Hello, Allpar Forums member or visitor! If you were a member, you would not see this ad!

Register or log in at the top right of the page...

  1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Fighting intermittent dash gauges

Discussion in 'Minivans · Pacifica' started by John Wood, Oct 14, 2017.

  1. John Wood

    John Wood Well-Known Member

    Joined:
    Nov 13, 2002
    Messages:
    11,078
    Likes:
    153
    I've been trying to sell the 96 Grand Voyager 3.0, 3 speed and have it listed on Craigs list. I had a real interested buyer until the gauges started going flaky. The speedometer, odometer, temp gauge and fuel gauge just decide to quit every once in a while.. So far, I have pulled the console out, disassembled, and re soldered the usual suspect joints with no success. Actually all the solder joints looked good. Banging on the dash and moving/probing wires has no effect whatsoever.

    Interestingly, with the key out, I get some fluttering of all the needle pointers which is weird since it should not even be powered up.

    I believe the vehicle CCD bus is working since I can connect my scanner to the DLC and read temperature and speed when the vehicle is moving and those readings look correct. I am currently trying to confirm if mileage is being added to the odometer when driving with the gauges and display being non-functional. I believe mileage is being added.

    When the console does work properly, I run the diagnostic test and get no error codes; just the 999. The needles go thru the self-test and position themselves in the correct position during the test. I just don't think the cluster unit is bad.

    Anyone have any ideas? The rest of the vehicle performs nearly perfect.
     
  2. ImperialCrown

    Level III Supporter

    Joined:
    Jun 8, 2008
    Messages:
    18,231
    Likes:
    2,660
    In many but not all cases, the problem is fractured solder joints. The larger pins and larger, heavier component leads are more susceptible to cracking than the lighter, smaller-leaded components.
    Connector terminal corrosion can also be 'invisible'. I have used a pick to 'dimple' or raise a terminal contact tongue that has lost its springiness. You want to 'scratch' a good connection into the terminals. Sometimes Deoxit or a shot of electrical contact cleaner may help. I haven't used dielectric grease as this is assumed to be a fairly dry area. Make sure that the cluster is fully seated on the connectors.
    The 13-pin and 8-pin connectors and large resistor were common touch-up points. Thumping the dash usually will re-establish a fractured connection, but only until the next occurrence.
    It could be a bad board? The cluster board may be intermittent and not exhibit a problem while under test. The instrument cluster is called the MIC in the service manual. MIC = Mechanical Instrument Cluster.
    The odometer window will only display PCM fault codes. There may be 'no communication' fault codes found in other modules like the BCM, ABS and AIRBAG modules. This has been my experience, but requires a more specialized scan tool. It may or may not light the AIRBAG light.
    Chapter 8E covers the instrument cluster. Follow the Diagnostics and Testing section as best you can. This is for a 1997, but should be similar to a 1996:
    http://oskin.ru/pub/chrysler-dodge/manuals/Service Manuals/1997_GS_Town&Country_Caravan_Voyager/EGS_8E.PDF
    As for a CCD bus going down, corrosion can cause a path to 'short' the low voltage/low current bus. I have seen vehicles in winter or heavy rain where the driver splashed (salty) water onto the DLC connector from his work boots and lost gauge action. I was able to rinse, dry and silicone spray the DLC to restore the bus. You may or may not see a 'green cheese' stain between the terminals.
     
    John Wood likes this.
  3. AllanC

    AllanC Well-Known Member

    Joined:
    Jan 6, 2003
    Messages:
    3,134
    Likes:
    400
    On the third generation minivans the odometer mileage is recorded and stored in the BCM. This information is sent through the network to the instrument cluster for display.

    There is a 13 pin connector which provides power and network communication to the instrument cluster. Check that connector (male and female components) for corrosion. Sometimes just one or more pins needs to be bent slightly to create better contact but do this VERY carefully as you do not want to break any solder joints.
     

    Attached Files:

  4. John Wood

    John Wood Well-Known Member

    Joined:
    Nov 13, 2002
    Messages:
    11,078
    Likes:
    153
    Thanks... I have tried most everything you mention. There is a cluster self diagnostics test that positions the needles into specific test points... lights all the lamps for a brief moment and displays cluster error codes. As mentioned, everything passes. Of course I can only do the test when it is working.

    After spending a couple of hours this morning doing voltage tests, I found that I have approximately .5 to .7 volts drop from the ignition switch output to fuse #2 in the junction block. Fuse 2 is one of the feeds to the cluster. 11.2 volts may be just at the point where the cluster won't power up. This voltage drop occurs when running or stopped. It could point to a dirty connection on the junction block (large grey connector P33) or perhaps one of the diodes which are in series with the power circuit, just before the fuse. Their could also be a bad solder joint in the junction block (diode or fuse holder socket).

    About an hour ago, everything started working again so I am shut down until it fails again which will likely be tomorrow morning if the pattern continues (morning dampness?). I still have close to .5 voltage drop when the cluster is working which just doesn't seem right to me. The data link connector battery voltage is solid 12.1.

