Corsa Temp Gauge stuck at 400F

[BruceWeeks]

Just bought '66 Corsa 140 and the temp gauge always says 400F. How do I diagnose where the issue lies?

If it helps, the tach can by jumpy at times also, reading high at times by about 30%.

Thanks for any leads on getting this going.

[BruceWeeks]

So far I'm thinking to check the wire at the thermistor to be sure it is connected.

Then ground the wire and see if the gauge reading changes when I key on.

If it does change, feed it 12V and see if the gauge goes the other way at key on.

If it doesn't change at all, verify the continuity of the wire from thermistor to gauge.

If continuous, and the gauge does not change, probably in the gauge.

If continuous and the gauge does change, probably in the thermistor. In that case - NUTZ! Hard to find another thermistor.

Other things to look for?

Thanx,

[bbodie52]

Since you mentioned that the tachometer is also misbehaving, I would start by checking the instrument panel grounding strap attachment point on your Corsa instrument panel. Without a good chassis ground the instruments can get pretty flaky.

Left-click on the image to enlarge...

1965-66 Corvair Instrument Panel Detail (Corsa)

[BruceWeeks]

Thanks Brad,

I get under the dash this weekend and report back.

If that doesn't work, how about the other plan of disconnecting the thermistor and alternately grounding or putting 12V to the wire?

[Scott V]

grounding the wire or leaveing it open are good tests. i dont think i would go w/12 volts as it may damage the gauge.

you could rig up a pot to simulate the thermistor & that way you could test the gauge & the wiring.

or...........ohm test the thermistor - cold it is about 5500 ohm. 200f - around 1200 & near 300f about 400 ohm.

-Scott V.

[bbodie52]

The voltage for the cylinder head temperature circuit in the Corsa COMES from the gauge — you would not want to apply voltage from the thermister to the gauge. In fact, if you disconnect the wire at the thermistor and turn the ignition key ON, you should be able to measure 12V DC coming from the gauge in the instrument panel between that disconnected wire and GROUND, using a multimeter. The thermistor in the left cylinder head is a variable resistor that varies the resistance to ground (the engine head) based on heat. The varying resistance to ground impacts current flow through the gauge, which changes indicator needle deflection that is calibrated in the desired temperature range based on the known behavior of the thermistor.

The fact that your gauge does not deflect either way and seems "stuck" at the midpoint would appear to indicate an open circuit. Either no voltage is getting to the gauge from the power circuit (loose wire, broken connection), or there is no ground being applied through the thermistor in the cylinder head (broken or loose wire, corroded or dirty connector near the engine, bad thermistor). Based on the schematic diagram, I don't believe a loose ground connection at the Corsa instrument panel would affect the temperature gauge, although it might impact the tachometer, fuel gauge, etc. I would still recommend checking the tachometer wiring connections and the Corsa instrument panel ground, but the temperature gauge may be a separate problem with either a faulty gauge, a faulty thermistor, or faulty wiring.

I wrote this post on the Corvair Forum about ten months ago. It was written discussing the temperature gauge in a 1963 Spyder, but it generally applies to the Corsa as well...

There are two cylinder head temperature sensors on your Spyder. One is a large temperature overheat switch that simply triggers an overheat warning light on the instrument panel. The other is a thermistor that looks like a small spark plug. This is the unit associated with your cylinder head temperature gauge. The thermistor is simply a variable resistor that changes resistance to ground characteristics as it gets hotter. Its characteristics can be seen at the following website:

[url]http://rad_davis.sent.com/Thermistor.html[/url]

Cylinder Head Temperature Gauge Thermistor Characteristics

This plot was done by Herb Berkman in 1974 and published once in the CORSA West newsletter, The Westwind. My copyright notice on the plot below is a declaration of my authorship of the electronic image (which is significantly different from the gray, photocopy-of-a-photocopy I recieved). In no way should this be construed to be a claim of authorship of the plot itself.

