1998 Chevrolet Malibu instrument panel temp gauge quit

Tiny
JIM198827
  • MEMBER
  • 1998 CHEVROLET MALIBU
  • 3.1L
  • V6
  • FWD
  • AUTOMATIC
Innova coder reader says 0118 high input temp sending unit this cross referenced to Haynes repair I replaced sensor properly checked previous and new sensor very close to same resistance values obtained test performed in car while running. Wire harness checks out to ohm value 1 ohm over test lead wire value 1.98 ohms PCM instrument panel does not look burnt the coil for needle has continuity I am unfamiliar with the movement driver IC or related components to verify operation is there a fixed value resistor I can check PCM with Haynes show electrical diagram as pin1 ground through PCM an signal to PCM book says voltage drop is method of PCM calculating temp but book does not mention voltage applied or current level used to reference the voltage drop to check temp sensor actively I'm using a multi meter on an 2000 ohm scale an book temp vs, resistance table to drive around. I know that the PCM can not run closed loop if it can't tell temp what can I do to fool the PCM into a continuous operating temp mode so it won't run in limp or is the last resort a trip to dealer for more money than the car is worth
Do you
have the same problem?
Yes
No
Thursday, June 18th, 2015 AT 5:11 PM

4 Replies

Tiny
CARADIODOC
  • MECHANIC
  • 33,699 POSTS
Wow, that's one really huge sentence! Please use some punctuation, otherwise it can be read three or four different ways, and you know I'm going to read it incorrectly.

You've already done way more testing than normal, but the mistake most people make is replacing parts based on fault codes. In fact, fault codes never say to replace parts or that they're bad. They only indicate the circuit or system that needs further diagnosis, or the unacceptable operating condition.

Also, don't get wrapped up in resistance readings. Those charts were of great importance to an instructor I used to work with because he didn't understand anything other than what was in textbooks. They're good for teaching theory, but beyond that, no mechanic refers to them to make a diagnosis. What you will have verified is the sensors were not open or shorted. No two sensors will ever be exactly the same, so don't get "wrapped around the axle" if you get valid, but different resistance values. One of the things Engine Computers do is learn the characteristics of various sensors when one at a time is replaced.

You'll get code 118 when there is an open circuit. It appears you have a better understanding of electrical theory than most people, so I'll add my wondrous circuit explanation. There's a series resistor inside the computer that voltage is dropped across when current is flowing. Suppose that resistor is 5,000 ohms. Temperature sensor circuits are fed internally from a 5.0 volt source. If the coolant temperature sensor were also reading 5,000 ohms, half of that 5.0 volts would be dropped across the internal resistor and half across the sensor. You'd see 2.5 volts at the signal / feed wire at the sensor, but only as long as it is plugged in.

If you unplug the sensor, there's no current flow, no internal voltage drop, and you'd see what you started with, 5.0 volts. That is an unacceptable condition and is what triggers the fault code. The ground side of the circuit can be open too, but other sensors share the same wire, so you're much more likely to have fault codes for multiple sensors if the ground is the problem.

It sounds like you're mixing up two sensors. GM used a variety of circuits, so I'm not sure which one you have. Most vehicles of this era used two totally different and unrelated coolant temperature sensors. They used a single-wire sensor for the dash gauge. That one won't set a fault code, at least not in the Engine Computer. Engine Computers always use a two-wire temperature sensor so the ground circuit can be monitored for proper current flow. You'll find very close to 0.2 volts on that circuit.

I left out the sensor itself up to now because failure of those is very uncommon. There is just one component inside them, a temperature-dependent resistor, (thermistor). Those rarely fail except on early '90s Ford products.

The place to start is by measuring the voltage on the sensor when the ignition switch is on. These readings must be taken by back-probing through the rubber seal, next to the wire, with the sensor plugged in. If you find between 0.5 and 4.5 volts, that is the acceptable range that won't set a fault code. Code 118 says you have 5.0 volts at the computer's terminal. If you find 0.0 volts, (actually 0.2 volts), at the sensor's signal terminal, there has to be a break in that wire between the sensor and the computer. If you're also seeing 5.0 volts at the sensor, measure on the ground wire. If you have 0.2 volts there, one of the connector terminals isn't making contact or is corroded. The sensor could be open too, but not two of them.

If you find 5.0 volts on the signal and the ground wires, there's a break in that ground wire between the sensor and where it is spliced to the ground wires for the other sensors that share that circuit.

If you DO find an acceptable signal voltage at the sensor, you can monitor that as the engine warms up. You'll see an interesting point of confusion though. As the temperature goes up, the sensor's resistance goes down, so the signal voltage will too. Typically what you'll see is something to the effect of, ... Oh, ... 4.0 volts at first, and it slowly drops to perhaps 1.5 volts by the time the engine reaches around 140 to 150 degrees, then all of a sudden the voltage will pop back up to around 3.5 volts, then start going down again. At a magic temperature, the computer switches in a different series resistor to reduce the range of the sensor and increase its accuracy. Once that occurs, a typical voltage at 200 degrees F. Is around 1.5 volts, but remember, that is going to be different for every sensor. To be absolutely certain, you have to use a scanner to view live data. That will show the actual voltage at the sensor and exactly what temperature is being interpreted.
Was this
answer
helpful?
Yes
No
Thursday, June 18th, 2015 AT 6:35 PM
Tiny
CARADIODOC
  • MECHANIC
  • 33,699 POSTS
Oops. It would appear I messed up again. I overlooked your title referring to the dash gauge. I found a wiring diagram and it does appear you have just the two-wire coolant temperature sensor. The Engine Computer sends data out on the "data buss" to the other computers, of which the instrument cluster is one. That would explain how you can have a dash gauge problem and a fault code caused by the same one sensor.

GM has a real common problem with failed "stepper" motors which are what turn the gauge pointers, but when you have a fault code too, we know we have to stick with the sensor circuit. Let me know what you find with those voltage readings.
Was this
answer
helpful?
Yes
No
Thursday, June 18th, 2015 AT 6:49 PM
Tiny
JIM198827
  • MEMBER
  • 2 POSTS
Thank u for responding. This is the first time I have received a response. I will donate to the site next check. The details you provided were exact. I found a bad connector. The heat of the exhaust charred through the casing. No ground, no sensor voltage ever would jump 3 8 gap. Previous owner changed harness an didn't zip tie away from exhaust. I never saw it under the ignition coils. Resistance I tested before asking was a result of ground to mainifold. Then I jumped the connection by running wires directly to sensor from my meter if u had not informed me I would have junked the car. Runs good thank u so much.
Was this
answer
helpful?
Yes
No
Monday, June 22nd, 2015 AT 5:18 AM
Tiny
CARADIODOC
  • MECHANIC
  • 33,699 POSTS
Dandy. Happy to hear it's solved and the car will live on.
Was this
answer
helpful?
Yes
No
Tuesday, June 23rd, 2015 AT 12:12 AM

Please login or register to post a reply.

Sponsored links