Toyota Forum banner

Trying to determine if 89 Camry is overheating, temperature gauge jumps around

1 reading
4.1K views 13 replies 4 participants last post by  white90dx  
#1 · (Edited)
Im trying to figure out if my 1989 Camry is overheating and if it is circulating coolant correctly.

My temperature gauge has had problems for a while. The needle will typically sit all the way to the left (cold). If it begins to get hot, it'll jump to a vertical position (12 o clock). More recently though it has started to behave even more oddly. The temp gauge will jump to 12 o clock immediately after starting up the car and then it woll abruptly drop back down to the left (cold) a few minutes of driving. It might pop back up to vertical if I'm sitting in traffic for a bit.

About a month ago the car ran low on coolant. The gauge popping up to 12 o clock prompted me to check the reservoir. The reservoir was almost empty. I filled up the reservoir and it hasn't been low in the month since I've filled it. I'm not noticing any big coolant spills on the ground so I'm not sure where the coolant was lost in the first place.

Between losing coolant and my temperature gauge acting funny, I'm concerned that the car might be overheating and I'm not aware of it. The fan up front never seems to kick on -- it has in the past but I can't recall the last time it did -- that could be a separate issue.

Is there a way that I can:

1. Fix the temperature gauge so provides a reliable reading.

2. Confirm that coolant is properly circulating through the system.

3. Determine that the car is running at proper temps or is overheating (given that the temp gauge is suspect).

I'm guessing that #1 might be too difficult for me to figure out, but maybe #2 and #3 are within my technical ability.

Thank you for any tips!
 
#2 ·
Sounds like something is wrong with the gauge, the wiring, or possibly the temperature sensor feeding it. My guess is either a ground wire or the sensor, maybe both.

The gauge should read just under halfway to the hot side when the engine is warmed up to operating temperature. The gauges on Toyotas of this era are also fairly slow, not the most accurate, and not super sensitive.

If you need to confirm coolant temperatures in the interim to make sure the engine is not overheating, I recommend purchasing an IR thermometer, like this: https://www.amazon.com/KIZEN-Infrared-Thermometer-Gun-Humans/dp/B07VSHR9M6
Point it at the thermostat housing while the engine is running and see what temperatures you get. It might also help to use it on the hot side radiator hose or some other key location on the engine block. I'd need a diagram of the coolant flow path to verify which locations would give the best and most accurate reading.
The thermometer should read 180F (or 82.2C), and if that is the case then the cooling system is functioning as intended. Though again, this assumes we measure it in the correct place.
Above that, and the engine is overheating, even if only slightly above (though not dangerous/panic level until 10-20F higher). Below that and the engine is running too cool, which is safer but still not ideal. Generally it is a thermostat issue preventing or allowing too much coolant flow; but on old cars like ours, the radiators can become clogged with scale and no longer work as effectively. I would lean more towards thermostat first, but it is something to consider.
 
owns 1994 Toyota Estima
#3 ·
and not super sensitive.
By design. It will sit in that 'just below half' position from nearly warmed up to just about to overheat...

I would suspect the gauge sender and wiring around that location first. The sender is one of the single-wire items on the water outlet neck (where the upper radiator hose connects to the engine).

-Charlie
 
#4 ·
You can idle the engine with the radiator cap off and you can watch the water. When cold itll just kinda sit there but when tge thermostat opens youll see the water move. Side benefit is thatll help let any trapped air out. You can also feel the hose temp warm up once the thermo opens.
Large swings of the gauge id suspect wiring/sensor. Its hard for coolant temperature to swing that fast.
 
#5 · (Edited)
@Eastman51 I took your advice and bought that IR thermometer. Had some mixed results with it. I let the car idle and heat up for ~5 minutes. The IR gun was against the engine block was reading temps all over the place -- anywhere from 160 degrees Fahrenheit to 300 degrees Fahrenheit, but on average somewhere around 220 Fahrenheit. Really hot. Hitting the IR gun on the top radiator hose gave much more consistent values -- ranging from 150 degrees to 185 degrees but with the average read somewhere around 178 degrees. Do you think that the IR thermometer gun is going to be more accurate on the black radiator hose than it would be on the metallic engine block?

