Introduction: Erratic Instrument Panel on GM Cars
This is my 1999 Oldsmobile Alero. In early September 2010 the instrument cluster (speedometer, tachometer, temperature gauge, fuel gauge, and various warning lights) stopped lighting up and registering data immediately when I start the engine. When that happened the radio and turn signals worked; but the power windows, door locks, and cruise control were inoperative.
The instrument cluster might work problem free for a few days, but then would not turn on until I had driven a block or two after starting the engine. It did always came on, eventually. Sometimes the instrument cluster failed to work properly each time I started the car over several days.
Over the years, connections and other things in circuits change to add more resistance to the circuit. This can have a large effect on circuits involving low voltage signal pulses. This Instructable is about removing one source of increased resistance, at least on this model of car. First, here is some general information on instructment cluster problems for slightly newer GM models:
General Motors cars had instrument cluster problems from 2003 through 2006. When I researched instrument cluster problems on the Internet what I found raised concerns in me. Many of those vehicles had problems that began with the instrument cluster, but eventually led to the engine refusing to start. Sometimes drivers were stranded on the road for at least 10 minutes. Those cases usually involved a security warning light showing itself on the dashboard. A friend's speedometer quit working on his Chevrolet Impala and the dealership quoted him a price of $800 to fix it. The car was no longer new, but still had some life in it. He did not want to spend that amount of money and traded the car for a new one.
The problem in those instrument clusters is a stepper motor design that fails. A number of services around the USA can be found on the Internet that will rebuild a GM instrument cluster with new stepper motors guaranteed never to fail during the remaining life of the car. Also, in 2006 GM began issuing phased recalls to fix the problems on vehicles no older than seven years and with fewer than 70,000 miles on the odometer. This business will rebuild your instrument cluster. This vendor will sell the new and improved stepper motors so that anyone with very modest soldering skills and a low wattage soldering iron could easily install the new motors on his vehicle's instrument cluster. You can also buy the new stepper motors at Amazon for about $6 each. If you have one of these GM vehicles and are having problems with your instrument cluster, here is a very good video on replacing the stepper motors.
Although my car works well in regard to everything else, it was not immediately clear to me whether my car used the same stepper motors as those models under the recall. A problem that might result in the engine refusing to start would mean I might have to sacrifice recent upgrade repairs I made (struts, tires, alternator) and shop for a new car. Certainly, I am not about to spend $800 on a car worth only about twice that much on a trade for a new car.
Step 1: Eliminate Other Possible Causes
A worn ignition switch could be the reason for erratic behavior at the instrument panel. But, removing the ignition switch to check it usually means removing the entire dashboard, and that is a very involved task to be avoided, if at all possible.
This is the left instrument panel cluster (IPC on schematics you might find in a manual for your car) junction block. It can be used to determine if the ignition switch is good or not. A 10 Amp. fuse (red) marked in the key as "IPC/ BFC ACC should be without power on both sides when the ignition is "off," but should be live on both sides when the ignition switch is "on" in any position. There is a ground connection with a black wire at the upper right corner of this junction block. Attach one lead of a multi-meter to the ground screw. Set the meter to a DC voltage setting greater than 12 volts. Use the probe on the other lead to touch the small metallic spots on the top surface of the fuse. If only one of these registers 12 volts on the meter when the key has been turned, the fuse is probably blown. If both sides register 12 volts with readings one after the other, it means the ignition switch should still be good.
Update: Although the tests I used indicated the ignition switch did send voltage through the proper fuse at the proper times. But, the problems I was having came again. I was able to remove the ignition switch and check it for continuity in a series of tests proscribed by the Haynes manual for my car. My ignition switch failed a couple of those more detailed tests. I have now replaced the ignition switch. I published another Instructable on replacing the ignition switch on this model without removing the dashboard.
Step 2: The Driver's View
This is my instrument cluster from the driver's seat of my car. Notice that the tachometer needle registers about 800 rpm at cold idle and the fuel gauge registers about 3/8 of a tank. The instrument cluster is working as it should. See also the second photo. When the key is first turned, all of these lights should come on. When the instrument cluster did not work properly, all needles pegged at zero while the engine was running. None of the lights shown in the second photo were lighted. The first task will be to access and remove the instrument cluster. Fortunately, that can be done very easily in just a few minutes by removing only six screws. There is no need to remove the whole dashboard.
Disconnect the car battery to protect against the air bag in the steering wheel from deploying accidentally.
See the yellow text boxes on the first photo to follow this description. Remove the top half of the steering column cowl. You can pop it off easily with just your fingertips. Loosen the steering column tilt lever and drop the steering column as far down as possible. Remove the two screws at the top portion of the bezel. Pull the bezel part of the way out. (You may need to loosen it with a screwdriver blade. See the text boxes.) Disconnect the wires from the trip meter switch to the left of the steering column. There are four screws holding the instrument cluster in place. Remove these screws and tip the instrument cluster so you can begin to remove it. A wiring harness attaches to it just above the speedometer dial. Pinch the blue tabs and pull the harness from the instrument cluster. (Before reconnecting this wiring harness I sprayed the female connector with TV tuner cleaner from Radio Shack. Then I plugged and unplugged it several times to make certain the connections at the harness were as good as possible.)
Step 3: Opening the Back of the Cluster
The instrument cluster opens very easily. Simply pry the retainer tabs open with a butter knife or a screwdriver blade while gently pulling the back of the cluster away from the body of the cluster as much as possible. There are about a dozen such releases around the perimeter of the cluster. Pry the black catches away from the white main body of the cluster in the direction of the red arrow.
I was unsure this fix would work at the time I was doing this project, so I did not make photos. Later, when it appeared that I had found a good solution to my problem, I decided to make it an Instructable and have used graphics I did in MS Paint for the next steps.
Step 4: Posts and Sockets
The left part of the graphic is a drawing of a post rising from the back of the motors that drive the needles on the instrument cluster gauges. The right part of the graphic shows the metal socket into which each post slides. These sockets are mounted on the circuit board. The socket here is shown inverted, or upside down. The red circle marks the hole through which the post slides to connect the posts to the circuit board.
Each motor has four posts. Some posts are not used, so a couple of the gauges have only three, not four, sockets on the circuit board.
A little extra resistance causes a poor circuit in the instrument cluster, and that causes various problems with the gauges. Spring connections of bare metal work well when they are new, but time means oxidation (corrosion) that may not even be visible to the eye. Yet, it is enough to add electrical resistance that can cause problems in electronic circuits. In researching instrument cluster problems I found a number of possible causes, including a change in the amount of electrical resistance in a circuit.
Step 5: Bridge the Posts and Spring Connectors With Solder
The graphic shows one of the spring connectors on the side of the circuit board seen when the instrument cluster is opened. The circuit board can be gently lifted off of the instrument cluster. I used a very fine sandpaper to brighten the ends of the posts rising out of the motors. Then I pressed the circuit board back in place.
I bridged the solder traces for the spring connectors on the circuit board to the ends of the motor posts using solder. I used a magnifying glass to inspect each bridge in order to guarantee a good solder joint. Where necessary, I used some small gauge bare stranded wire to assist in making the bridge. This gave a solid electrical connection between the circuit board terminals and the metallic posts that serve as leads to the motors.
Step 6: Sweet Success, But With a Twist
The Internet has many postings from owners of GM cars, especially Alero owners, who find erratic behavior on their instrument clusters and discover they also cannot use their power windows. There can be several reasons for this behavior. The instrument cluster on my 1999 GM worked well for a few days. Then began to act up again. I did some checking on the signs of a failing ignition switch. I decided to remove and test the ignition switch. As mentioned above, it failed some of those tests. I do not regret soldering connections on my instrument cluster, but it was not the whole problem.
I found a way to change the ignition switch without removing the dashboard. A new ignition switch did solve my problem. Looking back, some tests seemed to indicate my ignition switch was still good. I think it may have been a difference between voltage and current. The first is pressure and the second is volume. There may have been sufficient voltage, but not enough current to power everything going through the instrument cluster. That is just my guess.