Legacy Temp sender failures?

[GeneralDisorder]

I've got this '97 Legacy. Former owner said the temp gauge would peg out occasionally but it never overheated and then it would just drop. He described it as happening quickly. I marked it up to a bad sending unit. Drove home 60 miles with no problems.

 

So I go in for a timing belt/water pump, etc. Lo and behold - someone punched the guts out of the thermostat . I put in a new one.

 

After that I ran it up to operating temp in my garage. I noticed the temp climbing - pulled out my gun and the reading from the upper radiator hose is 175* F . Obviously somoene had bandaided the broken sending unit/gauge by punching out the thermostat. At this point I should have taken a resistance reading of the sender but I figured it was bad and this would be a slam dunk. I had a used coolant cross-over laying around, pulled the sending unit, cleaned it up and installed it. Drove around the neighborhood and all is well..... till I'm about to turn down my street and I look down and the gauge is pegged out again.... WTH?!? Temp gun says 195*F so I'm know I'm not overheating.

 

This time I figure.... like maybe a wireing short or something. Pull the wire off the sending unit and the needle drops to C. Ok so it's not a wireing short. Now I'm getting a little frustrated - pull out the DMM and the resistance is 20 Ohms. Checking with some PDF manuals I have this should indicate I'm at the top of the range for the sending unit (248*F).... but I'm clearly not. So the gauge is working and giving me the reading it see's based on the resistance of the sending unit - which appears to be wrong.

 

Is this just a random fluke? I've never read about many of these failing and this is the first time I've seen it. I've worked on and owned a lot of Legacy's and this is the first one (two?) that I've ever seen fail.

 

Am I missing something obvious or should I just pop for a new from the dealer?

 

GD

Edited May 28, 2010 by GeneralDisorder

[Cougar]

The only thing I can think of that may be causing the problem other than an intermittent sensor is a faulty headgasket. I suggest you try another sensor and hope that clears the trouble. If it still persists then I think you may have to work on the headgaskets.

[GeneralDisorder]

The only thing I can think of that may be causing the problem other than an intermittent sensor is a faulty headgasket. I suggest you try another sensor and hope that clears the trouble. If it still persists then I think you may have to work on the headgaskets.

 

That was one of the first things I considered. But it just doesn't add up.

 

How does that account for my temp gun reading correct (thermostat regulated) temp, and the sending unit reporting (via the 20 Ohm reading) that the temp is 248*F. The gun says 195*F. The gun is reading 4" away from the sending unit on the same coolant passage. I can't see a gradient of over 50*F from that distance. And nothing I am seeing says that it's really overheating.

 

It's an EJ22 with 147k on it. Runs great with good power and didn't lose a drop of coolant on the trip home with it. Overflow bottle is nice and clean...... I'm just not seeing signs of a HG failure here.

 

GD

[idosubaru]

no differences in sending units right - first gen to second or something?

[GeneralDisorder]

Turns out - I'm pretty sure the radiator is bad. Appears to be clogged and the flow through it drops off so rapidly that where I was measureing with the temp gun really IS 40*F cooler than where the sensor is at. I made a few changes, thought everything was good and went for a drive - it spiked and I pulled over. The radiator was cold in the middle and on the bottom. Top was hot.

 

Looks like a clogged radiator.

 

GD

[Cougar]

What you say about the radiator makes sense GD. Hopefully a new one fixes things up. I suspect it will at least.

[GeneralDisorder]

Yeah - I pulled over and let it cool. Then I was able to drive home with the heater on full blast and driving slowely.

 

The guy told me he had changed the coolant because it "looked nasty". And there was a brown coating on the inside of the old radiator hoses. Looks like this is (hopefully) a case of neglected cooling system and nothing more. These newer radiators are really thin - the whole core must be 1/2" to 3/4" thick. Doesn't look like they would stand much in the way of debris or corrosion.

 

The fact that is seemed to drive fine with no thermostat would tend to point to a lack of flow also. The temp wasn't low without the 'stat - it was where I would expect it to sit - just a shade below half.

 

GD

Edited May 29, 2010 by GeneralDisorder

[Cougar]

Considering the age of the car a new radiator sure wouldn't be an unexpected repair job or replacement part.

[WoodsWagon]

I'm having the same problem with a 98 forester. However, the gauge pegs as soon as you turn the key on and let it sit for a bit. With the engine running it comes up to temp and stays there, as evidenced by the OBDII scan tool data reading from the computers cts. The gauge just peggs out though. I swapped senders with a spare I had, and it still pegs so I've left it at that.

 

Hard to sell the car, esp with the DOHC 2.5, which is known for headgaskets, with a pegged temp gauge.

[Cougar]

I'm having the same problem with a 98 forester. However, the gauge pegs as soon as you turn the key on and let it sit for a bit. With the engine running it comes up to temp and stays there, as evidenced by the OBDII scan tool data reading from the computers cts. The gauge just peggs out though. I swapped senders with a spare I had, and it still pegs so I've left it at that.

 

Hard to sell the car, esp with the DOHC 2.5, which is known for headgaskets, with a pegged temp gauge.

 

It sounds like there is a problem with the wiring to the sensor. Check the voltage on the wire while it's connected to the sensor and then remove it and see if the voltage goes higher. If it stays about the same then the wire has a problem, possibly a shunt to ground.