I replaced the oil, oil filter, coolant, flushed the system with distilled water and re-filled the coolant. I replaced the water temp sensor because the old sensor was suspect -- the plastic attachment for the sensor would literally spin in place 360 degrees. Took the car out for a drive and it seemed to run well. Good power, no funny sounds, no overt signs of overheating. I parked it and let it idle for a few minutes -- the radiator fan turned on after a while -- the first time I had heard it in a long time -- so the water temp sensor seems to be doing it's job now.

@white90dx Still no luck with the gauge reading for coolant temp. I thought I had found the problem -- I noticed that the water temp sender (the sensor with the single wire) that screws directly into the engine block had become unscrewed by about a third of an inch. The screw threads of the temp sender were exposed and some oil was slowly leaking from it. I tightened it down til the sender head was nearly flush with the engine block. I turned on the car and the temp gauge immediately went vertical, but just to the left a little bit -- which (if I understand correctly) is exactly where the gauge is supposed to be for normal operating temps -- but after 5 minutes of driving the gauge suddenly dropped to the left (cold) side again. It dropped instantly, one second it's vertical, the next second it's all the way to the left. Fiddling with the single wire connected to the sender didn't seem to have an impact. Maybe the sender sensor itself is bad? Do you happen know what the part number is for the replacement or what the official name of that sensor is?

@root Thanks for the tip, I followed your advice and was able to confirm that the thermostat is functioning properly and that coolant starts moving properly through the system once it's warmed up. Confirmed that the radiator hoses heat up rapidly once the thermostat is open.
 
#6 ·
@Eastman51 I took your advice and bought that IR thermometer. Had some mixed results with it. I let the car idle and heat up for ~5 minutes. The IR gun was against the engine block was reading temps all over the place -- anywhere from 160 degrees Fahrenheit to 300 degrees Fahrenheit, but on average somewhere around 220 Fahrenheit. Really hot. Hitting the IR gun on the top radiator hose gave much more consistent values -- ranging from 150 degrees to 185 degrees but with the average read somewhere around 178 degrees. Do you think that the IR thermometer gun is going to be more accurate on the black radiator hose than it would be on the metallic engine block?

I replaced the oil, oil filter, coolant, flushed the system with distilled water and re-filled the coolant. I replaced the water temp sensor because the old sensor was suspect -- the plastic attachment for the sensor would literally spin in place 360 degrees. Took the car out for a drive and it seemed to run well. Good power, no funny sounds, no overt signs of overheating. I parked it and let it idle for a few minutes -- the radiator fan turned on after a while -- the first time I had heard it in a long time -- so the water temp sensor seems to be doing it's job now.
IR thermometers seem to be more accurate the closer you can get it to the place you are trying to measure (but not physically touching it); additionally, they are less accurate with painted surfaces and can be less accurate on metal depending on the material. On my RX-7's rotary engine for example, the very top center of the center iron is not very accurate; but using it on the rear housing offset as far to the drivers side of the car as possible (right below the oil filter housing) yields extremely accurate numbers. That location on the rear housing should be equalized in temperature to the thermostat as it is the exit of coolant from the engine.

Ideally, if you are measuring temperatures on the engine block, you get the thermometer very close to your target. Generally you should be looking at the thermostat housing or the engine's outlet passage (like that location on the rear housing of a rotary engine). The upper hose on the driver's side of the car for these Camry is the outlet passage. You measure the hot side to make sure coolant is not too hot as it leaves the engine, which would indicate overheating; and you measure the thermostat housing to give an indication that it is operating correctly; at least in theory they should be the same. You can measure the hoses instead of the metal if the numbers seem wacky, but you want to be as close to the engine side as possible.
I'm not sure what temperature the radiator fan is supposed to turn on or off at on these cars, but as long as it is cycling while at idle and temperatures are consistent then it should be functioning correctly.

The screw threads of the temp sender were exposed and some oil was slowly leaking from it. I tightened it down til the sender head was nearly flush with the engine block. I turned on the car and the temp gauge immediately went vertical, but just to the left a little bit -- which (if I understand correctly) is exactly where the gauge is supposed to be for normal operating temps
Single wire sensor with oil leaking does not sound like a water temp sensor. That might actually be the oil pressure switch, which would turn the oil light on if pressure got too low. I could be wrong, but water and oil don't mix...... and when they do its bad news. You could probably unplug the sensor and ground the wire to see if the oil light turns on. Single wire temp/pressure switch generally works by connecting the wire to ground when the fluid meets the set requirements; and once there's a ground, the warning light comes on. For a gauge, the sensor has a resistance sweep where different resistance values indicate different temperatures and cause the gauge to reflect that value; by measuring the resistance value between the sensor wire and a ground, you can determine if the sensor has failed. While it would help to know what the resistance values of the sensor should be, it should still be somewhat obvious the sensor has failed if the resistance values are ridiculous and there is a high disparity between hot and cold readings.

