[consaka]

Originally Posted by dcb

For typical diesel you need two, one to measure fuel going into the pump and one to measure it going out.

I already explained how to do this with one flow meter and to bypass the return line. But yes ideally two would be nice with a logic board to do the math for the output to the mpguino

I thought briefly that you could use one flowmeter, but the pumps (bosch at least) have internal bypasses that make it difficult, as John points out.

It can bypass internally all it wants since internally isnt a resevior with air in it it will make no difference on fuel demand from the tank. Any bypass or leakage internally is either going to stay in the pump or more likely go back to the tank on the return line. The only thing making it difficult is if they have more then one return line going back to the tank.

Rather than timing the injection pulse lengths, you just count the number of pulses from the flowmeter, as each pulse represents a certain volume of fuel (more or less). If you have one on input and one on output then you subtract the output count from the input. With the flow rate of the output it might have reasonable resolution. You might still want to time the length between pulses for more predictive instantaneous readings, but for the longer term trips you really just care aboout the total pulse count.

What you just described is what I have explained a few times already. As for the instantaneous reading the zemco was very precise in that regard. You are either using fuel or you aren't. Personally I dont think those old computers that computed the data from the zemco's was all that great, compared to the mpguino, but even on a carbed engine the response and accuracy was right there.

Though the flowmeter approach is probably better suited for the Carb thread since that is a single sensor solution for carbs, just don't have a solid on flow sensor sources.

You are right about a good source for flow sensors but I think if someone with a few pesos wanted to have some made in china cheap there would be a good little market for it similar to the mpguino market.
I personally think a zemco would be more accurate then trying to measure the pulsewidth. As your injectors get old and get deposits built up it can change their flow characteristics. If that happens then your accuracy went out the window. As we are all guessing our settings to some extent our calibration is a bit suspect as it is. With an accurate flowmeter or two none of that kind of thing would affect it at all. Once calibrated it shouldn't need calibrated again. The trick is to come up with a really accurate high resolution design that can be outsourced cheap enough.