In an email to me, Herb offered the following advice about this plot:

This data was never intended to be an accurate recreation of the thermistor
characteristics. It was only to get a baseline for one unit.
Anyone who looks at this curve needs to be advised:
a. It is from a sample of only one part.
b. The part that was tested was 10 years old and had been in use for about
60,000 miles.
c. The data was taken using instruments that had not been calibrated.
If this test was run on a greater number of parts, the data would probably vary
considerably. There might even be additional variation between new and used
parts.


Left-click on the image to enlarge...

As shown in the graph, the range of the temperature gauge starts at 200 degrees (1000 ohms) and has a maximum reading capability of 600 degrees (approximately 100 ohms at the thermistor). As the engine heats up, the resistance presented by the thermistor to ground decreases, which increases the gauge needle deflection -- reflecting an increase in engine temperature.

There are two electrical connections on the cylinder head temperature gauge. One is a simple 12 VDC power connection to power the gauge. The other is the wire connection between the gauge and the thermistor in the cylinder head. Assuming that the gauge is functional and has power (indicated by the initial needle deflection you described when you turn on the key), the first thing you need to do is check for electrical continuity between the gauge and the thermistor. This can be done by disconnecting the thermistor wire from the gauge at the instrument panel, and using a multimeter (set to measure resistance in ohms). You should be able to ground one test lead of the multimeter to a chassis ground, and touch the other test lead to the thermister wire with a resistance reading that reflects the cold reading that would be expected according to the above referenced graph (> 1000 ohms). (If you run the engine, this resistance reading should change as the engine heats up). If you measure an "Open" condition with your multimeter (no continuity) you will need to trace the wire and check for a damaged lead or corroded/loose connection at any connector between the instrument panel and the thermistor.

I found at least one article on the Internet that addresses troubleshooting the Corvair Corsa cylinder head temperature gauge circuit...

How to Troubleshoot your Corvair Corsa Engine Temperature

http://www.examiner.com/article/how-to- ... emperature

The attached book sampler extracted from the recently published Performance Corvairs: How to Hotrod the Corvair Engine and Chassis by Seth Emerson and Bill Fisher includes some interesting material on measuring cylinder head temperature and oil temperature.

[BruceWeeks]

Wow! A wealth of knowledge in this post!

Then consensus here at work from the sensor team was similar. The thermistor grounds the gage like a domelight or back-up lamps do in some cars of the era. Center of the gage may be a default position if signal is not available.

At least the idiot lamp lights during start (bulb check) so I know it works if the switch is working, then I would at least I would know of an overheat situation. Have to check that at some point too.

Looking for voltage at the thermistor end of the wire is a good and easy start. Checking the continuity and then resistance of the thermistor from the chart is another relatively easy step (I found that post earlier, it is a good reference - Thank you very much.)

Continuity from thermistor to the gage is a bit more difficult, but grounding the thermistor end of the wire and seeing if the gage changes at key-on would say something about that. The potentiometer would then be a step to determine if the gage works at least close to the way it is supposed to.

For some reason the Examiner article won't open, IE just goes in circles forever. But I get the gist from above.

Hopefully if something is faulty, it is either a wiring connection (easily fixed once found) or the gage as I have a far better chance of finding a working gage than a thermistor!

Failing that, the sensor guys here said bring them the chart and they'd see if they could find "something" that might work.

[bbodie52]

Remember, out of a total production of 1,786,243 Corvairs manufactured between 1960 and 1969 model years, only 78,924 (4.42%) were manufactured with cylinder head temperature gauges (Spyders and Corsas). All of the remainder survived primarily based on a fan belt warning light (GEN) and a TEMP warning light (I've yet to see that light come on because of an overheat condition). Basically, if the fan belt is driving the cooling fan, the engine does not overheat. Even when I've driven my 140 hp Corsa across the southern California desert in the middle of a very hot summer day at a constant 70-75 mph, fully loaded with four passengers and a lot of luggage, the temperature gauge never indicated excessive engine heat. Only the passengers were overheated!

