Audi 100 Won't Start: Breakthrough ???

[Marc Boucher]

Today I discovered something that has a strong likelihood of being the
problem with my car's starting.

I decided to measure the length of time that the cold start valve was
energized for.

Previously I had done the 'output' test of the ECU which essentially just
shows that it can be energized.  I'd also measured the resistance voltage
across the thermo switch which measured at about 3.5k Ohms and was
consistent with data supplied by others in this group.  Today I got out a
couple of radio shack resistors I'd never used in order to test.

The results: REGARDLESS of what resistance was detected by the wire coming
to the thermotime switch, each time the cold start valve was energized for
only about 1.5 SECONDS !!!  I have Hayes manual that shows that at current
temperature (50F) the cold start valve should be energized between 2 and 4
seconds.  Here is a table of my tests, my expected, and observed results:

Test Description                                            Expected Time
Observed Time
Thermotime Plugged in, 50F                            2-4 seconds
1.5 seconds
Thermotime Plugged out...open circuit                ?
1.5 seconds
11 KOhms resistor (simulates 15 F)                4-10 seconds
1.5 seconds
22 KOhms resistor                                             ?
1.5 seconds
wire shunted...zero ohms                                     ?
1.5 seconds
FI box of ECU unit removed                               ?
0 seconds

So unless I'm missing something, the cold start valve is not being energized
correctly according to the temperature.  Is there something that I could
have missed in order to get the cold start to be energized correctly
according to the temperature?

Would it be safe to energize the cold start valve with 12 volts directly and
try to start?

Anyone have any tips on where to go from here?  i.e. anyone have a wiring
diagram of the ECU's  I guess the next step it to trace the wiring ???


Marc



----- Original Message ----- 
From: "Marc Boucher" <mboucher70 at hotmail.com>
To: "SJ" <syljay at optonline.net>
Sent: Wednesday, January 28, 2004 8:03 PM
Subject: Re: SJ Audi 100 Won't Start (Vacuum Leak?)


It may be a few days before I can get starter fluid and a spark tester so I
thought I'd add some further information.  Each of these may be irrelevant
but for what its worth, here goes:

In lieu of an 'inline' spark tester, I pulled a plug wire, plugged it into a
spare spark plug, and turned the engine over a few times.  Each time, the
spark plug generated a healthy spark from the time at which I engaged the
starter until when the engine finally died.  I know its hard to tell
visually but it really seemed more like the engine died and so the sparks
stopped, rather than the sparks stopped and so the engine rolled to a stop.

I pulled one plug to check its condition.  It had a thin coating of  black
suit, as I'd expect from a car that's been turned over for several days but
rarely starts.  Otherwise the plug was in good condition, the gap was
correct, and there was no gap bridging or heavy deposits.  Plugs are Bosch
platinum 4477 with about 6000 miles on them.

While working on the car I noticed fluid underneath the tranny on the garage
floor.  Confirmed it was transmission fluid.  Not a lot but a noticeable
amount.  This car has never leaked a single fluid and the driveway and
garage have until now been pristine.  I wonder if repeated starting attempts
can cause transmission fluid to leak?

So I went and checked the other fluids...hydraulic fluid has dropped from
max to middle of fill line in less than a month despite never going down
before.  Again I wonder if the repeated starting attempts, or the time we
had to push it and steer without power assist can cause a leak to begin?

One more point to add: almost the last day that I drove the car trouble-free
also happened to be the coldest weather the car's ever been exposed to.  It
was parked outside for about 6 hours in -30C (-22F) temperature.  After 6
hours it turned over slower than usual but started on 2nd try (usually
starts on 1st) and took me home fine.  Its been in a warm garage since.
Maybe one or two other short successful drives, but something just didn't
'feel' right with how it started even then (given it was in a warm garage).

Before I had any trouble, there was an interesting 'tick tick tick' during
idle (actually a couple of them).  One could be recognizable as the tappets
and that would sometimes go away.  But the other one was the fuel injection.
I guess I'm gonna have to read up on fuel injection because I'm convinced
that the problem is that fuel isn't getting through the injectors into the
cylinders.  I'm tempted to try adjusting the mixture or pulling an injector
and if that doesn't work then getting it to a mechanic, or just shooting it
:-)

You asked if I had Bentley.  No unfortunately.  I have Hayes, which doesn't
cover adjusting the mixture of the fuel injection.

Marc

ps below when I said "Thus the current flow would be about 100mA once the
ignition was engaged,"  this was poorly explained.  A clearer explanation:

Once the ignition was engaged, the current flow was a steady 100mA.  While
cranking it remained at 100mA.  Once the key was released (released because
it seemed like the engine was starting) the current essentially followed the
rpm...ie.descended from 100mA to zero just as the rpm descended from 1000 to
zero. Time between releasing key and zero rpm was almost always under 1.5
seconds.



----- Original Message ----- 
From: "SJ" <syljay at optonline.net>
To: "Marc Boucher" <mboucher70 at hotmail.com>
Sent: Wednesday, January 28, 2004 5:03 PM
Subject: Re: SJ Audi 100 Won't Start (Vacuum Leak?)


