Cowtown Dave

Problem: my water temp gauge reads about 60 degrees higher than actual.

I tested the gauge as I have read here by disconnecting to zero out and grounding to max out. That happens as it should. I have replaced the temp sensor with two different ones. Actual temp I'm using a infrared thermometer shooting at the hose coming out of the therm housing. Any suggestions or something I'm missing?

Canuck62

Read on lectric limited web site that their temp sender is correct,,,, but i have never read from any forum member if it does in fact read as it should

Powershift

Could be the gage or the sender. Start here:

1. Disconnect wire to sender.

2. Measure resistance from sender tip to a good engine block or battery ground.

3. Run engine. During warmup, measure this resistance and also shoot the sending unit with your IR gun.

4. Do the same at normal operating temperature (thermostat open).

5. Post the results here.

We can then tell you where the problem is.

Larry

Jackfit

Most gages are no linear.... my gage reads on the unmarked line at 180 on a gun. But if you look at it it looks very close to the 150 mark and not 200. However, when my temp is at 200 by the gun, my gage says 200

Jack

Cowtown Dave

Car is at operating temp now and the resistance reads 98.5 -99.00 ohms. Temp is 180 at the sending unit. Gauge seems to be acting normally just off by 60 degrees. gauge reads 240 at operating, temp at the unit 180

GTOguy

I installed a Lectric Limited sending unit in my '65 GTO.....after 30 years of ownership, now have a gauge that reads actual engine coolant temp as verified with a pyrometer. The NAPA sending unit installed previously read 20-30 degrees high.

toddalin

Did you put Teflon tape on the sender threads? This has been known to cause false readings.

Cowtown Dave

One of the senders was a lectric Limited, the other was a Wells TU5. no tape just some anti-seize.

Cowtown Dave

90 ohms
temp at sending unit 167
gauge reads 180

car is cooling down and has been off for about 30 min

buns

Lots of info here:



http://www.lbfun.com/warehouse/tech_...eCF-Thread.pdf

jim lockwood

Any possible change in reading due to the teflon would make the gauge read low, not high, because the teflon would add resistance in series with the sender.

That said, since the threads on the sender are tapered and seal by a wedging action, they will also cut through the teflon (or anything else anyone puts on the threads) and make good contact with the manifold.

Jim

65GGvert

The readings aren't off far. When you disconnect the temp sensor wire, does the gauge read at the very bottom? I have never had a bad gauge, but the needle may not be the correct rotation on the shaft??? The Duralast TU5 has always been very accurate for me, from Autozone.

GTOguy

I respectfully have to disagree with Jim: temperature sending units are thermistors, and their resistance goes DOWN as they get warmer, allowing more current to pass through them. With most electronics, resistance increases with temperature, but not with thermistors.

GTOguy

Strike that.....Jim is correct, I misread his post and posted TOO SOON (again!)

65GGvert

One other thing I'd check. If the firewall connector has corrosion between the pins, you may be picking up some current from another source, or reading a lower resistance because of another nearby pin. I'd clean the pins on the firewall connectors before I started on the gauge. It's not a real common result to get a steady wrong reading with corrosion, but it's worth a shot. In fact, if you have someone watch the reading and wiggle the plug (the one closest to the centerline of the car), see if it makes the needle move.

DUB

I know Limited Lectric states their sending units are correct...but for the life of me....I do not trust any of them. Sometimes they are fine ad work correctly and then other times they fail badly.

When I am working on a temp gauge....I have seen the gauge work great to a point and then go crazy when teh coolant temp starts to get above 200 degrees.

And I have had a sending unit be 100 degrees off.

And not that many of you would go through this.....but some customers I have want the temp gauge to be RIGHT....so they pay me to invest my time in gettign it right...but I often times will use resistors and 'trick' the circuit and check to see what it takes to get the needle to go to 180 or 210, etc. I also check the integrity of the wire and circuit from the sending unit connector to the terminal on the temp gauge and see if I actually do have any issues in the bulkhead connector.

So those who have their temp gauge working great....consider yourself lucky because it can easily not be that way.

DUB

Powershift

Per GM Factory Drawings, a sending unit reading of 99 ohms corresponds to a temperature of about 190 F (and not 180 F).

A sending unit reading of 90 ohms corresponds to a temperature of about 200 F (and not 167 F).

Per GM design, as engine temperature increases, sending unit resistances should DECREASE. Yours seems to be going in the wrong direction, or your readings are suspect.

You need to go back and repeat your test. You also need a few additional readings as well to help nail things down.

Per Ken Anderson reading from GM blueprints, this is what the temperature and ohm curve should be. Ken is the guy who makes all the reproduction mid-year dash temperature gages.

Values shown in () are from Wilcox website for what Ernie tested for factory 1966-67 mid-year dash gages. 1965 temperature gage info from Ernie is shown in []. A little bit of differences all around.

Temperature (F).............................Sending Unit Ohms
100..................................... .........350.....(274)......[340]
180..................................... .........105.....(109)......[128]
210..................................... ..........76.......(79)
240..................................... ..........72....................[68]
250..................................... ..........52.......(51)


Larry

Cowtown Dave

Thanks Larry, I'll spend a little more time triple checking my readings at different points.

Ron Miller

I went through several and numerous past threads on the same problem, verified my gage resistance(s) required to produce varied temperature readings on the instrument panel gage, charted the readings on my temperature gage, used the closest matching sending unit, nada, nada, nada, . . . . . . nothing worked for me . . . .

The final solution after trying several sending units was the Wells TU 5 from Autozone, my temperature readings are spot on, just as in the above photo. This is on a 1966 Corvette, I believe there may be some difference in the 63-65 since the scale on the 66-67 gage face was revised from the 63-67's.

My experience, the only thing that worked for me. No need for an in-line "balancing" resistor such as are being peddled. (Also, past experimental test results using teflon tape have shown that use of even multiple wraps has NO effect on grounding of the sending unit, the threads will cut through and ground . . . . . this info has also been posted on the forum in past threads . . . .)

Just going from memory, but I think I recall on one of the past threads that JohnZ posted that the Wells TU5 also reported the correct temperatures on his '67 Corvette, he also stated that perhaps he just got lucky with that particular unit . . . .

After38Years

The hose may not be the best place to shoot. Depending on the surface finish and its insulating properties, it may not be an accurate indicator of the temp.
Also, does your IR thermometer have an adjustment for emissivity? Try putting a piece of flat black tape on the thermostat housing and shooting that to compare with the reading from the hose.