    Alan,I have inspected the cluster connector, put it in and taken it out several times and sprayed it with contact cleaner. I touched up the solder joints, even though they looked fine. I've done a lot of PC board soldering in my life, so I am fairly confident in my work there. I even took the BCM out and inspected it. The joints look good but it is much tougher to flow solder on that board because of the water proofing shellac they use which insulates the heat of the solder iron tip, making it tough to get good flow.

    I'm willing to put my time into this old van, but I am very reluctant to buy any parts for it since I am only asking $800 for it and the tires have only about 6000 miles on them. Even at $800 for a good running van, the buyer expects everything to work. We'll see what tomorrow brings. :)
     
  5. AllanC

    AllanC Well-Known Member

    Joined:
    Jan 6, 2003
    Messages:
    3,134
    Likes:
    400
    I looked at a wiring diagram for the 3rd generation minivan and it shows a diode in the switched ignition circuit upstream of fuse #2 in the junction block. This seems to be an intended engineering function and from my experience one would expect about a 0.5 volt drop across a diode. Voltage in the vehicle electrical system can and does on occasion go below 10 volts (cold start in extreme cold weather) so I do not think your intermittent gauge problem is caused by a 11.2 voltage source to the instrument cluster. Cavity #2 of the instrument panel electrical connector has full, unswitched battery voltage applied to the cluster in addition to switched power at cavity #11.

    Try this test. Measure the voltage at pin connector #2 at the instrument panel connector. It should be battery voltage. Fashion a jumper wire to connect pin #2 to pin #11. You will bypass the switch circuit at the ignition switch and your concern about having only 11.2 volts. Drive the vehicle. If the intermittent cluster problem disappears then low voltage on that switch circuit to the instrument cluster is the problem.

    But I am thinking that will NOT be the solution. If the jumper wire between pin #2 and pin #11 does not solve the problem, try this test. Fashion an additional wire from a good body ground (steering column tube) to pin #7 and pin #13. These are grounds for the instrument cluster. Drive the vehicle. If the intermittent problem disappears, then the issue is attributable to a poor ground connection.
     
    John Wood likes this.
  6. John Wood

    John Wood Well-Known Member

    Joined:
    Nov 13, 2002
    Messages:
    11,078
    Likes:
    153
    A bad ground is a logical explanation; especially since failure seems to occur with damp over night air. We can't seem to break the pattern of 100% night time humidity and muggy air here in Florida, even though it is mid October. :) I will probably pull the cluster out tomorrow and try a direct ground wire from pin 7 on the wiring harness. I actually already did that on pin 13 and when it didn't help, I studied the schematics a bit more and realized I should have also ran a ground to pin 7. I think the pin 13 ground is for the lamps. The pin 7 ground goes to a specific ground connector buried deep behind the dash (P25). I have had no luck trying to locate that connector. Possibly pulling off the top dash panel might help, but that thing is very brittle and any flexing will likely crack it. I'll just go with your recommendation to run a separate ground to a good clean piece of steel around the steering column frame near the inside firewall.

    It is a PITA pulling the cluster on my vehicle because of the mechanical indicator used on the 3 speed. I have to pull off the steel knee blocker for access to unhook the steel cable loop from the shifter mechanism, then unclip the adjuster and feed it all back behind the cluster as I pull the cluster out. They don't give you much slack in the wiring harness-to-cluster connector either, so it is hard to make 'live' voltage tests. i.e unit powered up when measuring.

    I appreciate you input and will give an update tomorrow if the unit does indeed fail again.
     
  7. John Wood

    John Wood Well-Known Member

    Joined:
    Nov 13, 2002
    Messages:
    11,078
    Likes:
    153
    Just a quick note. We had an early evening rain shower... i.e. 100% humidity and muggy. I just went out and the cluster has begun going into failure mode. It is intermittently trying to come on and then failing. The odometer display begins to light and then goes dark at the same time the fuel gauge starts to rise and then falls back to zero. It just seems like this has to be a dirty/corroded connection somewhere.

    I'm almost wondering if the windshield is seeping at the lower seam and perhaps getting on that ground connector (P25) under the dash. There is no evidence of water anywhere inside that I can see, but if that is the case, I may be facing more serious issues.
     
  8. AllanC

    AllanC Well-Known Member

    Joined:
    Jan 6, 2003
    Messages:
    3,134
    Likes:
    400
    Ground connector P25 is also referenced as G200 in the diagrams. It has a connector (not easily accessible) behind the dash / instrument panel.

    You are correct in that the schematic shows ground G300 as the ground path for the instrument panel lamps. It is grounded in the vicinity of the left kick panel area.
     
  9. John Wood

    John Wood Well-Known Member

    Joined:
    Nov 13, 2002
    Messages:
    11,078
    Likes:
    153
    No luck at all. ran new ground from pin 7 of the cluster plug. Voltages at pin 2 and pin 11 check out. Very damp morning, but I'm confident that it will start working in a couple of hours as the air dries out. BTW, the dash needles flicker or bounces when the key is out, even when the ignition switch plug is pulled out of the back of the ignition switch I guess that could indicate the B+ is a little unstable or the console has some internal component fault. I replaced the battery with another, cleaned up the terminals, and even replaced the positive battery terminal. Everything is shined up and making good contact.

