1996 DeVille Serial Data Failure

[TedK]

My 1996 DeVille is effectively dead due to the failure of circuit 1807 (serial data).  However, the failure is intermittent - it may disappear for an hour or so (meaning that the car starts fine), then will reappear (causing a no-start condition) - all without me touching anything.  For this reason, this seems to be a failure of one of the modules on the serial data line.  This doesn't seem to be a short to ground condition, as it doesn't ohm out to 0 resistance.  Rather, my dvm shows variable resistance / variable voltage.  I don't have a lab scope (hanging head in shame).

When the circuit is bad, my OBD II scanner won't work - it can't read the serial data (power & ground are present on the DLC - the tool just can't make sense of the garbage). However, I can pull the codes from the message console.  Then, out of the blue, the problem disappears and the car starts fine, with no current trouble codes!  However, after a few minutes, the problem returns and the car again won't start.

Here are the codes, pulled from the message console, when the car won't start:

• NO PCM DATA
• IP2750 C (PassKey II Data Comm Failure)
• AC2255 H
• AC2016 C
• No SDM Codes
• TC0072 C (IPC Serial Data Line Malfunction)
• TC0073 C (PCM / TCM - TCS Interface Malfunction)
• TC0077 C (PCM Serial Data Line Malfunction)
• No RSS Codes
• PZ2255 C (General Loss of Serial Data)
• IR2129 C
• No RFA Codes

I tried a different PCM, but the problem didn't resolve (so I put the original PCM back in).

My service manual says that circuit 1807 (serial data line) uses a Splice Pack (SP205, I think).  However, I cannot find this splice pack.  It's supposedly under the driver's side dash board, but I cannot find it!  I've removed the parking brake mechanism, pulled back the carpeting, removed lower dash components, and searched everywhere up under the dash, but can't find the splice pack.  This is an early-production 1996, so perhaps GM didn't actually use a splice pack until later-year productions?  The purple wire (1807) from the DLC enters a small wire harness, then seems to merge into a larger wire harness, and I lose track of it from there (but it "seems" to pass through the firewall - not sure if that purple wire is really 1807).

Please - I'm open to suggestions!  My next effort is to remove the rear seat back (again), and start unplugging controller modules one at a time, hoping for the problem to disappear with a module (I did this a year ago, without success; yes, the car has been down for a year).

 

[OldCadTech]

@TedK

 

This is one of the hardest codes to diagnose. We nicknamed it the 800 line cuz ALL the modules use it.

As you probably know it can be caused by ANY module on the 800 line or any combination of modules. You have started down the right path, as checking grounds is the first step to perform.

The parts department at the dealership I worked at would let us plug and play modules as this was the only way to determine the defective module. The codes you have are of no help as the rule of starting with the lowest code first does not apply with this 800 line malady...

The ground I would start with is behind the drivers kick panel. Look in the FSM at the grounds section and the locations section to determine locations. 

Personally, I would start with the pass key/security module as most of the others (except the PCM) would allow the vehicle to start but they would set a hard code and turn the light on.

[Logan]

Circuit 800 codes. Very difficult code to catch....and yours is intermittent too..

On my '93 it was the IPC cluster. 

I would go and inspect the ABS unit first. They are known for rotting away. Pretty sure it is on the network......maybe it rotted in a way it is messing up the network. 

 

[TedK]

7 hours ago, Logan said:

Circuit 800 codes. Very difficult code to catch....and yours is intermittent too..

On my '93 it was the IPC cluster. 

I would go and inspect the ABS unit first. They are known for rotting away. Pretty sure it is on the network......maybe it rotted in a way it is messing up the network. 

 

Thank you both for the quick posts!  When the car first failed to start (in Phoenix, on August 18, nowhere near shade, a day that broke the temp record), I eventually cleaned the resistor pellet contacts in the ignition cylinder, and (perhaps coincidentally) the car started.  The next time it failed, the failure was preceded by something like "traction control failure" on the message console; when I turned the car off, I was rewarded with this general system failure.  I inspected the ABS unit, but the coincidence is strong, so I'll revisit that module.  Just for good measure, I'll also install a new ignition lock cylinder.  And, I'll re-check all of the module grounds.

Whatever is causing this problem, it's causing a general failure of the entire 1807 network. One of these days I'll pick up a good lab scope.

Again, thanks!  I'm also open to other suggestions / ideas / recommendations!

[Logan]

On older models......the ABS computer was in the trunk. I think in '95 it was incorporated into the pump assembly under the hood.  

The underhood ones can get damaged by moisture etc. The electrical plug can get eaten away. 

All the computers are wired on a loop. So each computer has 2 network connections wires. So if a single wire got damaged the system would still work.  

If any of the computers went bad.....it can pull down the whole network. 

Don't think you need a lap scope....just a DVOM. On OBDI cars you could measure the network voltage at the ALDL......was suppose to vary like 2-5 volts all the time.....if you saw the voltage lock...the fault was happening....and within 30 seconds or so you would start getting the network errors and codes. 

To catch....you unplug each computer one at a time while watching the network voltage. When the bad one is unplugged and the voltage resumes varying again....you found the bad one. 

Of course.....the fault has to be present and happening to detect. For a while mine was a intermittent...once every 200 miles kind of issue.  

 

 

 

[rockfangd]

I agree completely on a module being bad.

ABS unit is located forward of transmission on the subframe.

Was there any issues before this started that may not even seem related? It could be a lead to the issue that is current