mileage computer for a carbureted engine

[mozul]

I have two flow meters only because the eBay auction was for two. I will only install one of them. There is no fuel return line. It just flows into the carb, gets burned up by the engine and pushes me down the road.

Perhaps I can make a second MPGuino with my other flow sensor for my motorcycle.

[AndrzejM]

For one flow meter the setup is much simpler. You count time between pulses for instant FE reading, and count pulses to have total amount of used fuel. Of course you have to combine that with a reading from VSS sensor.

[drpepper]

Hi
I have a 1975 GMC Moterhome. I am in desperate need of something to get MPG. O ya, I am poor. lol I found flow meter as log as $140 but nothing to read them unless I want to put in a fish finder. Did anyone put a project together?

[diesel_john]

one note: most low cost flow meters are not sensitive enough to accurately measure the small amount of fuel flow that an automotive engine uses at low power settings. A VW diesel for example uses 1 gallon per hour @ 50MPH. 128ounces/ 60 min.= 2.13 ounces per minute. Most flow meters can bypass that much without even measuring it. IMHO it would require very tight tolerances in positive displacement flow meter to that measure small flow. They exist it is just they are more expensive.

[mozul]

I have ordered the rest of the parts for my MPGuino. Once they come, Ill be able to make it read my flow meter. I can run a gallon of water through it and count the pulses. My flow meter is designed for fuel. It is for a boat to measure the gallons per hour that they use. Ill see how reliable it is. For a different vehicle example, my ultra gauge measures about 0.5 gallons per hour at idle for my Jeep Liberty.

[P-hack]

unsolicited thought:
by counting pulses, the sensor is inaccurate at slow speeds. Perhaps even a little off over the whole range.

If the timing changes of the pulses are considered, perhaps one could calibrate it and get more useable range and more accuracy from the sensor by using the pulse timing to look up calibrated values.

Or the sensor just stops turning at that slow of a speed.

After that the next gain in accuracy would be had trying to consider the acceleration of the rotor.

[TheTestPilot]

Greetings Gentlemen
I have been thinking about a suitable solution for flow sensors on & off for some time .
I have tested various types of flow sensors & as Diesel John mentions a few posts back ,low flow reliability comes at a cost ,I have been trying to locate a used flow scan sensor (model 264)that was used on a GM cars here in the 80's (apparently it was accurate) to see how it was made.

Here is another thought ,as the mpguino measure injector pulses(pwm) ,has any body thought about simulating the injector pulse train ?
This could be referenced from the ignition pulse ,perhaps along with a gm map sensor to vary the pulse length.
It wouldn't actually measure fuel flow but may be close enough (at cruise only) for real time comparison mpg's.
I have been testing this circuit for a few tanks now ,it is only referenced to the ignition pulses at the moment ,I don't have enough time to take this further .It may be worth further thought/development.

Tp

[P-hack]

I was looking at ultrasonic flow meters, and it seems likely that the sensor itself can be made with just a couple of piezos, that measure time of flight in one direction and then the other and gauge flow rate that way, it may need it's own cpu (at least for prototyping) to send and process the signals, but it might be the path forward for good range, accuracy, and reliability, and very low cost, ease of manufacture, and compatibility with lots of fluids. Don't know if it has been considered yet. My bike is the only non-efi left so I'm not real gung ho about it, just thinking about it.



Ultrasonic flow meter - Wikipedia, the free encyclopedia

[P-hack]

though I think if you arrange the piezos like this, it wont take so many maths to figure out the flow, and enhance sensitivity and range.


Or not, would take more transmit power perhaps. Having the sound in parallel with the flow doesn't take any trig and provides maximum difference between signal timings for each direction, and having a longer sonic path should increase the travel time of the signal through the fluid and thus the resolution of the signal.

[P-hack]

yah, 1/4" sensor within 1% accuracy from 2ml/minute to 3liters/min, made out of piezos and pipe (and processing power), definitely worth looking into.

http://content2.smcetech.com/pdf/LFU.pdf

that equates to 0.03 gph to 48 gph @ %1 accuracy?!?