    I had some other weird issues which is leading me to believe I have a BCM or junction block problem. As I switch the key from run to off, the power door locks suddenly lock the doors and when I go to unlock them, they go back to lock and the lock motor runs a bit longer than normal. The door locks function and sound normal when the key is in the run position. This happened for about 2-3 minutes and then things associated with the door locks became normal again. The odometer is just beginning to light up when I hold the trip reset buttons down (just flickers and then and goes black).

    So I guess I'll try looking for ground issues or bad connections around the BCM/junction block area.
     
  10. John Wood

    John Wood Well-Known Member

    Joined:
    Nov 13, 2002
    Messages:
    11,078
    Likes:
    153
    I'm fairly sure this has to be a cluster issue. There is an 'in between' condition that occurs before everything starts working properly. With the ignition switch in the 'on' position, the odometer display lights up with proper mileage, but when you move the key to run it dies and the speedo and temp gauges flutter a bit and come to near zero. Right now with this in-between condition, if I run the cluster diagnostic test, everything works until it gets to check 3. Then it dies and the odometer goes dark.

    I noticed that the large power resistor on the back of the board is too hot to touch, but I see no darkening of the PC board, so I guess the heat might be normal.

    I was able to monitor the two incoming voltages to the cluster connector with the unit hooked up by using an insulation piercing probe test stick that I have in my collection of tools. Everything checked good under all conditions. The CCD bus must be working since I get the proper odometer mileage during the 'in-between' condition' ... i.e. switch in run. Also, I can monitor coolant temperature with my scanner via the data link connection port (OBD II diagnostic connector).
     
  11. John Wood

    John Wood Well-Known Member

    Joined:
    Nov 13, 2002
    Messages:
    11,078
    Likes:
    153
    Totally gone as of today. Can not even get odometer mileage to display. Needles still bounce around.
    I'm searching for a used unit. No amount of flexing, banging or hot/cold air treatment is helping. There is a bad component and I can't find a cluster schematic on the web. If I thought it was a power resistor or a bad capacitor, I would try to fix it, but with no diagram there is no much hope. I did notice a gray small splotch inside the florescent display at the lower left side, suggesting possible arcing.
     
  12. Bob Lincoln

    Bob Lincoln "CHECK FAULT CODES"
    Level 2 Supporter

    Joined:
    Jul 10, 2002
    Messages:
    29,875
    Likes:
    3,255
    That small gray splotch is normal for VFDs. It is the getter material, a sacrificial deposit placed there to neutralize reactive materials.
    Getter - Wikipedia (at https://en.wikipedia.org/wiki/Getter )
    However, if you look closely at the display, you may find the grids bulged upward and touching the filament wires, or the filament wires may be burned or broken and coiled up. This is a common failure mode. One that you can't see is if the power converter or the driver ICs have failed.
     
    somber likes this.
  13. John Wood

    John Wood Well-Known Member

    Joined:
    Nov 13, 2002
    Messages:
    11,078
    Likes:
    153
    Very interesting about the gray splotch.... thanks for that info...learn something new everyday. :)
    I found a used cluster that is supposed to fit a 96 Voyager/Caravan, 3 speed. $50 shipped from an EBay seller. It should be in next week. Now I just had to put off another potential buyer. I told him to call me next Wednesday.
     
  14. John Wood

    John Wood Well-Known Member

    Joined:
    Nov 13, 2002
    Messages:
    11,078
    Likes:
    153
    Put in the used cluster and it works like new. No more jittering and bouncing needles. Comes right up and it passes the diagnostic tests. So glad this is fixed.

    What it appears with the old cluster, was that when it was cool, it failed (not the humid damp mornings like I first suspected), but as the day warmed up it started working. When I went to deliver to a buyer, I ran the AC for the 20 mile drive and that cooled it enough for it to start failing.

    I never took the time to measure the key out, MA draw, but with the power resistor remaining warm and the gauges jumping with the key out, there had to be some power being used.
     
    somber, valiant67 and Doug D like this.
  15. John Wood

    John Wood Well-Known Member

    Joined:
    Nov 13, 2002
    Messages:
    11,078
    Likes:
    153
    Had a young man show up at the house this morning with $850 cash. The 96 Grand Voyager is officially sold! It was a great family vehicle for us, taking us on a couple of vacation trips to New England, Arkansas, Indiana, and several trips to Georgia... all from Tampa, Florida. It will be missed.

    I hope the new owner makes out OK with it.

    Thanks for all the advice from Allpar board members over the last 13 years that I owned it. We bought it with 120,000 miles on the odometer and sold it with 252,000. Even with all the repairs, I still liked this vehicle for drive-ability, comfort, and utility (hauling) purposes.
     

Share This Page

Loading...