ThoughtBoxLabs@aol.com wrote:
> I have a 93 Civic 1.5L non-vtec.
My condolences. ;)
> It has a free flow intake and exhaust. I was thinking of adding a fuel
pressure riser and > a lean/rich fuel mixture monitor to keep tabs on the
engine. Yesterday my buddy told me
> that a more reliable way to view danger to your engine was by adding an
exhaust temperature > gauge. Which is the way to go?
Both.
The issue here is that niether gauge is able, in of its own, to give you
the whole picture.
Let's start with the O2 sensor gauge.
Modern O2 sensors are of the "narrow band, switching" type, which means
that they do not read linearly. Instead, they switch from almost-nothing to
almost-1V right at the stoch point (at about 14:1 air/fuel) This is very
useful if you are an OEM, because you can read the sensor as an on/off
switch, instead of having to read a value off the sensor and then do some
match on the value you read. By using an algorithm similar to that used to
adjust the fall of artillery fire (heh, how's THAT for obscure?) the OEM
computer can adjust to 14:1 by switching back and forth across the
transition point on the sensor - add fuel untill the sensor reads rich,
then pull fuel until the sensor reads lean. Repeat several times a second,
and you wind up averaging 14:1.
However, for making power, most engines like to run in the 12.5:1 to 12:1
air/fuel range, so the switching technique is no good for tuning. Instead,
you want to read a value off the sensor and work that into an air/fuel
ratio.
There are two problems here - while there tends to be a little "shoulder"
on the O2 vs V sensor response curve in the 13.5: 1 - 12:1 area that means
the sensor reads more-or-less linear in this area, the voltages we're
talking about are VERY small. Going from 0.85V to 0.95V on the sensor can
represent the difference between 13:1 and 11.5: 1 And secondly, the sensors
are VERY temperature-sensitive, reading rich while they are cold, and lean
while there are hot. (This is why they have heaters in them)
So if you use the typical AutoMeter 30-LED gauge, you have to know how hot
the sensor is in order to get a rough idea where the sensor is reading.
There ARE tricks where you can run a bias voltage through the sensor to
work out the current sensor temp, and then run that through a table to get
a corrected A/F value, but that's beyond the skill of most people. Motec
sells a box that does just that (using a Bosch LSM-11 wideband sensor) but
the box runs around $1500.
OK, so how about EGT then?
It is true that, all else being equal, EGTs rise as the A/F gets leaner,
and lower as the A/F gets richer. (Actually, it's more true to say that
EGTs get hotter as the mixture approaches 14:1, and gets cooler as you get
farther away in either direction) However, all else is usually NOT equal.
In particular, ignition timing plays a role in EGTs. The closer the
ignition event occurs to exhaust valve opens (the more retarded the timing)
the hotter the EGTs, and vice versa. If your engine has knock-retard (as
most modern EFI engines do) spark-knock-induced ignition retard can show up
as elevated EGTs.
So really, the best way to get a handle on what is really going on in your
engine is to have both gauges, and use them in conjunction with each other.
Incidently, the BEST possible case is to run 4 EGT probes, one in each
runner, and connect them to some sort of data logging device.
Now, let's look at your other issue - the use of fuel pressure to tune the
engine.
This is, unfortunately, a somewhat ham-handed way to adjust the A/F in your
engine.
The issue here is that A/F ratio is controlled by the injector on time, as
controlled by the ECU. The ECU "expects" a certain amount of fuel to be
injected per milisecond of injector on time. All the maps in the ECU are
programmed around this (along with some adaptive "fudge factor" to allow
the engine to adapt to clogging injectors, clogging fuel filters, and so
on) When you raise the fuel pressure, either with an adjustable regulator
or with an adjustable, rising-rate regulator, you are squirting more fuel
per event than the ECU expects. The adaptive fudge-factor code can usually
work around this, but if you exceed what the ECU can adapt to, you wind up
with substandard performance. On Hondas in particular, this can mean hard
starting, a really crappy idle, and lousy fuel milage. If you go really
nuts with the fuel pressure, you can even damage the injector solinoids.
The problem is that you're not injecting extra fuel only when you need it,
you're adding fuel ALL THE TIME.
The better way to do it is to leave the fuel pressure alone, and instead
adjust the injector on time. You can do this a number of different ways,
ranging from reprogramming the ECU, to aftermarket fuel computers, to
piggyback computers like the ApexI Super AFC (which is a pretty good unit,
and fairly cheap)
Good luck!
DG
****************************************************************************
The information contained in this transmission, which may be
confidential and proprietary, is only for the intended recipients.
Unauthorized use is strictly prohibited. If you receive this
transmission in error, please notify me immediately by telephone
or electronic mail and confirm that you deleted this transmission
and the reply from your electronic mail system.
****************************************************************************
/// autox@autox.team.net mailing list
///
/// To unsubscribe send a plain text message to majordomo@autox.team.net
/// with nothing in it but
///
/// unsubscribe autox
///
|