1992 Astra 2000L 8v SRi estate - temperature gauge reads low, new thermostat ensures 92degC should be 1/2 full scale meter but meter only reads 1/4 FS =82degC.
So either meter faulty, sender unit is faulty or the voltage regulator (associated with the temp gauge and fuel gauge) is faulty.
Can anyone tel me the specified VR output volts? either when connected to the sender unit or disconnected.
Also where is VR located?
|
with a 100 Ohm resistor between the disconnected blue temperature sensor lead and earth, the temperature gauge should read full scale.
On this engine, removing the sensor is the best way to bleed that air from the system when refilling; so the sensor will have been in and out a few times since 1992. It's the cheapest and easiest part of the system to replace, so if the test above works OK, I would begin by putting a new sensor in.
If the gauge has read like this since 1992, then I strongly suspect that nothing is wrong - they really aren't too accurate.
But, if the gauge has suddenly changed its behaviour, then you should be concerned.
The voltage regulator is an electronic device, screwed to the back of the instrument panel. If this part failed, you would see odd readings from your fuel gauge too, so I doubt there's anything wrong there.
Remember, dashboard mounted devices are all cheap and nasty - the absolute values they report are not to be trusted, but, sudden changes are important, and the accompanying fault, worth chasing down.
Number_Cruncher
|
Thanks, number Cruncher that all makes sense. There has been no sudden change in behaviour, though the incorrect low reading did mask a faulty thermostat for quite sometime..
I will try the 100 ohm resistor, any idea of resistance value for 1/2 full scale?
|
>>incorrect low reading did mask a faulty thermostat..
If your gauge hasn't changed its position, then I doubt your old thermostat really was faulty.
>>any idea of resistance value for 1/2 full scale?
No, the 100 Ohm at full scale was the only standard test as described by Vauxhall.
There isn't a 1/2 scale test because the meters simply aren't that accurate.
On Chevettes, for example, there used to be a warranty operation where if a customer complained of a consistently off reading temperature gauge, you simply warmed the car up, and tweaked the meter to read in the middle!
Thinking some more; is the point where the sensor screws in to the thrmostat housing properly earthed. It is possible that there is lots of corrosion and or gasket sealant gunk that is stopping the circuit being completed.
Number_Cruncher
|
The saga continues...I measured the Open circuit voltage at the temp sender= 11.16vdc with no ac content. This voltage drops as the coolant temp increases to around 6.4v at thermostat opening temperature,92C. At which point( viz., what should be 1/2 fullscale but on mine = 82C) the temp sender resistence =100ohms
Thus a 100ohm test = 1/2 full scale??
A s/c to earth wangs the meter over past full scale (110C)with the measdured current around 130mA.
I am of the conclusion that the temp gauge (meter)is faulty reading much too low.
I can live with it or put a resistor of appropriate value in parallel with the T/S to earth.
WBD
|
When you checked the resistance of the sensor, did you check between the terminal of the sensor and the earth terminal on the battery, or simply across the sensor?
The sensor should not have been at 100 Ohms at only 92C, its resistance should have been larger.
The meter behaves correctly when you short to earth, it should read beyond full scale deflection at that point.
I would suggest getting hold of a 100 Ohm resistor, and connect it to earth at the battery, and see what you get (There are a number of interfaces and connections between the thermostat housing and battery earth).
Number_Cruncher
|
|
Do not consider leaving a resistor is parallel with the sensor to fool your meter as the value does not change with temp on on cald start the car and the ECU will think is warm and will not initialise the choke and raised idle. Also do not s/c the gauge to earth as you can pop the gause doing this as a noral temp sensor does not go below 60 Ohms. a Nominal vlaue at 90 degress is around 220/210 ohms. Regards Peter
|
I agree with your advice Peter - there is, however, a seperate temp sensor for the injection. The instrument sensor, with the single blue wire is only used for the gauge.
Putting an extra resistor in parallel isn't a good idea.
Number_Cruncher
|
|
|
|
|
I think you are making the mistake that the gauge is very accurate in your Astra. I've owned a similar vintage Cavalier from new and the temp gauge goes to about between halfway between the top of the blue and the 90c mark in normal running -probably about a quarter deflection as you describe. It only goes up to the 90c mark (halfway) when the fan is about to kick in when stuck in traffic. When the stat was faulty it wouldn't get out of the blue area.
|
Gentlemen
Thanks for all the replies.
I measured the T/S resistance to the cylinder head immediately adjacent to the T/S, its a good point to measure to the battery negative terminal as a bad contact in this line could cause the low gauge reading I'm experiencing.
I appreciate that the engine ECU uses a different temp sensor, its just as well.
I just can't believe the the temp gauge reading can be this far out without a fault. The fan has its own seperate temp sensor/switch in the radiator which kicks in around 100C. If you can't trust your temp gauge and the fan fails the first indication of over heating will be steam, etc.,!
As to the complication of the thermosat that I have replaced I now think that the new one is of a slightly higher temperature at 92C or that the old original (1992) had 'drifted down' in its actuation temp?
Will measure the T/S resistence to Battery -ve and report back
regards
WBD
|
>>I measured the T/S resistance to the cylinder head immediately adjacent to the T/S
That would include the interfaces that I would be most concerned about. IIRC, the temp sensor is on the on the thermostat housing itself, so there are three joints;
sensor - thermostat housing
thermostat housing - "elbow" pipe
"elbow" pipe - cylinder head
which I would have been suspicious of, but you have already included in your test.
>>I just can't believe the the temp gauge reading can be this far out without a fault.
Oh, I can! In the 10 years I had my Nova 1.5TD, the temperature gauge only went above 1/4 twice!. Once in a beastly traffic jam on the scorching M1, and once going over Cam Fell in the Yorkshire Dales. I just got used to where the gauge normally sat, and only concerned myself if it moved elsewhere.
Looking back up the thread, you say that you don't get full scale deflection on the temp gauge with 100 Ohm - When you fill up with fuel, does the fuel gauge read full? If so, then you probably do have a problem with the gauge itself. If not, then look at the voltage regulator on the back of the instrument panel.
Some Vauxhalls did have adjustment screws behind the gauge panel that enabled "calibration" in the case of customer complaint - I don't know if yours will have them.
Number_Cruncher
|
|
Just remember many makers just use blue,white and red with no degrees shown;if it's not boiling,don't worry.
|
The Fuel gauge works OK.
I put a 99ohm resistor in place of the T/S and the temp gauge read 1/4FS (80C). Thus the gauge is off or requires a non-standard sender, unlikely. I guess I'll just have to live with it.
End of Story, thanks for all your support.
Funny, though my tachometer also reads low (10 -20%) but that's another story...
Regards
WBD
|
|
|
|