MPG meter. Could it be done another way?

[NEIL_N]

Hi all. I'm a new member and find these threads very interesting reading.

I had posted this another thread but thought it needed to be re-posted as a new thread.

I have a Mazda MX-6 (equiv to Ford Probe) 2.5 V6, And would like to know how much fuel I'm using. I do have an old Zemco ZT10 Fuel computer but it only works on carburettor engines with low pressure fuel delivery.

So. Could fuel consumption be measured using the following idea:

Generate two DC voltages. One proportional to injector pulse width, The other proportional to road speed. The DC voltage for pulse width being done by a gated integrator circuit, And the DC voltage being done by using an LM2907 Frequency to voltage convertor IC. The two voltages can then be applied to some form of comparator circuit that sums/subtracts/multiplies/divides or whatever in order to provide a DC voltage that correlates to the fuel being used for the speed (distance travelled).

Would a third reference voltage be needed (equating to the fuel flow rate)?Again all three voltage would need to be compared and an output a DC voltage required to indicate fuel used.

The DC voltage generated could be applied to a moving coil meter, Or an LED bargraph display (LM3914). Obviously the circuit would have to be calibrated. But once the all the DC input voltages have known relative values, calibration of the display device shouldn't be to difficult.

Any further comments or suggestions or ideas on the comparator circuit please

Cheers

Neil

[theholycow]

It sounds awfully difficult. It sounds like more work than using existing strategies. If you can do it and document it well, it would still be nice to have another option all figured out.

[GasSavers_RoadWarrior]

I'm usually "Mr Analog" when it comes to simple circuits, some of the Basic Stamp projects are "Mr Sledgehammer, meet Mr Walnut" affairs, but I do think it would be overall more advantageous to do it with a digital system. The duty cycle meter with a cap and a VM is simple enough, but once you're into building an analog computer with chips that are hard to get in DIL or anything you can see to solder, and they're hard to get in single unit quantities, then you get to thinking that maybe a PIC or something isn't that bad an idea after all.

[mrmad]

Unless you were implying your circuit already does this, I think the one thing you're missing is the tach signal. You would need to multiply the injector pulse width output by another proportional voltage form the tach signal to create the injector duty cycle.

I looked into doing something similar with analog circuits. I'm not an EE, so I wouldn't know how to design the individual circuits without this turning into quite a project, but I did manage to find circuits on the internet that did the following:

1. Create a proportional voltage output from the injector pulse width
2. Create a proportional output from the tach signal
3. Create a proportional output from the vehicle speed sensor (Vss).
4. An op amp based multiplier/divider circuit.

I was going to take the output voltage from the Vss and divide it by the voltage created by the product of the pulse width/tach signal circuits. In theory, the output of the multiplier/divider circuit would be mpg. This would only be an analog mpg "guage" and would do nothing to create average mpg.

I was going to calibrate it by using the portable 2 channel o-scope I have at work and take measurements of the pulse width and tach signals will driving under different conditions (heavy throttle/low throttle, etc) and calculate the fuel used by determining the duty cycle and multiplying this by the injector flow rate.

Of course I've never gotten around to doing this. I think it could be done for ~ $50 to $75 in parts depending on how elaborate the case and display would be.

If you're interested in getting some of the circuits I found, IM me and I can send them to you.

[theholycow]

Originally Posted by mrmad

Unless you were implying your circuit already does this, I think the one thing you're missing is the tach signal. You would need to multiply the injector pulse width output by another proportional voltage form the tach signal to create the injector duty cycle.

I thought the same thing, but in fact it's not true. Duty cycle is simply percent of time that the injector is open. A 50% duty cycle means that the injector is squirting 50% of the time, regardless of RPM. You could say it's open for half of every second, or for 30 seconds out of every minute; either way if the duty cycle is the same then it's flowing the same rate at any RPM.

I've been using a $33 dwell meter to measure my fuel injector duty cycle for a couple months and it's great, and what I've learned from it has in fact increased my average MPG.

[socalsupehero]

Somethign similar to what you are talking about is being developed by a few folks over on ecomodder.com. Basically involves tapping an injector to monitor pulse width and tapping the vehicle speed sensor.

[mrmad]

Originally Posted by theholycow

I thought the same thing, but in fact it's not true. Duty cycle is simply percent of time that the injector is open. A 50% duty cycle means that the injector is squirting 50% of the time, regardless of RPM. You could say it's open for half of every second, or for 30 seconds out of every minute; either way if the duty cycle is the same then it's flowing the same rate at any RPM.

I've been using a $33 dwell meter to measure my fuel injector duty cycle for a couple months and it's great, and what I've learned from it has in fact increased my average MPG.

I think this depends on how the circuit is measuring the signal. I'm pretty sure the circuit I found is measuring the pulse width not the duty cycle. This would be independent of the amount of pulses. Multiply the two and you would get a voltage proportional to duty cycle

[severach]

You'll need a fuel pressure meter for this to work. The fuel rail is pressure regulated. At heavy throttle the pressure is higher than at light throttle so the same pulse width will deliver a different amount of fuel. This helps get more dynamic range from low cost injectors. Then your fuel pump ages which decreases the pressure.

[theholycow]

That's the first I've heard of varying rail pressure. Everything I've read points to fuel rail pressure being constant (or as near constant as possible). Can you point to any documentation saying otherwise?

[mrmad]

Originally Posted by severach

You'll need a fuel pressure meter for this to work. The fuel rail is pressure regulated. At heavy throttle the pressure is higher than at light throttle so the same pulse width will deliver a different amount of fuel. This helps get more dynamic range from low cost injectors. Then your fuel pump ages which decreases the pressure.

Most fuel injected cars I've dealt with have an electrically driven fuel pump and the pressure is constant. Otherwise the fuel pressure would be monitored by the ECU and this would be caluclated in the fuel maps.