of VAG-COM and 200q20v's (part one)

[Brett Dikeman]

At 3:33 PM +1300 2/18/01, Dave Eaton wrote:

>how long did you let the car idle after you ran it?  if over 5 minutes have
>elapsed, you need to re-run the car - the manual states "if the engine runs
>at idle speed for a longer period, the specification may be exceeded without
>an actual fault existing".

Hmm.  Didn't notice this...are you sure this isn't something specific 
to the UrQ/RR?  I'll check in the Bently more closely though. 
Basically, I let the car warm up to the point where it hit(and 
stayed) within the range recommended for channel 1, motor 
temperature.  I don't think it ever topped 200.


>also, the actual value for display field 7 is amended in later 20vt manuals
>(aby, aan) to be from 41-61.  as the isv hasn't changed, you are ok anyway.

okay, sounds good.


>if required, the isv should be checked 2 ways - static resistance across the
>two terminals, and using the output tests of with the isv unbolted, and
>checking to see whether there is any "stickiness" to the movement of the
>rotary slide valve.  don't touch the slide valve - if looking a little
>sticky or slow, you can clean it with carb cleaner which may help.

the valve is completely clean; I actually cleaned it last night for 
the hell of it.  There was considerable dirt/whatever inside the 
connector and on the terminals, and a spot of oil goup inside the 
vane movement.

I did an output test with it back on the car, and the valve 
cleanly/totally opened, then snapped/bounced shut with a little "ker 
poing"; movement when pushed around with qtip was smooth/fluid and 
even, no sticky spots.

Idle still doesn't seem quite right; still occasionally hiccups and 
drops, labors a little, then recovers.  Sitting at idle at a traffic 
light, it goes smooth-rough-smooth-rough slightly.

At 2800 rpms or so in 3rd, engine speed will fluctuate back and 
forth,  maybe 2 times a second; I can watch the tach wobble slightly. 
This is very repeatable, as it's the speed I do through most of my 
daily commute.

Couple of people have suggested a faulty O2 sensor, but it's not that 
old, and none of the ECU values seem to indicate a problem with the 
sensor.

I drove the car home and letting it idle for several minutes while I 
played with VAG-COM; a little while later after pulling the valve and 
cleaning it out, I started up the car again and recorded a few 
values, they should be tab delimited and CR/LF'd at the end.

1	2	3	4	5	6	7	8	9	10
0	0	0	130	88	131	48	132	116	0
136	28	88	124	82	131	48	129	116	40
171	23	80-90	124	80	130	48	129	116	36
189	23	75-83	126	78	130	48	129	119	36
193	22	76-83	124	78	130	48	128	119	36

comments(for each line)
stopped, off for about hour, engine not running(many values don't 
show up until engine is actually running)
valve cleaned, car starting to warm up

warming up more
warmed up(in range)
same

Does anyone have docs or notes on what the single value readings are? 
Note, these are -not- the "channels" the bently refers to, these are 
internal values from the A/D converter according to Uwe.  The Bently 
is referring to Block 000 channel readings...

He and I tossed around the idea of building tables for various 
controllers that would let VAG-COM interpret them(same for the block 
000 values.)  It's a pretty big effort, because "practically every 
controller's registers are different."

I figure I can probably figure out a lot of single values on my car, 
simply by looking at a block and unplugging a sensor, moving the 
throttle, etc, to see if it .  Granted, it will take forever, but 
when VAG-COM supports logging(Uwe says "soon!") this could allow for 
a complete black-box style log of everything during a test drive. 
Call me nutty and slap me stupid, but I think that will cause a 
literal renessance in terms of the time it takes to troubleshoot Audi 
engine problems; 6 months from now, we could all be posting charts of 
what we're recording along with "it's surging slightly at 2800 rpms 
or so" and someone says, "hey, look at the O2 sensor line, it's doing 
xyz, there's your problem right there"...I think this will be 
especially useful for things like boost problems.
-- 
----
Brett Dikeman				Systems Engineer
ProAct Technologies Corporation		914-872-8043
(formerly CFN[formerly iClick, Inc])		914-872-8100(fax)
120 Bloomingdale Rd.			http://www.proacttechnologies.com
White Plains, NY 10605
PGP Fingerprint: 06C2 5D5B D2B4 7626  BB24 2BBC 9E4A C8B3
PGP Key location: http://pdikeman.ne.mediaone.net/pgp/brett.pgp