Help on turbo app.

[Garfield Willis]

On Tue, 9 May 2000 17:12:39 +0200, Jörgen Karlsson <jurg at pp.sbbs.se>
wrote:

>I look forward to compare a wide
>band O2 sensor and a conventional sensor. I would guess that 14.5:1 AFR at
>1100F EGT and 13:1 AFR at 900F EGT give the same O2 sensor voltage... Any
>comments?

Let's take a for-instance: spose you have an LSM-11 (Bosch 4-wire
"wider-band") sensor, and are just using it without fancy calibration &
temp compensation electronics. Then, over a range of 400degC to 650 degC
(750degF-1200degF if math hasn't failed me), the SAME sensor voltage
could be read for an AFR of from 0.85lambda to 0.937lambda
(12.4AFR-13.6AFR), or a slop of over 1 full AFR. The NTK current-pump
sensors in contrast are said to vary only 0.2AFR over their ENTIRE
temperature range. This is reported in the NTK SAE paper, over the full
range of heater voltage and operating range of EGT for these sensors.

For your question above, I'd say you're maybe being a tad pessimistic
tho, since your example only changes 200degF, so I don't think you'd
expect to see a full 1.5AFR slop from that small an EGT delta. If you
can hold your EGTs to this narrow a range during your run, then the
problem isn't the SLOP vrs. EGT per se, but just exactly WHAT AFR
actually corresponds to the Vout. So now (in your example) it's not so
much a problem of slop, but rather, accuracy; namely, what AFR range are
you actually in. The curves were published for the LSM-11 only because
that was intended to be used with more elaborate circuit means to
calibrate/compensate. The slop in any average everyday OEM switch-type
sensor is both bound to be more, AND you don't know what the actual AFR
relationship is even around that slop, because there are no AFR vrs.
Vout curves published for those devices.

That's a long-winded way of saying, yes you have a goodly amount of
imprecision in your measurement (SLOP), but you also have an ACCURACY
problem as well, since you don't have the AFR vrs. Vout curve for your
sensor either.

Precision vrs. Accuracy may seem like a convoluted tale, but basically
it boils down to how sloppy/repeatable vrs. how true/close the
measurement is to *actual*. If your O2 sensor Vout varies nary a tad
with EGT, but you don't know for your sensor, how it's output is related
to AFR, that's an accuracy/calibration problem. If the sensor Vout
varies widely with EGT (or some other variable, even sensor swapout),
that's a precision/repeatability problem. Ya gotta have both for a
measurement instrument.

However, having said all that blather, IF you know from other
indications (plug cuts, good timeslips, good detonation margin, etc.)
that your AFRs are about right (but you don't know nor care exactly WHAT
they are numerically), and you just want to make sure you're not
straying too far from them during a run, then you're probly going to be
off in the large fractional AFRs at worst (just a guess based on the
LSM-11; I have no idea how sloppy garden-variety OEM sensors actually
are with temp). Then the real question is can you keep your EGTs to this
range in the critical part of your run. It's really true that by
monitoring EGTs and using a switch-type O2 sensor, you can feel around
and home in on the sweet spot (it's done in experimental aircraft
running only mogas, all the time now); it's just alot easier if you have
a real AFR measurement device. When they become affordable very shortly,
the choice will be much easier between the hoop-jumping and fiddling
with EGTs/switching-O2 sensor combos, versus an actual AFR meter. Use of
an OEM O2 sensor alone without EGT monitoring really relies on the EGTs
being fairly repeatable. On some setups and applications, that may be.
Course, replace the sensor with another, and you start over again in
learning where things are.

HTH, (pardon the verbosity, brevity isn't one of my strengths :),
Gar


----------------------------------------------------------------------------
To unsubscribe from diy_efi, send "unsubscribe diy_efi" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org