And yes, just a tad under dead center is where it should stay once warmed up.
 
owns 1994 Toyota Estima
#7 ·
@Eastman51 Hmmm, that's really interesting, I wonder if my temp sender isn't hooked up at all...

Here's a photo of that area in my 89 Camry:

Image


The connector that I'm holding in my hand it just free-floating -- hasn't been connected to anything for years.

The sensor on the left with the rubber housing and single wire coming out of it -- that's the sensor that was partially unscrewed and oil was slowly leaking from. I tightened it down.

The sensor with the green adapter up top -- that's what (I think) is the coolant temp sensor. I replaced that sensor with a new one. That sensor feeds into the ECU, right?

Where's the coolant temp sensor that feeds into the dashboard gauge?

Thanks!
 
#13 ·
The connector that I'm holding in my hand it just free-floating -- hasn't been connected to anything for years.

The sensor on the left with the rubber housing and single wire coming out of it -- that's the sensor that was partially unscrewed and oil was slowly leaking from. I tightened it down.
That mystery connector is also not hooked to anything on my car either. It's probably for a sensor on a different trim level like Charlie mentioned.

The rubber booted sensor definitely looks like it'd be the oil pressure switch to me. You unknowingly fixed an oil leak!

@white90dx Thanks! I never would have found it. The cable was detached from the sender and obscured from view. The sender itself was completely covered in oil and grit -- didn't look like a sender at all.

Now when I turn on the car from cold, the gauge rises up to 12 o'clock (maybe a little past it). As the car warms up, the gauge gradually moves all the way into the red -- but the car is not overheating, it's running at standard temp. It seems like the temperature gauge is shifted to the hot side -- starts at 12 o'clock from cold instead of 9 o'clock and it settles in the red at 3 o'clock instead of vertical at 12 o'clock.

Does that sound like a bad sender to you?

Edit: Or (and this might be a stupid idea) can I just take the clear cover off the instrument panel and set the needle to the correct position when the engine is warmed and running at normal temp? It seems like the sensor is accurately measuring the rise in temp from cold to warm, but it starts too high to begin with.
That is sounding like a bad sender to me. Perhaps someone unplugged it in the past due to the whacky reading and no actual overheating.....

I would definitely test the resistance of the sender before proceeding further. It is not a very hard test to do.
 
owns 1994 Toyota Estima
#11 · (Edited)
@white90dx Thanks! I never would have found it. The cable was detached from the sender and obscured from view. The sender itself was completely covered in oil and grit -- didn't look like a sender at all.

I cleaned it off and re-attached the cable.

Now when I turn on the car from cold, the gauge rises up to 12 o'clock (maybe a little past it). As the car warms up, the gauge gradually moves all the way into the red -- but the car is not overheating, it's running at standard temp. It seems like the temperature gauge is shifted to the hot side -- starts at 12 o'clock from cold instead of 9 o'clock and it settles in the red at 3 o'clock instead of vertical at 12 o'clock.

Does that sound like a bad sender to you?

Edit: Or (and this might be a stupid idea) can I just take the clear cover off the instrument panel and set the needle to the correct position when the engine is warmed and running at normal temp? It seems like the sensor is accurately measuring the rise in temp from cold to warm, but it starts too high to begin with.
 
#12 ·
If you want to do it right, test the sender and the wiring before moving needles.

Here are a few values that might help with the tests:
Resistance : Temp *F
226 : 122
153 : 140
26.4 : 239

I spent a bit of time searching for more than that (the above numbers are what is in the wiring diagrams) but couldn't find anything worthwhile. Anyway, take a few readings of resistance and water outlet temp and report back...

If you do need to move the needle, you'll want to put a known resistance in there to 'calibrate' the gauge when you move the needle.

-Charlie