By Richard McCuistian

In the early world of automotive ‘black box’ technology, parallel interface was the only show in town to begin with.  What that means is that every sensor and actuator had to be hard-wired to the brain box, and quite interestingly, (other than dashboard-mounted electronic gadgets that didn’t do much real work) the automotive computer invaded the engine compartment first, where automotive engineers had to hammer out strategies and algorithms to handle the single most complicated job computers do on a vehicle; ignition timing and fuel control.  Anti-lock brakes became mandatory on vans and pickups in 1987, about the same time computers spread their electronic tentacles into the hot and slippery innards of our automatic transmissions and our suspension systems. 

While early automotive electronics worked fairly smoothly, OJT diagnosis was about as good as it got (factory schools and pinpoint tests offered limited success but could be confusing and ineffective in many cases) and could be a hair-pulling experience, even for the best technician.   The best resource an automotive electronics guy could have while chasing GM Computer Command Control, Ford Electronic Engine Control, or Chrysler Single Module Engine Controller flashout codes was first-hand knowledge of what the volts and ohms looked like at various test points on a normal system (it obviously varies with make, model, and year). In my early years as a dealership technician, I personally kept my own pocket notebook full of such figures.  Hither came GM’s scan tool-friendly system, which earned a great following in those early years.

Another example would be the way some networked traction control-equipped vehicles use the Anti-lock Brake module to prevent wheel spin by pulsing the drive wheel brakes while at the same time using the PCM to retard ignition timing, effectively reducing engine torque.  These two modules are both utilizing the same set of inputs to implement the desired strategy, namely, vehicle speed, wheel rotation speed, and engine load calculations.  The communication goes both ways in this circumstance, and the modules work together to achieve the desired result.

GM and Chrysler got smart in the early eighties and outfitted their engine control computers with special software and low speed, inexpensive UART (Universal Asynchronous Receiver/Transmitter) networks that gave GM and Chrysler techs a brand new window into what the ECM was seeing and doing. Chrysler’s UART protocol was dubbed “Serial Communication Interface” (SCI).  Ford lagged behind and kept Blue Oval techs fiddling with code-snatching diagnosis until 1988, when Dearborn’s engineers offered a Standard Corporate Protocol (SCP) window, delivered to the New Generation Star (NGS) tester through a twisted pair of wires, and while most Blue Oval vehicles were scan tool friendly by 1991, V8 Mustangs didn’t get scan tool communications until the mid-‘90’s.

In the meantime, GM and Chrysler had forged ahead with their multiplexed systems, allowing different modules to share information.  Two of the most commonly shared sensor inputs over a multiplex network are the Vehicle Speed Signal and Engine Coolant Temperature.

While all three automakers started out with simple Class A (10 Kbps) networks, their datastreams were proprietary and each system utilized a dedicated scan tool with a dedicated connector configuration.  With the onset of OBD II regulations, the data link connector and emissions-related DTC’s became standard on all vehicles sold in North America between 1994 and 1998.  And while a breakout box and a digital volt/ohmmeter are still as necessary as they’ve ever been, gathering (and recording) data with a powerful scan tool via a network connection is rapidly becoming the order of the day, although the best and most comprehensive software is still found on manufacturer-specific scan tools.  Nothing quite handles GM stuff like the Tech II, and the DRB III is the best choice for anything Daimler/Chrysler.  Ford vehicles still talk to the now-ancient Hickok New Generation Star tester through a redesigned interface module, but the Worldwide Diagnostic System is still the best choice for diagnostics on any new Ford.  

With OBDII regulations requiring reprogrammable PCM’s, serious independent shops are discovering that the capability to access manufacturer resources and re-flash engine and body controllers is becoming extremely important.  One problem is that the hardware to do such re-flashing is generally manufacturer specific and can add a thousand dollars or more to the price of an already expensive scan tool, not to mention what it costs to stay on the manufacturer’s software subscription list for re-flash programming information.  Many quirky problems (even erroneous instrument panel gauge readings) found on today’s vehicles are repaired by re-flashing the brain-box in question.

As some of the higher end cars have continued to add accessories and vehicle wiring became ever more complicated, engineers began to rely more heavily on multiplexing (multiplexing is a method of sending 2 or more signals simultaneously over a single circuit) in order to reduce the hundreds of pounds of wire harness required on those cars to make everything work.  There are a variety of networks (see above).

There is a fine line, from a technician’s point of view as to how much information is too much information (TMA) when we’re studying something like this.  While there are those among us who seek an in depth understanding of every part of what we do, there are others who simply want to know what they need to in order to fix the car.  The facts and figures in the table above might bore some of the more “cut to the chase” guys to tears.

Let’s have a nuts and bolts look at what’s going on behind that OBDII connector we’re so accustomed to seeing under the dash.

 Federal OBD-II regulations specify that stored fault codes be available through a diagnostic port via a standard protocol. Currently, OBD-II specifies J1850 (pins 2 and 10 in the J1962 DCL connector) and the European standard, ISO 9141-2, which is used by Chrysler in pin 7.

