Measuring low frequency average voltage

I have a question about measuring the average voltage being delivered
to a circuit. Basically what I want to do is tap into the line coming
from the ECU to the fuel injectors in my car. The ECU sends a pulse
signal to the injector to signal the injector to open and start
spraying fuel into the intake. I figure that if i can measure the
average voltage on this line then i can know what percent of the time
the injector is open. then knowing the fuel delievery rate of my
injector and my speed i should be able to calculate an instantaneous
gas milage value. Eventually i'm going to do this digitally with A/D's
and a LCD output so that i can monitor a number of variables. But
right now i want to do it all analog. That it's I'd just like to have
a (analog) voltmeter that reads a voltage that is proportional to the
average voltage being sent to the injector. I figure that for high
frequencies that any typical analog voltmeter is going to read the
average voltage, but what happens at idle when the frequency is going
to be something like 4-8Hz? does a typical analog voltmeter average
over that range too or does it start to look DC at 8Hz?

(for the car experts out there: i can't use OBD-II and get this info,
as my car is an '89 and predates OBD-II by many years).

if anyone has any ideas that would make my life easier, i'd appreciate
it!

mike

 

What you possibly want to do is to integrate the pwm signal going to the
injectors into a dc voltage. You can do this with a resistor in series
with a capacitor to ground. The other end of the resistor connects to
your injector signal. Start with about a 10k resistor and a .47uF to
1uF cap. At the junction of the resistor and cap you will have a
fluctuating DC voltage. By playing with the cap values you will affect
the response time of the derived DC voltage, a scope will let you see
the ripple. If you need to draw any current from this junction, then use
an op-amp with the gain set to 1 to boost the available current.

There are two ways to calculate the optimal values for your cap:
1) Use a bunch of math and factor in a ton of variables you probably
don't know yet
2) Tinker with it like a "real" engineer and watch it on a o-scope or
DMM till you like what you're seeing

Guess which way I would do it. ;-)

Depending upon the accuracy required, the "more correct" way to do this
may be to measure the number and width of pulses over a known time
interval; PIC chips come to mind. You should also factor in the
instantaneous voltage applied to the injectors since the ECU uses that
when calculating the pulse width to use. The injectors will open faster
at higher voltages letting in slightly more fuel for a given pulse width
and fuel pressure. IOW, just knowing the average voltage may not be
accurate enough for you. OTOH, it may be more than accurate enough.

Good luck, sounds like lots of fun. YMMV