[Diy_efi] WB free air calibration

[Garfield Willis]

On Wed, 29 May 2002 21:06:21 -0700, "Rob F" <rncfiles at hotmail.com>
wrote:

>From you testing of many diy-WB circuits, what has been your observed =
AFR
>error rate?  You quote percentages, but I am wondering how much is
>conjecture compared with scientific experimentation?

OK, since you asked (thank you thank you), I'll tell ya. We have two
'dweeb-WB' circuits here we use for demos for those customers that have
heard of it, and ask questions. Yeah, I guess we 'violated the so-called
user agreement' to built them, so...umm, sue me. :)  One using the
original board we got from Bob V. & kit from Steve C., and one from
Peter G. in Oz which we bought assembled, again to avoid suggestions
that one or t'other weren't built correctly. Friends on this very list
handled the buying of them for us, so we could do our research quietly.=20

They BOTH function exactly the same, since they're essentially carbons
of each other. Of course, neither works properly until you get up to ALT
output voltages, so they both expose the sensor to late warmup, and from
sensor to sensor, they each may take from 30secs to a couple minutes to
NEVER, for the sensor's heater to warm up properly, once the ALT is
turning. Depends on the sensor. The heater resistances especially vary
alot, the heater control in the dweeb-O2 is also just simply WRONG.
Wrong method for ramping up the temps, not ever recommended by NTK, and
wrong/no method for maintaining a controlled temperature. Consult the
original SAE article by NTK or just observe with a scope how any NTK or
Horiba box heats up the sensor, if you want to see how it's supposed to
be done. These dweeb-O2 guys apparently didn't even bother to look to
see how it was done correctly by everyone else.

I've run around 60+ different sensors now thru BOTH of these devices
since they were built/obtained, on our full-sweep gas calibration bench
and in client & test vehicles, mainly to avoid suggestions that
something 'just must be broken' with one or the other. On our main test
vehicles, the exhaust systems have MANY O2 and EGT bungs in every
imaginable location.

The errors with these two 'dweebs' using this large a spread of sensors
are HUGE. Sometimes as much as 1.5AFR, and that's ignoring EBP errors.
Partly due to the fact that the Cal R circuitry is just plain WRONG in
the design (which is why the free-air values are all over the place),
and partly due to the fact that the pump control circuitry used causes a
hysteresis effect whenever you quickly go rich or lean across stoich,
which throws the transient results especially off by as much as 1AFR.
Yes, one whole AFR. I have it on good authority that the dweed-O2 Cal R
circuitry was 'reverse-engineered' based on measuring just TWO sensors.
Yuh, good plan. I could quote the person involved, who didn't know he
was communicating with us at the time, but I wouldn't want to embarrass
him. :)  I even have drawings and explanatory notes; we do our intel
homework.

Oh, and if anyone questions the use of our gas calibration bench, let me
just tell you that we have a Horiba and two NTK blue-boxs here also,
that we run along side in our test chamber with EVERY calibration, just
in case. Next month we spend the big bucks for a Bosch LA-3, just for
the purpose of demonstrating to clients that we know how to measure AFRs
correctly along with the big high$$ boys.

Our full-sweep calibration bench allows us to sweep the full range of
mixture sample gas concentrations from less than 10AFR all the way to
free air. This way, we can measure & compare entire calibration CURVES
for each sensor, not just a few selected points. During this sweep, we
can also vary the exhaust back pressure by as much as 5 ATM if needed
(tho we normally only go up to 3 ATM absolute, unless you're one of
those crazy tractor drivers). This is how we can offer a full-gas
calibration and EBP calibration on our sensors. Consider the expense
we've gone to, in order to offer this level of QA. AFAIK, only Bosch and
Horiba offer this level of support, and then only to select customers.

>Oh and please don't use my questions or comments as a springboard for
>personal attacks on others.

Gee Rob, no, wouldn't think of it. I'm not attacking the person, just
their trash. Yeah, attacking the person spreading their trash would be,
umm, un-PC.

Gar

BTW, to illustrate a simple contrast, here's some specs for you data
hogs: our EGORmeters will operate on power down to 8V, FJO 12V, ECMs &
NTKs require 11V or better (same design), Horiba's 10V, and the Bosch
LA-3 big daddy ($$$) will actually run OK all the way down to 6V power!
'Dweeb-O2' requires 14+V to operate, and will NOT operate on even a
fully charged +Batt w/o the ALT running, so it won't even begin to warm
up the sensor until the exhaust is already flowing. What's wrong with
this picture?


_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi