small magnetic sensor that’s mounted near the flywheel ring-gear’s teeth, to detect their passing. Their output frequency, measured in Hertz (cycles per second) is a function of engine speed, namely equal to the RPM (Revolutions Per Minute) times the number of teeth on the ring gear, divided by 60.
   
   Most controllers are compatible with magnetic pickups, and some exclusively so, simply because their output signals are so very ‘clean’. However, their output voltage is proportional to the frequency, making very low speeds harder to detect. (In some situations, one may want to distinguish between low speed and NO speed.)
   
   A clear disadvantage of magnetic pickups is difficult installation: drilling, tapping, mounting, then adjusting position for good signal.
   
   
2. Flywheel Alternator
   A Flywheel Alternator is a low-cost and convenient means of getting a speed signal from a small engine. Typically, on a single-cylinder engine, there is a magnet embedded in the flywheel, which induces one pulse per revolution onto a stationary coil. At a 600 RPM (idling) speed, for instance, the pulse frequency will be 10 Hz, and at 3600 RPM, 60 Hz.
   
   However, this signal is very ‘spiky’ (high harmonic content), with peak voltage values in the 100 to 200 volts. Although appropriate filtering could ‘clean-up’ such a wave-form and use it, few controllers boast the ability.
   
   
3. Engine Alternator
   An Engine Alternator is an AC generator, driven by the engine, often via a belt, whose output is subsequently converted to DC for battery charging. Now, most engine alternators also feature a tachometer output, which provides a signal of frequency proportional to engine speed, suitable to most engine controllers.
   
   The frequency output of engine alternators varies significantly depending on engine speed, pulley ratio, number of alternator poles and normal belt slip; and their output voltage stays within battery charging range. One major concern with this sensing method, however, is the belt drive: The belt could break, or slip excessively. Should it break, a well designed engine controller will detect the sudden loss of speed signal, and perform a safe engine shut down, and even indicate loss of speed signal as the cause of it. However, if the belt merely slips, a speed signal will still be there but it will indicate that the engine speed is lower than it actually is. As a consequence, an over-speed situation could go unnoticed, leading to engine damage or destruction.
   
   
4. Generator Output
   This applies to engine-driven generators. When at operating speed, a generator’s output frequency is adjusted to be 50 or 60 Hz, at 115 or 230 volts. Very few controllers are designed to withstand such voltages across the speed sensing input, so usually a step-down transformer is used.
   
   The main problem with this option is a very low starting voltage. Most AC generators start with nominally "no field" and "no output", as the generator’s output is, itself, used to power the field winding. In truth, there is an output voltage, however feeble, due to residual magnetization, which is what, after a few seconds, resolves this chicken-and-egg paradox.
   
   Typically, this type of generator produces anywhere from 0.1V to 1V during starting. After the transformer, that may become 0.025V to 0.25V; at cranking frequencies of about 10 to 20 Hz. Some engine controllers can readily use such a signal.
   
   

Speed Related Issues

Speed signal voltage:
Some engine controllers’ rated maximum is 50 V: they cannot be used with a flywheel alternator. Some can only sense signals down to about one volt: these can’t use transformer coupling from generator output.

Wave-form:
A controller may withstand flywheel alternator voltage levels, but may be unable to work with its wave-form.

Frequency Range:

Again, mere range numbers are not enough: A controller may be able to detect a 20 Hz signal at 3V, say, but not at 0.3V; and yet the same controller may be able to sense a 2 kHz signal at 0.3V. In assessing detectable-range adequacy of products under consideration, therefore, voltage sensitivity as a function of frequency, and how it graphically compares to the output characteristics of the sensor used, should ‘give one a better picture’.

One may ask, also, what happens above the maximum frequency? Does the sensor saturate (continue to see the maximum frequency it can)? Or does it suddenly see zero speed? It is very dangerous to set the Over-Speed (see below) at a value near a ‘drop-to-zero’ maximum; and a well designed controller won’t allow it.

Ease of adjustment:
Two parameters a user needs to set, in connection with engine speed, are:

• Starter Motor Disengagement Speed, called ‘Crank Disconnect’ in the trade. This is the engine speed up to which the starter motor remains engaged.
• Over-Speed Trip Point, or ‘Over-Speed’, for short. Usually set at about 15% above normal running speed, is used for fail-safe engine shut-down, should a mechanical or other failure cause the engine to speed-up out of control.

