View Full Version : Temperature gauge behaviour
02-09-2014, 07:04 PM
No doubt like most of you I drive with one eye on my temperature gauge especially in Sydney's hot weather.
My gauge has always read a bit high - a bit over half way when driving. But on a hot day (around 90 degrees) it climbs to about three quarters of the way and even higher if idling in traffic. There's no evidence of boiling over and I never need to top up the expansion tank. The engine has done about 350 miles since the rebuild.
I have an accurate CVR reading dead on 6 volts (thanks to Dave) and I've checked that the sender is working. A check with an IR temperature gun shows between 180 and 195 depending on where it's pointed ie. the expansion tank, radiator, hoses etc. This all points to the gauge, doesn't it? I've read the Tech Section on adjusting the temperature gauge. Has anyone had a good result from doing this?
02-09-2014, 08:29 PM
The way I do it is to make note of how much distance there is between the needle and the C when the key is in the on position and the engine is cold. I then take the wire off the sender and ground the wire to the block. The needle will go all the way over to H. I then make note of how far the needle is from the H. If the distance between the needle and letters is about the same then the gauge is most likely calibrated. I usually find that the needle is too far one way or the other and it will have to be adjusted. At best the gauge is only roughly accurate. I would trust your IR gun at the block close to the sender for the real temperature.
02-10-2014, 12:31 AM
Disconnecting the sender wire is the same as turning your key off. No current goes through the gauge.
John, doesn't the needle drop below 'cold' when the key is off?
The correct way to measure is by pulling the wire off the sending unit so you can measure the current with a meter. Put the meter scale on milliamps and put one lead on the wire with the other wire on the sending unit. Make sure you use the 'amps' plug for your lead.
When the engine is cold, the sending unit should resist 100 ohms. (volts divided by resistance = amps) So 6/100=.06amps (or 60 milliamps). At this current the needle is below 'Cold'. HOT should show 6/10=.6amps (or 600 milliamps). These values depend on the health of your sending unit and gauge so they may be approximate. At least you know what to expect.
When your gun shows 180 degrees at the thermostat, measure your gauge current, and notice your gauge needle. Again, the same current flows through the gauge and the sending unit. Now, you can determine which is causing a false reading.
We can accurately make the gauge needle rest at any point. Adding a simple resistor to that sending unit wire will lower the scale on your gauge. Is that really what you want?
Temperatures fluxuate dramatically in your engine. The thermostat is mechanical device that opens at a preset value IF it works right. Thermostats are slow-moving in both directions. I have seen them stick in the closed position, which normally causes an overheating situation.
My Escort had a broken thermostat (yes, the brass arms broke half-in-two) which never allowed the engine to come up to heat, and caused terrible gas mileage. My temp gauge was 'cold' all the time. So, thermostats can break, either open or closed.
Your TEMP gauge should show what is really happening in your cooling system. Sometimes the engine does run a little hot which causes the temp gauge to raise, but still be in the normal range. Your gauges should report the truth. - Dave
02-10-2014, 10:59 AM
Disconnecting the sender wire is the same as turning your key off. No current goes through the gauge. - Dave
I said to disconnect the wire and ground it. That's the same as the temperature sender circuit being totally closed. I know it's not the most scientific way but then the gauge isn't the most precise of gauges either. It's always worked for me as good as any other method.
02-10-2014, 12:30 PM
I'm getting senile in my old age, John. Sorry. - Dave
02-10-2014, 01:00 PM
I'm getting senile in my old age, John. Sorry. - Dave
I can relate to that. I just call it selective memory.
02-12-2014, 06:07 PM
I'll give what you suggest a try. I'm considering installing an accurate digital temperature gauge somewhere inconspicuous as I really like to know what's happening under the bonnet (hood).
02-12-2014, 07:52 PM
Harbor Freight sells a nice infrared thermometer gun for US$38.00. Just point, shoot, and it reports the surface temp:
There are no cords so you can take it anywhere, and point at anything from a distance so you don't hurt yourself. I use mine at the radiator for detecting plugged cores or at the exhaust manifold to show 'dead' cylinders.
Your gauges are designed to give you an average reading. For example, if you're climbing a mountain the gas gauge should be slow to move. Same holds true for the temperature gauge. The car's thermostat moves slowly but does a good job of cooling. The temp gauge should not show the immediate changes. On a cold day, you may feel heat before the needle starts moving. When things 'normalize' the gauges show a good average. (which is really what you want). - Dave
02-25-2014, 08:23 PM
I finally had a chance to test my temperature gauge as per your instructions. The needle rests on the C when the sender wire is off but goes quite a bit past the H when the wire is grounded. This makes sense as the reading with an IR temperature gun pointed near the sender says it is around 160F when the needle is more than half way. Running temperature on a hot day is between 175 and 190 but the needle is close to H.
I didn't get a chance to test the resistance as I'd left my multimeter on and the battery was dead and I didn't have a replacement 9 volt replacement at the time.
I've read up on calibrating the gauge in the Technical Resources section and Dave also mentioned using a small resistor on the sender wire. What value resistor would be required?
02-26-2014, 12:13 AM
Ok, let's do this again. Your dash meters are current devices. They have heater elements inside. When more current passes, the needle goes higher.
Let's measure current with a meter. Put the meter scale on milliamps and put one prod on the detached wire with the other prod on the sending unit. Make sure you use the 'amps' plug on your meter for your red meter prod.
When the engine is cold, the sending unit should resist 100 ohms. (volts divided by resistance = amps) So 6/100=.06amps (or 60 milliamps). At this current the needle is below 'Cold' or 'T'emp.
What current are you reading when the needle just starts to move?
"HOT" or TEM'P' should show: 6/10=.6amps (or 6-volts divided by 10-ohms = .6-amps or 600 milliamps) on your meter.
What current are you reading when the engine is up to temperature?
These values depend on the health of your sending unit and gauge so they may be approximate. At least you know what to expect.
I cannot recommend a resistor value without knowing your current readings. If your sending unit proves to be bad, remedy the root cause and change it. If your sending unit proves to be good, we can talk about gauge adjustments or the addition of a resistor. - Dave
vBulletin® v3.6.8, Copyright ©2000-2018, Jelsoft Enterprises Ltd.