I missed dinner last night... :) (long)

["Linus D. Toy" <linust@interramp.com>]

or rather, missed eating dinner with my wife.  But no matter how much she
might have yelled at me (she didn't), you couldn't wipe that silly grin off
my face last night.  it's still there this morning.  you'll see why in a minute.

Yesterday afternoon, I got a call from a certain Olympian (as in Olympia, WA
resident) involved with ECUs.  Yes, Mr. Intended Acceleration.  "Linus, can
you do me a favor?"   huh?  me?  Given the vast car computer knowledge
disparity between Ned and me, "I'll try" was about the most I could answer.

Ned wanted to test some computers from customer cars that reportedly didn't
work right.  (If any of you want to fess up on your projects, go ahead :)
They tested fine in his car, but as he put it, his car wasn't an ideal test
bed.  He was looking for a stock test bed.  In return, I could try out a few
different computers.  hey--that's a no-brainer, come on over!

About 7:00pm, Ned and his daughter Marnie showed up with his Pearl 5kcstq.
"Some" computers was four computers for 20v turbo engines, in various stages
of programming.  The plan was to see how my car behaved with them.  Sounds
simple--just find a nice hill where to load the car and drive.  For kicks,
he also pulled the computer from his car, so we had five computers to try,
plus my original.

To start, Ned plugged in his V.A.G. 1551 to check my car's operation.  The
1551 is the diagnostic computer that interfaces with the car, downloads
fault codes, and fixes dinner for  you.  Some of you may recall I previously
reported poor cold engine performance with my car.  Sure enough, when we
fired it up, it ran rough for the first few minutes.  The 1551 reported no
faults, but did show fluctuating idle speed and a widely fluctuating O2
response value.  hmmm bad O2 sensor?  but no faults.  we'll see this again
later.  So, all plugged in, we headed for a nearby hill.  All tests were for
drivability, not to test raw power output, so top speed wasn't the
objective, no times were kept--the goal was seat of the pants feel.  The
tests were NOT identical in terms of load, speed, road condition, etc.  Part
of that is operator error--Ned asked that I drive (hmmm...)  Given the
nature of testing (full throttle up short hills sometimes contending with
traffic), it was fortunate that Marnie sat in the back seat to watch the
boost gauge.

Original computer:  Car warmed up, we started with my computer.  The car ran
as it always has, with the boost "coming on" with a kick at about 2500 rpm
(on this hill).  OK, this was our "control."

Test I:  This box was programmed to IA Stage III levels.  This test was run
on the same hill near my home (SE 24th for those of you familiar with Mercer
Island).  This box produced a similar kick to stock, just more of it, with a
fuller feel.  The pressure transducer on this computer goes to 2.5 kPa
instead of the stock 2.0.  This causes the dash boost gauge to read low--the
most I saw on the boost gauge here was about 1.7, but i think that
translates to about 2.1 or 2.2 bar boost.  Something new that I heard:
under heavy boost up the hill, a low frequency air "puff" as the computer
works to control the wastegate more than normal. 

Test II:  This box was marked "90 Coupe Quattro."  Someone out there has a
real sleeper.  Marnie used the word "S2" in association with this box.
There were no customer complaints about this box--see Test III.  I don't
know the differences between this box and the Stage III box, but it felt
better.  EXTREMELY SMOOTH power delivery, less of a kick than the Stage III
box, but stronger, fuller.  This was probably the nicest box I drove--When
(not if) I modify my computer, this is the one I want.  I didn't observe the
boost gauge much with this box, but Marnie said it peaked at 1.9--again,
this is low by 20-25%, maybe more.  

Test III:  This box was programmed identically to the Test II box.  After
putting this box in, Ned checked the 1551 again.  Guess what--I've got a bad
O2 sensor signal.  No big--most of our testing is at WOT anyway <G>  We
moved our testing over to the Issaquah-Fall City Road for a longer, steeper
hill, less traffic.  This box and the 90 CQ box were tested by the customer
and sent back for comparison.  No problems with the 90 CQ box, but this box
caused jerks and hesitation in 1st and 2nd gears.  No such behaviour in my
car.  In top gear, at WOT, we did notice a slight anomaly, almost like a
miss;  nothing constant or regular, just once or twice on I-90--oops, 90mph,
better back off.  For some reason, this box was not quite as smooth as the
Test II box--you could feel more "kick".  Ned attributes this to production
tolerances in the analog devices (resistors, etc) and the pressure
transducer (which, btw, is hand made). Similar power delivery though.

Test IV:  This was the last customer box.  It was designed to be used with a
different exhaust manifold and turbo, which my car obviously doesn't have,
so we didn't reach its full potential.  The kick was back.  Wow--this thing
builds boost fast.  I had previously ridden in another '91 200q with
software similar to this box--the owner lit up the back tire going around a
corner!  I didn't try that, but this was a noticeably stronger box in the
mid-range.  Hey--I'm halfway up the hill doing 80!  Didn't get a peak boost
reading off this one--saw 1.6 and 1.7 easily.  For my daily use, this box
would probably be OK, but at the track, this one could cau$e problem$.  

Test V:  Computer from Ned's car.  This is the box that Scott Mockry wrote
about from our Bremerton event.  This is a whole new ball game...the 2.5 kPa
transducer is replaced by a 3.0 kPa.  Again, this box needs a different
manifold and turbo to really shine.  For our limited tests, this shouldn't
be a problem.  The real problem was--no suprise--the driver--I've never used
up first or second gears so fast!  I took my foot out of it before long--I
don't have an R$D budget for my car!

End result of the testing:  all the computers behaved just fine in my car.
Problems in the other cars?  One car had had its engine washed, so Ned
thinks the throttle sensor is shorting out (car would not build boost, but
computer did fine in my car).  Other computers were from project
cars--remember the '90 Coupe Q?  Never came with a 20v turbo, at least not
in the US.  This is a project car of unknown pedigree--the usual caveats
apply (how hard was the donor car hit, how long did it sit, etc.)

During the course of the evening, Ned and Marnie had some great stories
about their recent travels (Europe, East coast), purchases, future projects
(Marnie's car to get a new engine--someday).  A hilarious story about a
couple in Seattle who each had 5k turbos--her stock sedan, his modified
wagon.  He kept the secret for a couple years until one day when she managed
to borrow the wagon from him--"what did you do to your car?!?"  I got to see
a few neat tools, including an EPROM burner with about a dozen stickers from
a competing company on the back.  When Ned modifies (make that, "fixes") a
previously modifed computer, he keeps the sticker, kinda like tally marks on
a fighter!  Interestingly, only one company's stickers were there.  Guess
which one.  Coincidentally, the EPROM burner is an "S4" model--Ned put an
Audi sticker just above the model label :)

One last tech item:  those of you with earlier cars (like Ned's) have an
easy time getting your computer in and out of the car--it simply sits on a
shelf/tray under the kick panel.  On the '91 200q, the computer is held by
two nuts on the trailing edge and a spring steel clip on the leading
edge--that clip is almost impossible to get off--Ned spent a good 5 minutes
on that clip alone.

The evening at a close, Ned reassembled my car (bummer--he used my original
computer), I thanked Ned for the chance to sample his wares, and headed in
for a late dinner--silly grin and all!

your faithful reporter--
----------------------------------------------------------------------
* linus toy           email:  linust@interramp.com                   *
----------------------------------------------------------------------