Mystery Solved: was No Start: Audi 100, 1990, I5 NF engine, non-turbo

Fri Sep 9 21:43:29 PDT 2011

After checking and rechecking every connection to the ECU (comprised of a 25-pin Ignition Control Unit (ICU) and the 35-pin Fuel Injection Control Unit (FICU) ), and everything else possible, I decided it was time to order replacement units.  I decided to completely pull the ICU unit from the car to be certain I had the correct part number, and to get a good look at it. 

When I got it out I noticed something: Someone had previously opened the sealed ICU !!!  First, there's a screw with a drop of red paint on one screw and black plastic to show if its been opened.  Second, all of the screws looked like they'd been pulled by someone who didn't quite have the right set of hex-head (allen key) screwdrivers.  

When I got the circuit board out of the plastic case, there was one spot on the backplane that looked 'different'.  This was a long thin strip on the circuit board that connected a number of pins directly together.  What I'll surmise is that this strip on the circuit board had previously failed.   Someone had repaired the strip by flowing a thin strip of solder.  Since I've had the car for 11 years, this repair was done at least 11 years ago.  When I used an ohmmeter probe to touch what I think was the solder repair to measure connectivity, the repair simply crumbled away into ashes.  I believe that this is a thin area of the circuit board that carries the bulk of the current that the ICU handles, and hence the soldier repair eventually failed also.

I repaired it by soldiering a copper wire into place, ensuring that I recreated the original circuit.   (Those familiar with the type 44 might also have had to do a similar repair on their instrument panel to solve the common 'bobbing speedometer' problem).

Put the unit back in place, and the car started and ran like a charm, WITHOUT the 'ground-to-hall sender' repair that the garage had done.

If I end it there, we have a pretty good Hollywood ending.  But, alas, there's more.  I decided to run the "Output" tests for the ECU...the ones that test that it correctly cycles the ISV, cold start valve, carbon canister, and differential pressure regulator.  Only the carbon canister correctly cycled.  I rechecked all the other components for these tests that I could determine (WOT, idle throttle, O2sensor heater ground) and all appeared correct.  So it looks like I'll still need to order new ECU units.

The FICU had also been taken apart previously.  However just by visual inspection and some basic tests with an ohmmeter, I couldn't determine any problems with the FICU, or any additional problems with the ICU other than the one damaged path on the circuit board of the ICU, which I repaired.  No signs of water, corrosion, cracks, or other broken solder joints on either board. 

Anyways, a big thanks to Huw and Tony for their help in diagnosing this problem, and to Peter for suggesting that I take the ECU apart and look at the insides.

MC

--------------------------------------------------
From: "Peter Orban" <orbanp1 at yahoo.com>
Sent: Friday, September 02, 2011 10:14 AM
To: <quattro at audifans.com>
Subject: ECU repair, was No Start: Audi 100, 1990, I5 NF engine, non-turbo

> Hi Marc,
> 
> Check that big main connector to the ECU. Are all the female pins tight? If they are not that could cause that 100 Ohm resistance.
> If the connector checks out OK, take apart the ECU and see how are things inside. Check if there was any water intrusion into the box. There would be a thin O-ring around the perimeter the make the watertight seal. 
> Track down that path that read 100 some Ohms instead of zero. Water intrusion could ruin the connection between the ground point and where you measured that. If there was water inside, there could be some corrosion of the traces, component leads. Follow through that trace visually. Most probably that ECU is only on a two-layer printed circuit. If there is some corrosion from water, chances are that it could be repaired (I did manage to repair corrosion damage in a TV that a tech had given up on it.)
> 
> See if there are any broken solder joints there, they would appear gray as opposed to shiny. Reflow them (with a soldering iron) if they look suspicious. Use a magnifier to check this. some of the ICs (memory, processor) would use sockets, that is also a source of unreliability. Remove and re-seat the integrated circuits, that would "clean" the connections.
> 
> Good luck, Peter
> 
> --- On Thu, 9/1/11, quattro-request at audifans.com <quattro-request at audifans.com> wrote:
> ------------------------------
> 
> Message: 3
> Date: Thu, 1 Sep 2011 23:18:03 -0400
> From: Marc Boucher <mboucher70 at hotmail.com>
> To: <audi at humanspeakers.com>,    "Tony Hoffman" <auditony at gmail.com>
> Cc: quattro at audifans.com
> Subject: Re: No Start: Audi 100, 1990, I5 NF engine, non-turbo,
>     non-quattro,    CIS-E-111 (California)
> Message-ID: <SNT106-DS101EE6EC1DC80E66B2946C1180 at phx.gbl>
> Content-Type: text/plain; format=flowed; charset="Windows-1252";
>     reply-type=response
> 
> At this point, I've located a replacement ICU for a reasonable price.  I 
> also believe that I've exhausted all reasonable tests that would point to it 
> being something other than the ECU (ICU/Fuel-Injection CU combination).  So 
> I'll probably order it.  Just wanted to first detail the additional tests 
> that I've conducted, and information that I've gathered, and see if there 
> were any final tests anyone has to recommend.
> 
> All three wires from hall sender to ICU were measured, once again, for 
> resistance.  This time on the Rx1 scale.  There is no measureable resistance 
> in any of these wires between the plug in the engine bay that connects to 
> the hall sender and the plug that connects to the ICU.
> 
> Previously, I'd measured the voltage between the red/black wire and the 
> brown/white wire of the hall sender, and got the expected 12 volts with the 
> ignition on.  But then it occurred to to check the voltage between the 
> red/black wire and the the car's ground...this is also 12 volts with the 
> ignition on.  So then whats's the resistance between the brown wire and the 
> car's ground with the ignition on?  Its infinite...open circuit.  How can 
> red be 12 volts to both the car's ground and 12 volts to the brown wire, but 
> the brown wire has no connection to the car's ground?
> 
> Next I put the multiimeter between the car's ground and the hall sender's 
> brown wire and cranked the engine.  It read about 5 volts.  Then I switched 
> to a current scale which would allow current to pass and hence the car to 
> start.  Car started, but I couldn't read the current, likely because its 
> switched, not DC.
> 
> Just to be complete, here's a list of additional actions that I've taken:
> 
> Remove the ground wires attached to the intake manifold and test for 
> continuity with ground at the ICU.  Even though it was good, I nonetheless 
> cleaned the ring connectors and reattached them with a bit of dielectric 
> grease.
> 
> Checked the continuity of every wire between the ICU and the fuel injection 
> control unit !!!
> 
> Previously all grounds were checked successfully, but you'd raised the 
> possibility of a wire showing a good connection using an ohmmeter but not 
> being able to carry a sufficient current.  To eliminate this possibility I 
> ran a second, parallel ground to each of the ICU and Fuel Injection unit's 
> ground terminals.
> 
> Any last tests that anyone care to recommend before I get the replacement 
> ICU?
> 
> If I do order the replacement ICU, any other parts that I should probably 
> pick up while I'm at it?  Not specifically for this problem, but since I 
> plan on keeping this car a few more years, and since shipping parts 1 by 1 
> gets pricey, I may as well stock up on a few things if I place the ICU order 
> from Force 5, that needs to ship it.
> 
> Thanks again
> 
> MC
> 
> End of quattro Digest, Vol 95, Issue 1
> **************************************
> _______________________________________________
> quattro mailing list
> http://www.audifans.com/mailman/listinfo/quattro
> http://www.audifans.com/kb/List_information
>