1966 L79 Temp gauge/sender observations.

[Rich G.]

I’ve had the 66 for 22 years and normally runs at 210 on the gauge and higher in traffic. That’s according to the gauge. According to my IR thermometer it’s running at 180 at the hottest point in the cooling system when the gauge says 210.

There was an excellent article in the Restorer a few issues back that dove into this, theorizing as the temp goes up, the resistance goes down and the internal resistor (thermistor) adds heat. The theory in the article was that the resistor is not internally thermally bonded to the case, which makes it less likely to be able to dissipate the heat it is generating itself, which lowered the resistance even more making the gauge read too high and climb while the true temp of the engine is stable.

So, it’s cold, foggy and generally a day for experimenting. First I got the car up to temp and made sure the thermostat was open. The gauge read 210 IR thermometer read 180 at the base of the sender. Hottest spot I could find. I shut down and quickly disconnected the lead to the sender and measured its resistance. Engine not running. It read 120 ohms which corresponds to 180 degrees on the table I have.

Then with key on but not running I substituted a resistor box set to 120 ohms and observed the gauge read 180 degrees. WTF?

The other thing I have noticed is the tip of the sender if hotter than the body of the sender when reading with the IR thermometer which would support the theory that the resistor is not thermally bonded to the case properly.

I think, because of that the resistor cools more quickly and by the time I measure the resistance with the power off it had stabilized.

If you do Ohms law with the system voltage ar 14 volts and temp sender at 120 ohms (180 degrees) that resistor has to dissipate 1.6 watts. When the resistance gets down to 80 Ohms (which equates to 210 degrees on my chart) that resistor has to dissipate 2.45 Watts. That’s a lot of power for a resistor to dissipate not in free air with no heat sink.

I did notice that if I measure resistance from unconnected lead to ground it reads 70 Ohm. Clearly it’s reading back through gauge (shunt) but since the gauge reads close across the whole range and the sender reads the correct resistance at 180 degrees and without having a schematic of the whole circuit, I chose to chalk that up to something else I don’t understand. Like how my wife thinks.

I do have an NOS sender I got 20 years ago. I’ve given it the boiling water test and see it’s ok resistance wise there but there was no current going through it so who knows what it will do in the car? One theory from the restorer article was that original and/older senders had better thermal bonding.

[Mark F.]

Re: 1966 L79 Temp gauge/sender observations.

[Rich G.]

[Mark F.]

Re: 1966 L79 Temp gauge/sender observations.

Mark, thanks. What you point out makes complete sense...What do you think a regular thermometer with the proper range suspended in the coolant tank would read if my gauge is reading 210 and my IR is saying 180? Closer to one or the other? Rich.

[Rich G.]

Re: 1966 L79 Temp gauge/sender observations.

Mark

Yes, the L79 is the SB 327/350 and has the separate expansion tank. The car does not ever dump any coolant. This adventure was partially motivated by trying to understand the difference between the IR readings and the temp gauge. Your explanation satisfied my curiosity especially in the light of the fact that I’m not having a problem. When I first got this car, it did. The addition of the correct radiator from DeWitt, the correct fan clutch, proper vacuum can and timing and adding the proper seals around the radiator fixed that.

The other thing that got me thinking was the Restorer article about some aftermarket senders exhibiting self heating once the temp got high enough to have the resistance be on the low end.

After my last post I realized my proposed next experiment wasn’t going to prove anything. If the weather was better I’d be driving instead :-)

I used to own a 68 427/435 and I agree with your observation about how much (little) coolant was the correct level to be maintained. That car ran at 200 degrees on the hottest days. I don’t think I ever checked it with the IR thermometer. Of all the issues it had, overheating wasn’t one of them.

I never knew why some configurations used the expansion tank and some didnt. If I recall, maybe had to do with copper vs aluminum radiators. My BB radiator was copper (I think) I did replace it with a DeWiits repro when I had it.

