Return of the diy_EGOmeter (YES, it's finally ALIVE!!)

[garfield at pilgrimhouse.com]

On Wed, 08 Apr 1998 17:00:31 -0700, garfield at pilgrimhouse.com wrote:

>it's a comin,
>honest engine. Give me another week before you tar and feather me for
>dragging my virus-infested butt. Whimper. Man how can such tiny lil
>critters make you feel like sucha pieceOyou-know-what.

Man, that was almost a week ago, and I'm way worse than I was then. What
a sissy! I'm tellin yous man, these virii vermin are kickin me
buttocks!! My doctor wants to hospitalize me if I develop even a hint of
a secondary lung infection; I'm right on the doorstep of viral
pnuemonia! Whimper is right! Crapola, what a BORE.

BUT, do I have good news for YOUS GUYS!!! I got in some quality time
today testing our diy_EGOmeter, and I just got finished sending off a
missive to Frank with the nitty gritty details, but suffice it to say,
it looks like we have a *weiner*. The free-air O2 ion pumping current my
circuit settles at (the so-called free-air calibration point) is only 5%
off from what Frank gets with his NTK sensor and interface box!!!! This
is stunning confirmation that we are go-for-launch, and that the Honda
OEM sensors are indeed essentially the same as the NTK sensors. I
measured 6.3 mA @0.445V, and he (Frank) gets 6.0 mA @ 0.428V. Before you
think them stoich voltages are somewhat different, lemme tell ya that we
discovered there's a real black-hole steep Ip null right at stoich, just
like that cliff-edge part of the conventional Lambda curve. The voltage
of the measurement cell, VsCell, goes along niceNlinearly heading for
stoich, as you crank up the O2 cell current, IpCell, but just as you get
to 0.450V, all hell breaks loose, and even a tiny lil tweak of IpCell,
and you're on the other side of stoich by 200-400mV. So what I'm saying
is that Frank's instrument's feedback loop locks stoich up at 0.428V,
and mine at 0.445V (hey, mine's more precise than NTK's, hee hee;
alright, it's also a tad twitchy at the moment), but for all intents and
purposes, it's smack on the same point in the curve of IpCell vrs.
VsCell. Hence, the free-air calibration currents diff. by only 5%.  You
couldn't ask for a better reassurance that the Honda OEM sensor and the
NTK "special salsa" version are essentially the same, except for their
method/values of embedding CAL resistances. The most important thing is
that since the sensors perform basically identical, we can just plan on
calibrating the boxes each of us build with this "free-air
self-calibration" procedure, and care not a wit about what the internal
CAL Rs are set at, for their various original usage. "We don't need no
stinking CAL Rs", is the bottom line. Besides, as I mentioned to Frank,
use of the CAL Rs also can't compensate for sensor aging nor
deterioration due to lead/dirt contamination, like a self-calibration in
free-air can, so it's better we don't mess with them Rs anyway, eh?

Actually, I have a theory about what the CAL Rs in the Honda sensors
actually are, and will test me hypothesis when I get a couple more, but
the main thing is we aren't dependent on even knowing what they mean,
since the sensors can be self-calibrated in free-air. In fact, this is
THE method of choice for the Horiba boxes, since they don't ship their
sensors with ANY embedded CAL R in the cable. They ASSUME you're going
to self-calibrate it yourself.

I mentioned to Frank that I have a couple tweaks still needed to the
circuit, one to detune the hyperactive feedback; I didn't count on the
Ip vrs Vs relationship being so sensitive near stoich (hence it twitches
on it's way to settling, instead of gracefully landing via a smooth
approach, when you first turn it on), and I need to replace one of the
parts in the "warm-up" circuit with one I've got on order; the current
part doesn't like anything but 12V supply (some times it switches,
sometimes it lounges), and I want the whole thing to function on 5V. But
the bottom line is, at this point, looks like we get our O2 meters
finally! As soon as I get the last tweaks done and everything checked
out completely, I'll post the schematic & parts list up in my group's
aviation DropBox, which I'll post the URL for when the stuff's ready.
Give me another week or so to recover and clean up things, and we can
all go about building a clone of ole diy_hUEGO. [Too bad they don't make
the O2 sensor tips in the shape of a cone; that would be so fitting].

Cheers,
Garfield