Fuel Management Summary:
All fuel managements seem to have pro's and con's,
it's just which one has the least compromises for the engine you're
assembling.
In our case, pressure sensing, or speed density
as it's also know, shines when it comes to pro's, where all the
little con's of the other methods seem to pop out. Pressure sensing
gives us the ability to calibrate turbocharged engines
in the way that the engine itself works, i.e. Pressure. The engine
utilizes boost pressure to create more power, and along
with that boost, comes changes required in fuel and timing. Pressure
sensing uses pressure as the primary input and is
easily calibratable for these changes and can accurately and easily
obtain a very precise tune for all the conditions which
turbocharging poses to the engine.
On the other hand, mass flow, air flow meter, and
throttle position sensing have the key disadvantage that they
don't know the pressure. They know many other things, and
all have their own advantages, but with turbocharging
comes another dimension that they just don't seem to handle very well.
Mass flow works well with OEM, and lower horsepower
applications, but when it comes to turbocharging, and high
output engines there are quite a few disadvantages to consider.
The first would be the increased power of the system
and the ability for the mass flow sensor to operated within it's range.
If you're beyond the very limited capabilities of the
original mass flow sensor, you can replace it with a larger unit, but
this also has it's disadvantages. For instance, a OEM
mass flow sensor capable of monitoring a 250hp engine needs to monitor
somewhere between 25 to 30lb/min of air at it's
highest point, but this isn't even the problem. the problem comes
into play at the low end. If the engine makes it's 250hp
at about 6000rpm's, and consumes approximately 30lb/min, then at 1000
rpm's it'll consume 1/6th that amount or 5
lb/min. Now, that doesn't sound to bad, but if we only have an
8 bit resolution, or 256 points, to work with, that gives us
about 40 or so points for the full range from off to full throttle
at 1000 rpm's, vs. the full 256 points of resolution that a
pressure sensing system would have throughout the range. Now,
this still isn't that bad because we're only metering
250hp, but let's turbocharge the thing and make 500 hp. This
now doubles our mass flow that we have to meter within,
so at 1000 rpm's, we are reduced to under 20 or so points from full
off to full on throttle, and the air fuel curve is beginning
to look more like a set of stairs rather than a linear line.
As well as, at idle, with a/c on and off and small changes are
barely or not even noticed and the air fuel ratio does not stay accurate.
At 6000 rpm's, where you'd be making full power,
from full off to full on would get you nearly all the points regardless
of the range of the mass flow sensor and it would give
you a very nice representation of the actual air flow input to the
engine.
With pressure sensing, since we work with pressure
as the primary input, regardless of the engine output from
250hp to 1000hp, as long as we're working within the same pressure
range, the resolution hasn't changed and
drivability isn't sacrificed. On the other hand, if you want
to go to Formula 1 and run 5 bar boost pressure, you'll
definitely be needing more than 8 bits of resolution on you analog
to digital converters in the computer.
Throttle position sensing might get you the full
256 points of resolution from off to full on the throttle, but at 1000
rpm's, you're at full power with very little throttle due to the fact
that you just don't need to flow that much air down there.
So, you run into the same problem again, you're trying to get an accurate
air/fuel curve based on a very limited resolution.
Another interesting problem with mass flow sensors
is their ability to measure air flow in both directions. This
presents an interesting problem on turbocharged engines, in that, when
you come off boost, the air flow reverses and
comes back out of the air flow sensor. This is more of a problem
with very large intercoolers, and even with the largest
blow off valves recirculating air back into the intake, you still get
a considerable amount of air coming back out the mass
flow. This presents a problem. The computer sees air,
it gives fuel, but the air isn't going IN the engine, it's coming
OUT! But the computer is putting fuel in, so you usually end
up with a very poor driving condition, or even killing the
engine with excessive amounts of fuel. The solution to this is
to put the mass flow sensor very close to the throttle body
so the air flow reversal is all happening upstream of the mass flow
sensor. But now, you end up with pulsations and very
uneven air flow metering due to the throttle body and proximity to
the intake manifold, so your air flow curve is a mess.
You also become limited to a single mass flow sensor that must flow
an enormous amount of air and the low speed
resolution gets pathetic again. Around and around we go, the
answers seem to be pointing in the obvious direction...
Dump the mass flow sensor and go pressure sensing!