Hi
I am new to this forum, I live very far from you (Romania), and I am interested too in an accurate, digital fuel consumption sender/gauge. This is a mandatory first step in order to monitor any fuel saving device.

I've decided to ask some questions in here because there are a few high skilled, motivated members that I am sure can help in my project.

My car is on carburetor. I know, you're into fuel injection solutions, but hear me out, my ideas can apply to all cars.

Basically, I solved the fuel sender poor accuracy by replacing the wire resistor type with a capacitive one. I am in the process of designing a simple PIC-based digital fuel gauge that takes into account the irregular shape of fuel tank.

Now comes the tricky part. There is a feed line coming from fuel pump, and a return line going to fuel tank. I should measure feed flow, subtract return flow, then divide by distance. When idling division is made with respect to time (GPH). My main problem is that I need a differential fuel flow meter for small flows and smart enough to detect if there are bubbles of air with the fuel.

I attached 2 images that describe the installation.
The main advantage is that you neither count injector pulses, nor listen injector clicks. Both are highly inaccurate because you never know for sure how much fuel squirts each injector. All we want is (fuel feed - fuel return)/distance.

Any thoughts?

If your car has a carburetor, there shouldn't be a fuel line returning to the tank. You should just have a mechanical fuel pump going directly to the carburetor.

I know my carburetor and many other European carburetors by heart.
Trust me, there is a return line to send back to tank any fuel in excess.
We don't want fuel to overflow past needle float valve, don't we?

I guess. I've just never seen a configuration like that. Only ones I've seen with fuel return lines were fuel injected. There is a line from the carb to the tank in my carbed vehicles, but that is intended for vapor only, and if you actually get fuel in it, you have problems. They are all 80's vintage GM vehicles though, late model European vehicles with carbs may be different.

There used to be an add on mileage computer available in the 80's for carbed vehicles. It included an optical sensor that was installed inline on the fuel line, and was able to measure the fuel rate. They were somewhat problematic, as when the light bulb burned out, the sensor stopped working. I've read about folks who have switched the bulbs out for LED's though.

Please have a look in here: https://skontrol.ru/node373/ and tell me if it is the most reliable, accurate, instant fuel consumption sender or not. IMHO it is, since you don't guess anymore if and how much each injector squirts.

It looks pretty good, but as far as being the best or most reliable, that's totally dependent on the design and build quality, which I cannot comment on as I do not have any experience with that product.

I guess I wasn't explicit enough. I rephrase: is that a more accurate and easier method to measure instant MPG than what has been discussed till now?

It appears to be something that could prove to be very useful, is there a trip computer that will readily hook up to this device's output?

Not to my knowledge, because I didn't look for one.
My intention is to design and build an on-board multifunctional computer.

I'm not sure I understand your concern about guesswork. Knowing fuel injector duty cycle and fuel pressure gives you dependable consistent data. If you're doing DIY work and want absolute numbers then you need to calibrate your measurements, though you can probably get accurate numbers without calibration if you factor in your fuel injector's specifications.