Engine ECU Oscilloscope Traces

Here are some ocilloscope traces taken from wires at the Engine ECU.  I used a Tektronics TDS 360 borrowed from my former employer.  Only channel 1 was active and I selected 20 MHz input bandwidth.  Ground was taken from a screw in the chassis near the ECU.  I clipped the probe to a very small screwdriver which I was able to poke into the back of the connectors until it made contact with the pin assembly.  I didn't worry about triggering since by pushing a button the ocilloscope stored a bitmap of the screen.  These bitmaps, rotated and cropped, are linked to this page.  Although the images are small in file size, they are large on the screen so I have not placed them directly in this web page and you will need to open them from the hyperlinks.

Injector Signals

The Engine ECU sends a signal to each of the four injectors.  I took two traces from injector 1, the signals to the other three being similar.  The first trace was taken as soon as possible after starting the car, I believe during the "pre-heat" phase when the oxygen sensors and catalytic converter are heated and the engine does not power the car.  The second trace was taken a little while after the first when I noticed that the pulses had become narrower.  Presumably, the "pre-heat" phase had ended and the engine was idling normally.  In both traces, you can see that the pulse period is about 92.5 ms.  This indicates that the engine was running at about 1300 r.p.m. (there is one pulse at each injector per two crankshaft revolutions).  The signal inactive level is around 14 volts, the voltage of the auxiliary battery when the car is on.  It appears to pulse to a low voltage to open the injectors.  The pulse width is about 6 ms in the first trace and 2 or 3 ms in the second trace.

There seems to be the potential for measuring fuel flow here.  The four injector signals could be combined with an old-fashioned diode-transistor logic NAND gate.  Using discrete diodes would allow them to be located near the Engine ECU so that only one wire would be needed running away from it.  The positive pulses that result could be simply passed through a low-pass filter to extract a signal representing the proportion of time for which one of the injectors is open.  Although fuel flow in grammes per second would not be known, the signal should be approximately proportional to fuel flow and would serve as an excellent indicator.

Igniter Signals

The Engine ECU sends a signal to each of the four igniters and a fifth signal to all four together.  I took a trace from igniter 1, the signals to the other three being similar.  You can see that the pulse period is about 91 ms.  This indicates that the engine was running at about 1320 r.p.m. (there is one pulse at each injector per two crankshaft revolutions).  The signal inactive level is close to zero volts and positive going pulses to about four volts appear to be the command to fire the igniter.  The pulse width is about 2.5 ms.  I also took a trace from the signal to all igniters.  Here, the inactive level is a little over 5 volts, perhaps the supply voltage of some logic circuits.  Negative going pulses to about 0.8 volts appear to be the firing commands.  These pulses are much narrower than the individual igniter signals at around 1 ms.  Their period is a little more than 25 ms, indicating that the engine was then running at just over 1200 r.p.m.  I am led to believe that people have used this all-igniter signal as the input to a tachometer.


Last edited November 14, 2002.  All material Copyright   2002 Graham Davies.  No liability accepted.