Question for Electrical / Computer Gurus - Hall Effect sensor for engine position

[GG54172]

I have been researching fuel injection systems for a while now, and there is one thing I cannot figure out. Most aftermarket fuel injection systems such as MOTEC use a hall affect sensor with a toothed sprocket attached to the crank or cam shaft to monitor engine speed and position. I understand the logic for engine speed, as the teeth spinning by the sensor create a voltage proportional to the engine speed.

 

To detect an initial engine position, the sprocket will have one or two teeth missing. How does the computer detect and determine that the missing teeth is a larger gap than all of the other gaps? Let's say you had a 60 tooth sprocket that is missing two teeth attached to a crankshaft spinning at 10,000rpm. The teeth will be spinning by the hall effect sensor with an AC voltage at 10,000Hz, and two teeth missing will create a zero voltage incident only 0.0002 seconds long. How does the computer notice this is different from the other pulses? I understand that with a 2MHz processor 200microseconds is a long time, but what does the computer logic look like?

[Squeeze]

The Nature of an AC Current is that it is waveing between Peak high, zero Voltage and Peak low.

 

The EFI-Computer senses the missing Tooth as a Peak Level with an unusual Lenght and therefore interpreting this unusual Lenght as the Point which it refers to. The Computer usually not only operates on the Peak Levels, but takes Zero as it is (Value = 0) and uses both Peaks to compute. This means, by moving one Tooth and one Gap next to the Sensor you're generating more Signals. When starting in the middle of the Gap, Level is Peak low, when the start of Thooth is in Range, the Voltage goes up through Zero Level and reaching Peak high. These are three Signals. When the Thooth leaves the Sensor, the Level will fall through Zero and reaching Peak Low. Another two Signals created.

 

What this Point means to the EFI, is up to the Software and the Definition thereof. I. E. it could define TDC on 1st Cylinder.

 

 

I added two Pics from a EFI Vmax. You can clearly see the missing Thooth. You can disregard the Enlightment of this heavily modyfied Rotor (weights only 1.6 kilogramm, stock is 3.8 kg) and the new (and never ever again) Starter Clutch Assembly.

[KiteSquid]

The computer might not be concerned with the voltage being seen, but it may count the number of pulses from the sensor in a given time period, this would indicate cam shaft RPM.

 

With the missing tooth, it knows were TDC for cylinder #1 is, then it knows in its' look up table how many pulses to TDC for the other cylinders.

 

It uses this to know when to operate the fuel injectors, how long to have it operating for, and when to fire the spark plug.

 

Over time, it can take information from the 02 sensor(s) to adjust the air to fuel mixture in an attempt to achieve the most efficient timing of the fuel and spark for the conditions at hand for that particular motor, IF the engineers put enough thought into the firmware. this would provide the operator with maximum power and maximum fuel economy for that engine.

 

Of course our engines don't have the feedback loops like a formula 1 engine does, like one O2 sensor per cylinder.........but what if they did???????????????????

[GG54172]

I have been chatting with some computer science major friends and you are right, Kitesquid. Let's see if I can paraphrase what they told me....

 

The hall effect sensor doesn't detect the tooth, rather it detects the edge of the tooth on the sprocket. The ECU has the crank position sensor wired into a special kind of interupt port. When ever a tooth passes by the sensor, the voltage spikes high or low depending on the computer program and wether they want to detect the leading or trailing edge of the tooth. For the sake of discussion, let's assume that the program detects the voltage high.

 

The special port has an interupt capability. Every time the port reads a voltage spike, the interupt tells the computer to pause what it is curently doing. Then the computer goes to the internal clock to read and record the current time in a memory location A. The second time the sensor spikes the interupt records the time again in memory location A, but the previous time is pushed to memory location B. The time recorded in location A continues to get pushed into location B as often as a tooth passes the hall effect sensor.

 

So later, in the progam, you can call up the time recorded in memory location A and B. That can be used to calculate rpm. The current calculated rpm and the previous rpm can then be compared...if the calculated rpm changes by a significant amount, like 150%, then the computer knows it is at the gap. The next tooth becomes your zero reference point.

 

Something like that anyways I think...