Space wouldn’t permit an exhaustive set of tables like the one above, so in the interest of space, the table above simply compares a 2006 Thunderbird to a basic Daimler/Chrysler pinout.  Remember that on every connector, pin 16 is B+ and pin 5 is ground.   The knowledge that pin 7 is dedicated to the ISO 9141 network, and on some vehicles (Chrysler particularly), the OBDII generic information is delivered here via the “ISO K” Serial Communication Interface (SCI) line, while the OEM enhanced data comes out on pins 2 and 10 via the J1850 network. 

On Fords, generic data comes out on the same line as enhanced data, but remember, there are two separate ‘rooms’ in OBDII compliant PCM’s.  While you may or may not have seen this scenario, just be aware that it is possible to have an illuminated Check Engine light and no DTC’s stored in the OEM Enhanced part of the PCM, yet still have a misfire code in the OBDII ‘room’ that must be cleared before the Check Engine light will go away.  The OEM Enhanced ‘room’ generally has more comprehensive data, but codes are defined and stored in each ‘room’ using different criteria.  It’s wise to peek into both ‘rooms’ on every vehicle if you don’t want to miss anything that might be important.

 By now, most knowledgeable scan tool junkies will know that P-codes are powertrain related, B-codes are body, C-codes are chassis, and U-codes are network related concerns.  It’s also important to understand that a no-communication issue isn’t always network related.  For example, a no-start/non-communicative PCM can be the result of a shorted Reference Voltage (VREF) wire.

So what do we look for at the DLC connector if we have a non-com issue?

 Gather Good Data

 As is my mantra, a good technician will spend a few moments whenever he or she can in gathering pertinent data from vehicles that are working right.  With a mental library of dependable readings, any tech is better equipped to diagnose a concern.  Here are some examples of what a tech might see with a voltmeter at certain circuits found in the DCL.  Scope patterns taken here are interesting too.

Some networks, even when they’re working right, will have no voltage unless communication is under way.  On Ford’s two-wire SCP bus, both wires read 5 volts until a scan tool is connected. At that point, the voltage averages about 2.5 volts on each line, and an automotive o-scope will show what appears to be a line of W’s.  On some networks, the bus voltage is zero unless a scan tool is connected, and some scan tools actually use voltage from DLC pin 16, filter it, and use it to provide ‘bias’ voltage to the bus.  More about that in a minute.  Chrysler’s PCI bus has about 7.5 volts (1/2 charging system voltage) when the bus is transmitting data, and 0 volts when the bus is silent.

    On Chrysler’s two-wire SCI bus, DLC pin 12 (SCI receive) should show 5 volts with no scan tool connected, and DLC pin 7 (SCI Transmit) should so zero volts.  SCI transmit voltage should jump to 5 volts when scan tool communication begins.

All networks have a ‘bias’ voltage that originates at certain modules (Chrysler likes to call these ‘dominant nodes’’ and the receiving modules are referred to as ‘recessive nodes.’).  The network is set up not to broadcast recessive node transmissions when a dominant node transmission is under way. That’s what “collision detection” is all about as related to networking.  In a word, when multiple modules transmit messages at the same time, the possibility exists that messages could collide with one another, as peculiar as that may sound. Therefore, messages generated simultaneously are sent in an order based on priority.  Low priority messages go to the back of the line and wait in a buffer until the information highway is clear, but that clearing usually doesn’t take but just a few microseconds. 

   CAN bus networks operate similar to the previously described networks, but communication takes place a lot faster, thus more information can be handled in a much shorter period of time.  It’s sort of like the difference between dial-up internet and a DSL connection.  At least one of the secrets behind the CAN bus’ ability to communicate faster is the fact that CAN network modules have a dedicated processor for communication jobs. This arrangement allows the main module processor in each black box to handle the tasks it is designed for unhindered. On D/C products, however, the high speed (500 kbps) CAN C bus is only functional with the key on.  The slower CAN B network may or may not be active with the key off, depending on what the modules on that network may require. These configurations may vary from vehicle to vehicle.  Some of the benefits of CAN networks over previous protocol architectures are:  

the shelf) customer features, a large number of configurations serviced with a fewer number of parts, and cost savings from quantity production of common parts. Last but not least, there are a fewer numbers of parts necessary in production plants

           When repairing CAN bus wiring, make sure you keep it away from wires that carry high current. 

Chrysler CAN-B Controller Area Network Data Bus - Lower Speed, used for less essential functions such as those listed below.

• The two CAN bus wires are generally referred to as CAN H (+) and CAN L (-).
• Ford calls their two CAN busses “medium speed’ and ‘high speed,’ but each bus has plus and minus wires.

Some CAN busses have a central controller that separates the busses connected to it so they can operate at different speeds.  This module also relays data from one module to another or to the scan tool.

If you have multiple in-car systems that don’t operate properly or at all, the CAN B bus might be at fault.

An exhaustive study could fill a lot more pages.

One way or another, networks are getting better, faster, and smarter, and that’s a good thing.            R.W.M.