High temp on the gauge, but not overheating?

[starshine84]

I have a Westerbeke 71c diesel on a 42' sailboat. Last weekend, about 30 minutes after leaving the mooring, I noticed the temp gauge was reading about 230 (but not pegged). We usually run at 190 on the gauge. The audible high temp alarm had not sounded (it's supposed to sound an alarm at 212).

I shut down the engine, checked the strainer (which was clear), checked the coolant level (fine). I started the engine and confirmed that the usual amount of cooling water was being pumped out of the exhaust.

Shut it down again, and sailed most of the way back to the mooring. The last bit is a narrow channel with the wind on the nose, so we started the engine and idled back in. The temp gauge started about 150 (the engine had cooled while we were sailing), but climbed back up above 220.

Once back at the mooring, we checked the gauge (shorted the S pin to ground, it pegged to the right; disconnected S, pegged to the left). It seems the gauge has not failed catastrophically, if it has failed.

Next, I removed the wire from the temp alarm switch and shorted it to ground, and the alarm sounded properly. Of course this doesn't prove that the switch is OK, but it eliminates the alarm buzzer.

Next, I disconnected the gauge at the temp sender on the engine and started the engine, and measured the resistance to ground as the engine warmed up (at 1000 rpm or so in neutral, not loaded). I took temperature measurements on the engine using a laser infrared thermometer, at the block next to the temp sender and at the thermostat cover. As expected, the temp vs. resistance curve was a nice logarithmic curve. The temp readings never went over 177, even after 20 minutes or so (when it leveled off). The highest temp I could find on the engine was the oil pan, which was 188 (by the way, all of this argues for taking these readings on your engine before something goes wrong, so you have a baseline!).

Here are my conclusions, and I'm hoping for confirmation or an explanation of why I'm wrong:

1. The engine isn't overheating. The block next to the temp sender was 175 degrees. The alarm didn't go off. Plenty of coolant, and plenty of sea water being pumped.

2. The gauge or the sender is bad. I wish I had more confidence about this. My thought is that when either of these fail, they fail completely. I haven't been able to find an exact temp vs. resistance curve for my sender. The only one that I found was similar to my curve, but shifted (it had a lower temp for same resistance). My plan is to replace the sender as a first step.

Could it be something completely different, like a thermostat problem or an air bubble somewhere?

The engine has been working fine all season, and I haven't done anything to it that would have caused this. Two years ago I had the heat exchanger off and flushed the cooling system, so I would think that any air bubbles would have worked their way through the system by now.

Any ideas are much appreciated. Also, when I replace the sender, do I have to worry about air getting into the system?

[Blue Stocking]

My Westerbeke W46 runs at about 200 on the gauge when the water temp is in the 80s, and about 180/185 when water temp is high 60s/low70s. My IR gun confirms this. What are your water temps ?

[hellosailor]

With some patience, you CAN cross-check the sender and gauge to pin this down. Pull the sender, and take it home. Put it (just the business end) in a pot of boiling water, and attach an ohm meter. Note the resistance at boiling, which will be 212F, for most of us. Let the water cool a bit, use your IR thermometer or cooking thermometer to measure the water and again note the resistance at, perhaps, 200-190-180-170 degrees. OK, now pick a couple of convenient points. The sender might range from 240 ohms cold to 0 ohms extrahot, with "normal" temps around 100? 150? ohms. Youll need to check a manual. In any case, you get a bag of assorted resistors and...yup. Put the same value you found at 180F in the boat, from ground to the gauge. Do the same for 170 and 190. See if your gauge comes up showing the same temperatures. Odds are that one or the other will disagree very much with what the manual says. The service procedures usually state how many ohms a specific temperature should be, so finding out which one of the two is off, should tell you the one that needs work. FWIW I have had a brand new, brand name, perfectly good temp sender be DOA, and that certainly screwed us up for a while.

[newhaul]

From what you did and posted here I would have to say likely the sending unit failed or is failing . When in doubt change it out.
Cheap insurance. ( usually when gauges fail they don't work at all . )

[starshine84]

Thanks for the quick responses.

Our current water temps (northern Chesapeake) are high 70's, and mid 80's in the middle of the summer. 190 is the usual temp with water temps in mid 80's.

I will change the sender out, and run the experiment that hellosailor suggests. I hadn't thought of just using some resistors to check the gauge; great idea.

Would everyone agree that if the block temp near the sender and the temp of the thermostat housing are in the high 170s that the engine is not in fact overheating?

[hellosailor]

"Would everyone agree that if the block temp near the sender and the temp of the thermostat housing are in the high 170s that the engine is not in fact overheating?"
"Agree" might be too strong a word but I'd tend to suspect it wasn't overheating.


You can also go to Ace Hardware. they sell black rubber stoppers, like the kind you'd use in a chemistry lab. Find a nice robust one with a hole in it, or drill a hole in it, and stick that handy kitchen thermometer through the cork. Shove it snugly into the place the sender came out of, and again, you've got a way to get a reality check on how the coolant itself is doing in there. You may need to pull the thermometer out & stick a big eye hook in the cook to pull IT back out again, but that's no big trick.

[roverhi]

Could it be thermostat issues?? Pulling the thermostat and putting it in boiling water will tell if it is working. A little investigation on a stove with a meat thermometer will tell at what temp it's opening.

[pesarsten]

I wonder if some corrosion at the connection point to the sensor could cause the problem. Perhaps you should replace the wiring connection.

[starshine84]

I thought about the wiring at the sender, and trimmed it off to clean wire and crimped on a new ring connector. But since the sender has lower resistance as the temperature gets higher, a poor connection or corrosion would (I think) lead to a low reading on the gauge, not high.

I also thought about the thermostat and will consider that next if replacing the sender doesn't solve the problem.

Thanks again to everyone for the feedback and suggestions!

[GordMay]

Approximate Teleflex sender resistance values are listed:
Here ➥ http://www.cruisersforum.com/gallery...=500&userid=79

See also ➥ http://www.cruisersforum.com/forums/...sting-645.html

If sender is shorted (0 ohms) gauge will read above 240F.
If sender has infinite resistance (Open) gauge will read below 120F.

[a64pilot]

Your not overheating, you have a bad gauge.
I have a story of a guy who spent tens of thousands and months chasing an oil overheat on a PT6-34. I went out to help, I had told him to not start the aircraft until I go there, I got in the airplane and turned on the battery and oil temp was 60C, I looked at him and told him I had found his problem, his expert mechanic had never checked the gauge.

When the engine was stone cold and at ambient temp, the gauge read way high, in his it was the gauge, it usually is the gauge with aircraft, I expect the same for boats too, but I'd replace them both, too easy to do and the other component may be on its last legs too.

You have an IR temp gun, something that I have found invaluable.

But with anything, low voltage, oil pressure or temps whatever, first verify the gauge, often the problem is right there.

[bdbcat]

Check the ground circuit at the gauge.

Bypassing the wiring harness, use an ohmmeter between the engine block and the gauge ground terminal. If this is higher than a few ohms, the gauge will read high, and the harness has a connector fault. BTDT.

Dave

[starshine84]

The curves of temperature vs. resistance match very closely for my old sender and a new one (the old sender was tested with an ohmmeter and IR gun on the engine, and the new one in a pot of water on the stove with a digital meat thermometer).

This points to the gauge or wiring. I'll get some resistors the check the gauge, and check the resistance in the ground wire at the gauge.