The usual means provided is miniature potentiometers, single-turn type, to 20-turns precision.

However, with adjustable frequencies spanning between 10 and 9,000 Hz in a single range, precisely setting a low value on a linear potentiometer, even a 20-turn type, is a difficult task. Selection of a suitable sub-range via DIP-switches is provided in some controllers.

Some products feature a ‘Setup’ mode, which temporarily shifts frequency parameters by about 15% in order to help adjust the Over-Speed setting without actually over-speeding the engine.

Temperature drift:
e.g. By how much will the set speed parameters’ real values drift as temperature dips to, say, forty degrees below zero?

Electromagnetic Interference:
EMI is another consideration: The wire connecting the speed sensing input to the sensor output can, and will, act as an antenna, and catch ignition and relay spark noise. Proper filtering can remove much, but not all, of this noise. Thus, other preventive solutions are often recommended, some more effective than others…

One of the least effective, yet very popular, is the use of ‘shielded’ wire. It is a common misconception that the tube-mesh around a "shielded" wire can somehow block EMI. Such incorrect notions are borne from lack of understanding of how coaxial cable works: The "shield" is not supposed to be just "grounded"; it is supposed to conduct the (differential) signal’s reference or return path (the ‘other’ wire), such that differential currents magnetically induced between signal and return, on one side of the "shield", effectively cancel out those induced on the other side. Coaxial cable thus ‘rejects’ magnetically-induced, differential noise currents by virtue of its axial symmetry. Properly applied, coaxial would be excessively more than adequate for a simple speed sensing application; and yet, the way it’s often used ("shield" grounded), it is futile.

A good controller’s speed sensing circuit should provide two input terminals for differential sensing via twisted-pair wire. This makes for a superior installation: simpler, less costly, and very effective: EMI signals picked up by the two wires simply cancel each other out.

Quality of Operation

There’s no better way to assess design quality of an engine controller than to get a sample and test it, both: on the bench and on typical engines. Functional testing could include the following:

• Normal Cycle:
  Testing the unit for consistent, correct behavior through every operational state, with an engine that responds as would be expected.
  
  
• Exception Handling:
  Testing how the unit manages when things go ‘rather unexpectedly’. Some ‘surprises’ one could try on a controller, to ‘see how it reacts’, are:
  
  
  • No speed signal (wire out) during cranking.
    (Controller should ‘think’: "Starter solenoid is not closing, or perhaps speed signal wire fell off, or starter pinion is not engaging ring-gear properly. In any case, I shut down ‘No Speed’." Otherwise, starter damage could result, or engine might start without over-speed protection.)
  • Engine starts, then stalls.
    (Does the controller consider this an engine stall, and fails; or a failed crank attempt, and tries again? If the latter, does it count it towards total crank attempts? Hint: Shall we not try at least once more…?)
  • Engine starts, attains full running speed, then slows down to a slow idle and stays there.
    (Controller should ‘think’: "Too heavy a load? Governor is malfunctioning? If situation doesn’t improve quickly, I’m shutting down.")
  • Etceteras…
    (See what tricky situations you can conceive.)
  • Extreme Conditions:
    Finds-out what the unit does when subjected to abnormalities ‘below’ monitored level, while operating in normal cycle, and during every phase of operation. –e.g.: Loosen the ground wire and try some ‘Morse Code’ while the engine is running, or during starting. Try also with the battery-positive wire.

For crude, but effective EMI testing, try moving the controller next to a ‘live’ spark-plug or ignition coil, and/or wrap some of its connecting wires around spark-plug wires, until there are obvious behavioral effects: such as engine stutters and hick-ups, and/or randomly flickering lights: Does the controller recover consistently? Or does it sometimes energize the starter while the engine is going at full-speed, or shut down with a false failure indication –e.g.: Over-Speed?

A capacity for recovery should weight more than mere resilience, because no matter what… the right jolt, or bolt, will eventually come.