Thanks for your input. I’m always learning…..or trying. :-)

Rich.

[Jim L.]

Re: 1966 L79 Temp gauge/sender observations.

That was my article, Rich. Thanks for the kind words.

You are right to believe that the thermistor inside the sending unit cools too rapidly for you to measure it's self-heated resistance.

The way to test for self heating is as follows:

1. Warm up the engine to normal operating temperature. Going for a drive would be my choice. Note the indicated gauge temperature.
2. Shut off the engine.
3. Wait 20 - 30 seconds.
4. Turn on the ignition but do not start the engine. Again note the gauge temperature and watch the gauge needle.

If the gauge initially indicates the correct engine temperature and the needle begins moving toward higher temps, you are observing self-heating of the sending unit.

It's not hard to understand why self-heating happens with modern, replacement senders. The picture below compares an original Delco sender to a NAPA replacement. The Delco is on top:

8088.jpg

Note the comparative sizes of the thermistor elements. The original Delco thermistor is simply larger which gives it a greater ability to withstand heat.

Now look more closely at the NAPA sending unit:

8085.jpg

Looking at the small insert it's immediately obvious that the thermistor element is cocked and not in good thermal contact with the bottom of the sender case. Not obvious from this picture is that the bottom of the thermistor element also seems to be slightly convex which reduces its thermal contact to a very small foot print.

Bottom line here is that the thermistor in this replacement sender can NOT shed heat that builds up due to current passing through it whereas the original Delco sender can.

Disclaimer: These are the only two sending units I've destroyed in my effort to understand why modern senders result in high indicated gauge temperatures while otherwise passing all the usual resistance vs. temp characterization tests. So I can't say for sure that other senders are constructed similarly to the NAPA sender but I suspect they are.

In any event, the Arduino-based solution to the self heating problem to which I alluded in my article solves the self heating problem two ways:

1. The current being passed through the thermistor is reduced by more than an order of magnitude.... less overall heat results.
2. The source of current through the sender is a current source rather than the Voltage source as our old Chebbys were originally designed.

#2 is particularly significant.

If the sender resistance tries to decrease due to self heating, the actual power dissipated by the thermistor element actually DECREASES which reduces the self heating thereby INCREASING the sender resistance. For those who understand such phenomena this is what's called negative feedback which stabilizes the overall system.

I hope this helps those who struggle with the problem of so-called "inaccurate" temperature senders.

Jim

[Mark F.]

Re: 1966 L79 Temp gauge/sender observations.

What issue of the Restorer is your article in, Jim ?

[Jim L.]

Re: 1966 L79 Temp gauge/sender observations.

What issue of the Restorer is your article in, Jim ?

Fall 2023, Vol 50, #2

[Mark F.]

Re: 1966 L79 Temp gauge/sender observations.

Fall 2023, Vol 50, #2

[Jim L.]

Re: 1966 L79 Temp gauge/sender observations.

Thanks Mark.

I used a cooking thermometer. I don't have a liquid-based thermometer which can register temperatures as high as I needed to measure.

Arduinos..... seems like every time I turn around, I find another application for one. And, yep, at approximately $10 each, they are sooooo cheap.

[Mark E.]

Re: 1966 L79 Temp gauge/sender observations.

Apologies if I'm stating the obvious or repeating something already stated (I haven't read the article)...

Years ago, I corrected the gauge reading in a '70 Chevelle by adding a rheostat to the sending unit lead and adjusting it as needed. Once dialed in, I replaced it with the appropriate resistor. They're small and can be hidden.

[Rich G.]

Re: 1966 L79 Temp gauge/sender observations.

Apologies if I'm stating the obvious or repeating something already stated (I haven't read the article)...

Years ago, I corrected the gauge reading in a '70 Chevelle by adding a rheostat to the sending unit lead and adjusting it as needed. Once dialed in, I replaced it with the appropriate resistor. They're small and can be hidden.