Gordon,

The attached image shows signals generated by an EC2 for a 13B engine at 5000 RPM. The signals are identified by color in the legend at the right of the image.

Like Ed, I have built fuel totalizing and flow rate instruments based on discrete counter IC's and based on a microprocessor. The unit based on discrete counters was accurate only under the conditions at which the calibration was made. The reason for this is the effect of the dead time of the injector preventing a simple correlation of the length of an injector pulse with the amount of fuel delivered during that pulse when the length of the pulse is variable. Since the aviation application often involved operation at the conditions used for calibration, this limitation was not often very significant.

The system using the microprocessor enabled the accurate calculation of the amount of fuel delivered by each pulse at any injector pulse length by taking into account the injector dead time. The injector dead time was determined for the injectors used and its value was part of the programming. The microprocessor version uses inputs from the primary and secondary injectors and displays the RPM, fuel flow rate, and the amount of fuel remaining in each tank since I have a duplex fuel valve with a switch telling the microprocessor which tank is in use. I chose to use the Texas Instruments Launchpad development system which cost about $5 and included two microprocessors.  Of course, other components were required for the final unit, the most expensive part being the display.