Engine (and Load) Safety

Typical dangerous or fatal engine conditions monitored are:

• Low Oil Pressure
• High Coolant Temperature
• Engine Over-Speed

Most engine controllers perform these functions, but all are not equal when it comes to details:

Oil Switch Verification:
It should be a trivial task for an engine controller to verify that the oil pressure switch is signaling NO Oil Pressure when an engine is off (and we are planning to start it). If this is NOT the case (i.e. if the Oil Pressure Switch says "Oil pressure is okay"), then either the pressure switch is damaged or the connecting wire fell off, and in this case, we do not want to start the engine until the problem is rectified.

Most controllers fail to do this simple verification. This is a very serious oversight, because if something is wrong, you would not know it until oil runs out and, being unable to detect the condition, the engine is destroyed.

On the other hand, note that the engine cannot start unless there’s a temporary bypassing of the Low Oil Pressure signal. An oil bypass delay is normally offered as an adjustable time delay after engine start, during which the low oil pressure signal is ignored. A more sophisticated approach actually ‘counts revolutions’, rather than seconds, which better tracks with oil pressure rise. This approach can eliminate the need for field adjustment of oil bypass period.

Additional Failure Input:
Take the case of an engine driving a pump: it may be desirable to stop the engine on loss of prime or of discharge pressure. This could be done by ‘paralleling’ a sensor to, say, the Low Oil Pressure switch; but, should such a failure occur, the indicator lights’ status might suggest oil failure. An Additional Failure Input simplifies diagnosis.

Other safety-related issues:

• Is there a restart delay? Without it, the starter may be engaged while the engine is coasting, resulting in starter pinion and ring gear damage. However, a fixed restart delay, if too long, may impose an unnecessarily long wait; and if too short, may cause damage. Preferably, engine speed is monitored at all times, and starter engagement preempted by any non-zero speed: zero speed restart.
• If there’s damage to the speed sensing circuit, does the controller detect this loss of speed signal, and shut down the engine? Is a failure from loss of speed signal indicated distinctly from engine stall (gradual coast to a stop)?
• In a typical engine stall situation, oil pressure falls as speed decreases. Is low oil pressure sometimes falsely substituted for engine stall as the indicated cause for the shut-down?

Extra Features

Over-temperature Grace period:
Permits engine to start and run for a time during a High Temperature condition, to allow circulation of freshly added coolant.

Pre-Heat ‘Timer’:
If a diesel engine is being considered, is there a built-in pre-heat timer function? If so, is preheat applied in the unlikely event of starting during an over-heating condition? (See Over-temperature Grace period, above.)

Choke ‘Timer’:
If a gas engine is being considered, is there a built-in output for a carburetor choke solenoid? If so, is choke applied when the engine had been running and was stopped only for a moment, then restarted?; or on the first crank attempt? (NOTE: Choking a warm engine can lead to flooding, a condition that would be very hard to reverse during subsequent cranking attempts. Without a thermostat, choking on the first attempt, therefore, is a gamble.)

Microprocessors and EMI

Some highly reputable engine controller makers have been known to boast of having no microprocessors in their products; as they seem to feel that low scale integration technologies are less susceptible to EMI.

Consider this: if microprocessors were really all that bad, surely not so many of them would find uses in mission-critical, military and aerospace applications.

Electromagnetic Compatibility is rather a matter of design quality.

Output Technology

Some engine controllers include miniature relays for driving outputs; others use solid state devices.

Relays are subject to wear by normal operation, and their contacts may suffer additional wear from sparking during release and contact-bounce.

Solid State outputs are superior in that there is no limit to their number of operations, and have no bounce and no sparking characteristics.

There is a popular perception that, at equal ratings, a relay is more ‘rugged’ than a solid state device. In addition, solid state driver circuits are, out of necessity, often designed with built-in inductive kick-back protection diodes; which not only help protect the devices, but also reduce generated EMI.

In Summary:

The process of selecting an engine controller requires, not just ‘thoroughness’, as in comparing and verifying specs, but also attention to detail and imagination: It necessitates the asking (and testing) of some of the "What if…?" questions that were, or perhaps should have been, asked during a product’s design.

From PowerLine Magazine (ESGA)

back

Standard Terms & Conditions of Sale

Home   |    About Us   |    Products & Support   |    Buy   |    Contact Us