If 12V DC is present on the wire coming from the gauge, when measured at the connector in the engine compartment area, this would tend to indicate good power to the gauge itself, and good electrical continuity all the way to the engine compartment connector. On the other side, a cold engine should show approximately 5000 ohms resistance between the wire connector and ground. This resistance reading should be roughly 300-500 ohms with the engine fully warned up. If the thermistor is providing appropriate resistance readings to ground, and voltage is detected coming from the gauge on the other side of the connector, the gauge should indicate an appropriate reading when the harness is reconnected (unless ther gauge itself is defective). If no voltage is present coming from the gauge, the wiring powering the gauge at the instrument panel must be checked, and if 12V DC is found going into the gauge and coming out of the gauge, the thermistor wire itself may have a broken lead. Of course, if the thermistor does not display appropriate resistance readings, it is likely faulty. In any case, a physical inspection and some testing with a multimeter should point to the cause of the malfunction.

Let us know what you find.

[BruceWeeks]

Yes, I will. One of tomorrow's primary chores is to root cause the gage. Plus check the brake linings, put on the new wheels and tires, the second pair of halogen headlights, pull the driver's bucket for recovering and root cause the reason the signlas don't self cancel (the lever is floppy in the column, is it simply loose or is the switch or cancel cam broken? - They still work just don't cancel.) And of course it's the tele column so a new switch is expensive. Hopefully just loose and maybe a cancel cam for $12 will fix it.

[BruceWeeks]

I did some diagnosis this weekend. The thermistor has resistance, but not sure I got a good reading as the only battery I had on hand for my VOM expired this month. plus I don't have a calibration resistor. But at least it is not open or shorted.

Then I checked the wire at the thermistor - no voltage coming to it. Tried to figure it out but haven't located the problem yet.

I checked fuses just to be sure, none blown. From the wiring diagram, it is hard to tell just from where the power to the gage comes. I did notice a loose electrical connector under the dash, chased it in the manual and appears the warning buzzer is MIA. There is an unoccupied 3 prong harness plug with only two wires in it hanging down above the dimmer switch. But the wires from memory don't match the wiring diagram for the buzzer. Both had color stripes on them )though I should have written them down, from the pictorial it seems to be in the right place.)

From the wiring diagram, it does not appear that the buzzer would interrupt the gage as it only goes to the idiot light side. So more chasing by seeing if power is at the gage or not, then along the length of the wire.

The turn signal switch appears complete and unbroken just really gummed up. I'll clean all the old dry grease out and replace the cancel plate as a preventative measure. Plan on using a synthetic Teflon lube that is compatable with plastics.

[BruceWeeks]

This weekend, trying to figure out how to access the back of the gage to see if it has power, I simply pulled the cluster. Low and behold the plug was missing off the gage and the wires had been cut off the harness. So now I know the "how" it doesn't work, but not the "why."

Bought a 5K potentiometer to test the gage. If that seems to be OK (<200F at 5000 ohms, ~ 400F at 250 ohms, ~600 at 90 ohms) then I'll reconnect it and see if I get 12V at the thermistor. If not, it'll be chade the wire to find the break (another cut?).

I'm begining to realize it's probably a bad thermistor with an abnormally high resistance: not open or shorted. A new battery in the VOM still gace a reading near 50K instead of 5.5K ohms.

I'd better make sure the snap switch is working as back-up in case I can't get the temp gage working for suimmer cruise season.

Any suggestions on testing and diagnosing the snap switch? Haven't tried to research that yet.

[BruceWeeks]

I set the 5K pot up and tested the gage. As soon as you crack the pot off 5K (~4800 ohms) the gage starts going to the high end and stays there. Move to about 4500 ohms and it goes way past the peg to virtually horizontal. I'm thinking the the gage is bad.

And since the thermistor seems to be reading very high also, it is probably bad too.

Opinions?

Also, how do I check the idiot light switch? Bulb check at key-up shows bulb works OK. Just want to be sure the switch will turn it on if I get in trouble until the gage can be permanently fixed, one way or the other.