I had a bash at cleaning my MAF with the aid of OBD graphing if anyone is
interested in what happened when I did:
Test starts, my MAF is reading 0.524 lbs/min airflow, the car is at 850 rpm
and cold, the A/C is switched off.
The flow rate gradually reduces, and as the engine warms up goes to 0.508 and
down below 0.500, reaches 0.497.
Now the engine has warmed up to normal and the MAF is reading 0.440 lb/min
rate. Time to rev it up a bit.
At 2397 rpm th MAF is reading 1.366 lb/min and returning the revs to idle, at
872 rpm the MAF is reading 0.413 which is about as low a reading as it has given
More revving, 3087 rpm gives 2.157 lb/min, back to 0.420 at tickover. This
seems to be stable now at low 400 thousands of a pound per minute at tickover.
Putting on the air conditioning, and the MAF reading rises to 0.575 at 850
rpm. Don't know why the AC causes the reading to rise, unless it is to do with
the cooling effect on the airbox which is near the AC valve.
AC back off, 857 rpm and 0.407 lb/min.
Of to mess with the MAF...
Airbox off first, five screws and a couple of clips hold it to the bottom of
the airbox and the MAF itself, I took the opportunity
to replace the air filter at the same time although the old one looked ok
really, and I kept that as an emergency spare. This may have affected airflow in
the sense that more air will pass through the new filter than through the old
one. As will be seen, the opposite effect is observed, which should imply either
the filter made no difference or the cleaning was even more significant.
A jubilee clip holds the MAF to the next element in
the airflow route, some sort of cube with big air hoses. There is also the
electrical connection to the MAF, which is the usual
ford type with a retaining clip, press in the metal wire clip and the connection
comes off easy as anything.
The MAF is a short tube with a cross member, one
end of which has a small tube containing the two wires which are the sensors
giving the reading. When I get this off, the wires are a dull grey or brown,
being quite hidden away the colour isn't obvious.
Ok, so out with the "carb and injector cleaner" spray can, I fasten a little
tube on the end of the nozzle, test it, and then spritz away liberally. Cleaner
drips out of the other end of the sensor cross member, but the wires are
But not quite clean, now that the wires and chamber are mostly shiny silver,
a few dregs of dark matter remain, so some more fizzing with the spray can until
they are completely clean. Then a couple of squirts down the tube into the
engine, clean up the MAF fittings and a drop of light
oil to ensure a tight seal around the rubber ring and the whole thing goes back
together the way it was to begin with. This isn't entirely trivial, the air
boxes are flexible and move to the right as you have the air filter box off, so
you have to shove it all back to the left before it fits properly, not too hard
to do but worth expecting this.
Also, don't forget to connect the electrics, this is probably the one that
will be forgotten if anything is. Who knows what happens if the maf isn't wired
Ok, fired up again, the OBD is reporting 0.418 at 860 rpm. This is unchanged,
at first glance, but the engine has had the opportunity to cool off a bit.
Revving things up, at 1947 rpm the reading is dead on 1.000 and at 2962 has
risen to 1.460 lb/min and these figures are considerably lower than they were
Returning to tickover, 867 rpm and the airflow is now 0.377, the lowest
reading so far. Rev up and down, the setting is still on 0.377 at 868 rpm, a lot
more consistent than before but that's probably just chance.
Air conditioning on, and the reading goes to 0.554 at 861 rpm. Again, this is
a bit lower than before. Running with revs and the AC on, the OBD shows 1.858
lb/min at a steady 3000 rpm, higher than with the AC off, lower than with the AC
off and before the cleaning.
Turning off the AC, at 2910 rpm the MAF is measuring 1.440 lb/min.
Conclusion of all this is simple enough. The MAF contains two wires, one is
very hot and any muck entering the unit will basically burn onto the wire and
reduce its function. Basically, more heat must be pumped into the wire to keep
it at the set temperature because it also has to heat up the dirt, this
translates to the sensor imagining that more air is passing by the wire to make
it that much cooler.
And, since this unit is not cleaned by an automatic process or during routine
servicing, it is inevitable that over the years the MAF will become gradually
dirtier and this will cause it to read more air passing into the engine than is
really happening, and the fancy automatic computer control metering the fuel and
everything will be misled and will obviously pump more fuel into the engine than
is actually needed. Conclusion, clean your MAF regularly because it will become
inaccurate and it won't be cleaned by anyone else, and the effects of a major
engine sensor giving readings that are up to 30% or more higher than they should
can obviously be quite extreme, e.g. strange autobox selection profiles are just
the tip of the iceberg, and heavy fuel consumption ought to be expected.