Tailpipe Probing of EGO

[garwillis@msn.com (Garfield Willis)]

Pardon e'moi,
I tweaked the subject to segue to the more techo side.

Gar

On Wed, 28 Jun 2000 18:46:43 -0400, "Jason Bonie" <gmecmml@cfl.rr.com>
wrote:

>quick question about the probe shown in the pictures.  does it have
>to have that much "pipe"? sticking out of the exhaust?  what if you 
>wanted to take a spin while watching the meter?

Well, yeah, that particular probe WAS used to take quite a few spins. :)
Some of them in that twin-turbo Mazda were pretty dizzying! :)  Why
would you think not? BTW, that particular length we settled on for
conservatism's sake is only 16" long. Might look longer in the pix than
it really is?

Here's the deal about the "aft end length" of any proboscus (I posted
about this before, but 'oh well'). You MUST make it long enough aft the
pipe that you get NO reversion from exhaust pressure pulses. I've seen
some other vendors shorten theirs to near 6", and all I can say is,
"fine, make it as short as you can, but make sure you see NO difference
if it's 6" or 26"". After all, it's a MEASUREMENT we're after here, so
it wouldn't do to wank it up by skimping on the aft-length and suck
reversion. Personally, until it really gets in the way, I'd make it as
long as possible. As long as that probe might seem to be, all's that's
happened in MANY test runs is we get quite a LOT of "oogles" on the
road. In fact, I've raised up the Meter to eye level on a couple
occasions to answer some VERY curious looks from other proformance
"afficiandos". They seem to instantly sense some tuning operation is
going on. :)

I can tell you from measuring "in the bung" near the exhaust port or
header pipe, compared with using the proboscus, that except for that
narrow band around STOICH, the CAT really IS outta the picture (thank
you, Dr. Heywood). Sure, you cain't hardly see the closed-loop cross
counts, and that's good proof that the cat IS working as it should, but
for any OTHER range of AFR, outside 1AFR centered around STOICH, you
most definitely CAN use a temporary proboscus to get your numbers for OL
tuning, because the CAT during those runs IS outta the picture. Nice to
be able to "switch" it out by the press of the pedal, eh?

WELL, in fairness and for fun's sake, I MUST mention another exception
to the above. There IS a very real value to at least looking at the AFRs
pre-CAT. We had occasion to tune an opposed 4-cyl experimental aircraft
engine (water-cooled) that was mounted in an auto for testing, and were
able to see a mixture imbalance from side-to-side at the conjunction of
the Y-pipe just prior to CAT, from the way the AFR jittered backNforth
at ALL rpms (it was worse under load; ick :). This says that despite
questions about the sensor's response time, it most certainly is
adequate to see if you have cyl-to-cyl AFR variation, if you look just
aft the collector. BTW, testing on another exact same XA engine but with
different induction, no jitter. We were able thereby to point the finger
at the different/"newNbetter" induction setup, which was why one side
was running leaner than the other. OTOH, looking at the same aggregate
mixture at the tailpipe, the gas had mixed and averaged to the point
where we got the same AVERAGE readings for AFR, and you couldn't see the
jitter any more.

Soooooo, to make a short story long, like I said some time back, we
don't INSIST you use EGOR with a probe, just that once you've sniffed
the mix pre-cat, you CAN practically go back to sniffing the tailpipe
for most of the rest of your tuning. It's alot safer for the WB sensor
if you baby it and treat it as part of your "instrumentation gear".
We'll supply you with the schematics for building a "stoich-crossing"
simulator to drive your OEM ECU, if you INSIST on keeping your WB sensor
"in the bung" perpetually to replace the OEM O2 sensor, but frankly, we
see NO reason to do so at this time. When they're (the Honda/NTK
sensors) priced like Pez candy (hee), then sure knock yerself out and
run the WBO2 sensor all the time, but at $90+/pop, it presently seems
"ill adviced". :)

HTH,
Gar


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