BallerCamaro21

I swap a 5.3 into my 96 trans am and I hook my snap on scanner to it and the scanner says the coolant temp is at 217 with the car not running. so I check my wires and I have good ground but the yellow wire says I only have 1.8 volts. Any idea on what would cause this problem? I also check the wire at the pcm on the blue connecter pin 74 and it also said 1.8 volts. Maybe a bad pcm?

stevieturbo

The sensor should have a sensor ground from the ecu, and the signal wire will have a voltage present, this will vary depending on the internal pullup resistance used.

And are you measuring with the sensor plugged in, or unplugged ? And does this make a difference to the temperature reading ?

BallerCamaro21

when I had the sensor plug in the scanner said 217 and I used a power probe and tap into each with and the yellow wire had only 1.8 volts and the grey had a ground to this wire. And when I unplug the connector everything said the same thing and nothing change

stevieturbo

sounds more likely either a broken or shorter wire.

Does temperature reading change if you short between the two wires ?

BallerCamaro21

If I get the yellow wire 12 volts then the reading on the scanner goes down to -38 and stays at that temp

BallerCamaro21

I can see that it may be a broken wire and all but wouldn't that mean if I have a broken wire to the sensor the reading read the way it does but shouldn't the wire coming right out of the pcm read 5 volts?

stevieturbo

I certainly wouldnt be applying 12v to either of the wires, that could cause damage to the ecu.

Typically yes the signal wire to/from the ecu would normally be closer to 5v, but again it depends on the internal pullups within the ecu, some will be different.

And if testing for ground at the other wire, do not use a test light or anything that will draw current otherwise that too could cause damage.

As said, do the test I stated, short the wires and see if that changes the reading. That will at least verify wiring integrity from ecu to where you are testing.

Although from the 12v test that does suggest at least that wire is connected to the ecu ok.

Some older ecu's may reference the sensor via 5v, so you'd have a 5v supply then a signal return. In that case you wouldnt expect 5v or anywhere near on that return wire. But it may not be 0v either.

BallerCamaro21

Ok when I test I didn't use a 12 volt test light I use power probe 3 and when I ground the yellow wire on the connector the reading didn't change at all. It still read 217

BallerCamaro21

Is that 5 volt ref wire tie into another 5 volt ref ?

stevieturbo

The 5v supply from the ecu to sensors will be a common 5v supply for all sensors that require 5v.

BallerCamaro21

So if all the 5 volt are on the same wire then I should be able to cut the wire by the pcm and if the wire going back to the pcm has 5 volts after I cut it then I have a shorted wire from the pcm to sensor but if it doesn't change to 5 volts then I have a bad pcm?

stevieturbo

I wouldnt just go cutting important wires for some strange test, you'll end up getting loads of error codes and problems thrown up.

Just do the simple 5 second test I stated.

And without a wiring diagram, there is no point making assumptions. The 5v supply whilst common, may be present on more than one ecu pin. Looms are all made differently so routing etc will vary from car to car.

Why not get a wiring diagram and see exactly how your particular car is configured etc.

Project GatTagO

iTrader: (7)

Coolant temp has two wires and measures resistance. Three wire sensors are typically 0-5v signal.

Andrew

stevieturbo

Correct, but the ecu will typically apply 5v across the sensor and see what comes out the other end via that resistance.

But there are two methods, a sensor ground at one end and a voltage close to 5v applied via an internal pullup resistor.

Or a plain and simple reference 5v to one side with a straightforward return on the other, although that is far less common and usually only very old cars that take that approach.

Hence it is normal with an unplugged sensor for one wire to be 0v, ie ground, and the other close to 5v but not quite 5v in the case of the pullup method, or it should be almost bang on 5v for if it actually has that voltage applied.

Although he seems to suggest he is seeing 1.8v at one of the wires with the sensor unplugged which is odd.

ECU failures or problems are rare, but possible. But certainly possible to cause damage to them.

Most basic check to confirm wiring integrity is to simply short the two wires and see if the ecu recognises this.

BallerCamaro21

Thank you for all the help
The problem was a broken signal wire