Circuit Fault Causes Premature Component Failure

Randy Peterson - Diagnostician

Diagnostician - Randy Peterson Randy has worked for Certified Transmission for over twenty four years and is an ASE Certified Master Technician, including L-1. He has been in the automotive industry for over 30 years.

This story begins with a 4WD 1998 Honda Passport equipped with a 3.2 V6 engine and 4L30E transmission. The shift indicator in the cluster was not functioning correctly and transmission was shifting hard. The shift indicator showed 'P' in Park, and '1' in D3, D2, or D1. The indicator didn't light up in any other gear. It seemed like a straightforward diagnosis since the mode switch (range sensor) has a high failure rate on these vehicles. However, a new OEM mode switch from Honda had been previously installed for this issue, so it was time to start the real work of accurately diagnosing the cause. Little did I know that this project would become, "The Nightmare on Oak Street."

There was no data stream available to view this vehicle using our trusty MT2500 Snap-On scanner, so some pinpoint testing was the next order of business. To start, I printed out the wire diagrams and familiarized myself with the component and connector locations. This system is a little more complex than others. The mode switch sends information to the PCM as well as the Alarm and Relay Control Unit (aka Multiplex Control Unit). The PCM uses this information to regulate the transmission operation, the ARCU sends the information to the IP cluster. (Figure 1)

Article Figure 1

Figure 1

First things first. I cleaned the science experiment at the battery terminals. (Figure 2)

Article Figure 2

Figure 2

I then went on to check the power and ground circuits. How many times have I skipped these steps, only to get bitten by it later? The PCM was below the center of the dash, under the console. (Figure 3)

Article Figure 3

Figure 3

The Alarm and Relay Control Unit was above the right kick panel (it was hidden). (Figure 4)

Article Figure 4

Figure 4

I gained access to both modules for circuit checks. Power is needed at both the Mode Switch and the ARCU. (Figure 5)

Article Figure 5

Figure 5

Now that I had access to testing, it was time to break out the DVOM. The backup lamps and turn signals worked, which meant the fuse (shared by the Mode Switch and ARCU) was good. I checked for B+ at the Mode Switch. There was B+ to the Mode Switch, and B+ and ignition B+ to the ARCU. I checked the ground circuit from the battery to the PCM and the ARCU. At this point, all that I tested was good. (Figure 6) (Figure7)

Article Figure 6

Figure 6

Article Figure 7

Figure 7

With the voltage supply and grounds testing good, it was time to see if I was getting signals from the Mode Switch to the PCM and ARCU. The Mode Switch has four internal switches. Power is supplied to the switch on pin 8, and as the lever is moved it directs power through the internal switches to their assigned circuits, as follows:

  • Pin 8 (Lt Blue): B+ from the back up/turn signal fuse.
  • Pin 2 (white): effective in P,N,3,L
  • Pin 5 (purple): effective in P,R,3,2
  • Pin 6 (gray): effective in R,N,D,3
  • Pin 7 (org/blk): effective in D,3, 2,L

When testing each pin, there should be continuity in its respective ranges. I also tested the voltage from the Mode Switch at the PCM and ARCU with them disconnected. Pins 5, 7, and 2 tested okay. Pin 6 tested open in all ranges. Bingo! It appeared that I had found the problem. Could there be two Mode Switches with the same problem? Remember, this vehicle arrived to us with a new, OEM Mode Switch installed. Could this new switch be bad?

We purchased a new mode switch and installed it on the vehicle. Now the indicator was functioning correctly, and the shifts were normal. We always perform a final road test before the vehicle is delivered to the owner. For the first seven miles, all was well, then as I turned down the home stretch, the indicator went blank, and the transmission wouldn't shift. What happened? Everything was working so well up to that point.

I limped it back to the shop. I ran the lever through the ranges. I witnessed the same exact problem as before. What did I miss? I ran through the tests again. Again, open on pin 6. What's going on? I took the mode switch apart and there it was: a piece of the thin metal used on the circuit board to make the slide connection was missing. The only thing that I could think was, there had to be too many amps on that circuit that were burning up the circuit on the board.

I used a fused jumper (5-amp fuse) with an ammeter in-line to test the load on pins 5, 6, and 7. Pins 5 and 7 had .012-amp draw. But pin 6 blew my 5-amp in-line fuse. I disconnected the ARCU and replaced the fuse. Blew the fuse again. I disconnected the PCM, and it blew that fuse. It seemed to indicate that I had a short to ground somewhere.

I started the hunt at the Mode Switch. The wiring harness passes by the lower left side of the bell housing with a shield to protect the harness. I removed the shield and there it was: a shiny piece of copper staring at me. The shield had rubbed through the conduit, tape, and finally the insulation causing the short. (Figure 8>

Article Figure 8

Figure 8

I repaired the wire, taped the harness, and installed a new piece of conduit. At the same time, I bent the shield so there were no sharp edges to wear through again. I retested the circuit and installed a new Mode Switch, and this time it passed the 15-mile final road test.

Moving the harness around during the Mode Switch replacement must have been enough to move the bare wire away from the shield for a time. Then after driving, the vibrations moved it back to its original resting place and shorted to the ground, repeating the fault.

When I took the Mode Switch off and took it apart, I can only conclude that the printed circuit was acting as a fuse, but the only problem was it failed with a lot less amperage than what the fuse that powers the circuit is rated at. In the end, the root cause was found, and the issue was corrected. The customer was pleased and that's always what matters the most.