31st January 2004

The speed sensor input issue had been resolved by feeding the signal from the tachometer output back to the speed sensor input through a buffering amplifier with a gain of 0.5:1. The tach output provides three pulses per engine revolution, switching from 0 volts to 12 - 14 volts. The speed sensor input is expecting four pulses per revolution of the speed sensor with a level switching from 0 volts to 5 volts. Al Wick had told me that his experience had been to exceed the five volts and go below the 0 volts for the input to work effectively. I used a single supply, rail to rail, operational amplifier (OP284) with a gain of a half to buffer the signal and to act as a level shifter. Since I installed this device I have not had the error code 33 reappear. I conclude that this method for spoofing the speed sensor input is a good one.

The engine was started and run for just under two minutes. During this period the engine temperature would rise from room temperature to approximately 100 degrees Celsius at which point the engine was shut down. Data from the EIS and the oscilloscope were recorded.

Engine & Propeller RPM

The following chart indicates the rpm as a function of time. The RPM was intentionally brought up slowly to ensure reliable data at the EIS.

Manifold Absolute Pressure (MAP)

The MAP is indicated in the chart below. The engine is started twelve seconds after the recording starts. The filters in the acquisition system slow the response time for the measurements and this can be seen during this first twelve seconds where the manifold asymptotes to the environmental pressure. As the engine starts, the manifold assumes a strong vacuum or low MAP. After the start the engine idles as indicated above followed by a slow increase in RPM and MAP. At 29 inches of Hg the engine is only turning 3000 RPM. As time progresses, the engine speed increases slowly and there is considerable backfiring. The MAP data and the RPM Data do not seem to correlate. The initial though was that electronics that had been installed to overcome the speed sensor input, was upsetting the ECU. After some discussion the fuel tank was checked. The temporary fuel tank is semi transparent so the fuel level was visible but with some difficulty, there seemed to have some two inches of fuel. On closer inspection the fuel take off pipe starts about two inches from the bottom of the temporary fuel tank. Calculations on fuel burn suggest that about one litre per minute is the expected so at least two litres (half a US gallon) is needed to conduct the test. The max fuel capacity for this temporary tank was one gallon but the tank had not been refuelled prior to the test. Clearly, as with every flying experience, it is necessary to check the fuel before each adventure.

Last Updated:    Thursday August 31, 2006