>
> Here are the results of today's testing (pulls out his log book :-)
>
> TESTING OF THERMOTIME SWITCH:
> Both pins measured about 3.5k Ohms from ground to the pin.  At about 50F
> this is quite close to your results.
>
> TESTING OF IDLE and FULL THROTTLE SWITCHES:
> Confirmed...on idle the switch is engaged (zero ohms) between pin 1 and 2.
> On full throttle we get zero ohms between pin 2 and pin 3.  (I know its
not
> technically zero ohms...guess that's called 'closed circuit' ?)
>
> TESTING TO ELIMINATE FUEL PUMP RELAY AS CAUSE:
> This had been done yesterday and results reported.  Key point is that on
> removal of the relay and shunting, the fuel pump can be heard to operate,
> and is measured to draw about six Amps.  Starting of car while using the
> shunt produces a result that is NO different than using relay.
Conclusion:
> Relay is likely not the problem.
>
> TESTING OF FUEL PUMP OUTPUT:
> The fuel pump was disconnected at the entry to the filter (filter is
new-was
> changed as a first step in trying to solve this problem) and carefully
> directed into a container in which I could measure flow.  Fuel pump was
> energized for about 2-3 seconds...during this time it pumped about 60ml (2
> fluid ounces).  Note that I don't have the equipment to accurately and
> safely test flow, or to test it under pressure.  But this test gives me
some
> confidence that its pumping and more than likely the pump is not the
> problem.
>
> TESTING OF CURRENT FLOW THROUGH DIFFERENTIAL PRESSURE REGULATOR:
> Today I was not successful in getting the car started.  Thus the current
> flow would be about 100mA once the ignition was engaged, remain at that
> during cranking, and once the key was released (released because it seemed
> like it was starting) the current essentially followed the rpm...i.e.
> descended from 100mA to zero just as the rpm descended from 1000 to zero.
> Time between releasing key and zero rpm was almost always under 1.5
seconds.
***** "Thus"??  "would be"??
Did you measure the current or not?
Current should be at 100ma with igniton "on", but engine not cranking. Is
it?


>
> TESTING OF POTENTIOMETER DURING STARTUP:
> Previously I'd tested the potentiometer resistance with the bonnet open.
> I'd found that The resistance between the bottom and the middle contact is
> about 2 kOhms at rest and rises to a max of about 13kOhms near full open
> falling slightly to 11kOhms at full open.  Today I decided to measure this
> resistance during an attempted start just to ensure that the plunger is
> actually moving.  The resistance moved from about 2 kOhms up to about 4 or
5
> kOhms. when it appeared that the engine was about to start and then as it
> sputtered the resistance fell back to 2kOhms.
>
> TESTING THE EFFECT OF UNPLUGGING OXYGEN SENSOR:
> O2 sensor disconnected (heater and sensor plugs).  Result...no change in
any
> observed behavior.
>
> Two of your suggestions...the inline spark plug tester and the starter
fluid
> will have to wait until I can get a ride to a car parts store.  The
spark's
> been verified but you're right, it could be intermittent.  I'll try to get
> both as soon as practical...note however, with regard to the starter
fluid,
> its really too cold to barbecue up here right now :-)
***** Ok, we need to check these two items in order to eliminate spark as
the problem and fix the fuel system as the problem.

After you do the above > spark plug tester, starter fluid:
Verify again that you have no vacuum leaks. You said two lines had leaks . .
pinch them off to make sure the intake has no leaks. Since this is the #1
problem with hard starting, really make sure this is not the problem.

Redo the ground connections at the back end of the intake manifold. Take it
apart, sand the lugs, reinstall.

Do you have the Bentley manual?

Does the Cold start valve fire? You said it tested ok when you did the
diagnostic tests.
If its not too much trouble, take out the cold start valve, and see if it
squirts when you crank the engine. If the engine fires up with starter
fluid, this will be the area to check next.

Take out the plugs also and check them. See if they are fouled or wet. If
they are wet(fuel) I would get a new set. Clean and dry out the old ones in
the oven.


>
> Marc
>
>
>
> ----- Original Message -----
> From: "SJ" <syljay at optonline.net>
> To: "Marc Boucher" <mboucher70 at hotmail.com>
> Sent: Wednesday, January 28, 2004 11:38 AM
> Subject: Re: SJ Audi 100 Won't Start (Vacuum Leak?)
>
>
> Theoretically, the car should run without the fuel injection ECU. The
system
> will run in a mechanical mode.
> Basically, all the ECU does is vary the Diff Press Reg current to fine
tune
> it for emissions.
> The other ECU outputs are minor, ISV, vapor canister, cold start valve,
etc.
> Many of the inputs to the ECU can be "faked" . . ie, resistor in place of
> temp sensor, jumper for idle switch, etc
>
> Are you keeping a log of your tests?
>
> Things that you arent doing and should be doing.
> 1. Get that $7 in line spark plug tester . .and leave it on. Just in case
> you have an intermittant ignition problem.
>
> 2. Hook up the ammeter for the Diff Press Reg. I use a cheap analog meter

>  radio shack. You can calibrate the analog against the DMM. Always monitor
> this current . . . you should have a list of current values and
conditions.
> I sent you an email with these values.
> Both my cars have insulated male and female spade lugs crimped on one of
the
> regulator wires. And I made up lug type test leads to interconnect to the
> wiring and my meter. Now its very simple to hook up a meter.
>
> 3. Get starter fluid and rig up rubber hose so you can inject the fluid.
Its
> a lot easier to troubleshoot if you can get the engine to run. The starter
> fluid will bypass cold start valve and lean mix from pressure regulator.