Mfts/guage Testing

[twoqu]

Why don't you use a linear potentiometer to create the desired resistance values for testing purposes. Just adjust the variable resistor to the value you want using a DMM.

If the MFTS is showing the expected results out of circuit using the boling water method you describe but not functioning when reinstalled in the car I would suspect a problem with either pin corrosion on the connector or a wire break somewhere.

I'll look into that. If thay are affordable then it would be a useful addition to the toolbox. The frustrating thing has been the fact that the VAG 1301 setting was not the actual resistance value. Now we have a better idea of what the values equate to then it should make troubleshooting that bit easier but the subsitution of a known resistance.

I think the MFTS died reinstalling it as it no longer has any resistance which it should have when measuring between earth and the T terminal. I have proved the circuit from the connector to the gauge works with the 50 ohm resistor.

[twoqu]

*IF* you take the time to re-read the info I provided *and* mesh it with the Bentley VW1301 info link that you provided (now "503"'d ), and reverse your thinking, I believe that you have your answer as to what resistances should be produced by the MFTS to generate a given temp signal.

Again had you read my first post you will see that indeed I had already deduced what the MFTS should output. Thats not the issue. The issue is whether you can measure the resistance of the MFTS out of the vehicle by the method I have outlined. It does also allow, now we know the resistance of the MFTS to troubleshoot it directly rather than eliminate it by subsitution. If the engine is hot we know that the MFTS T terminal should show a resistance of 50 ohms or less between it and earth. Its unlikely we would get a much different reading as the engine would need to be about 110C to get the 38 ohms output.
If you don't know the answer to a question thats fine! If you want to take an educated guess thats ok as well! There is no obligation to contribute to a thread but at least read the thread first and then contribute, not see the title and post everything you have archived about it.
In my opinion the answer is yes, but I am willing to take into account anybody elses opinion as to whether its possible or not and why!

[UrS4boy]

IF I had a spare working MFTS, hot water (to say 90 C), a thermometer and a DMM, would you be interested in the ohm readings from the "T" pin as the temps cooled to 50 C?

[twoqu]

IF I had a spare working MFTS, hot water (to say 90 C), a thermometer and a DMM, would you be interested in the ohm readings from the "T" pin as the temps cooled to 50 C?

Well it would answer the question I posed in the first post!
All I am trying to establish is whether it indicates a working MFTS. If you immerse the bulb part of the MFTS into recently boiled water and you can read a resistance from the T terminal and the MFTS in the 50 ohms range or less then we know that might be useful to save fitting a sensor that is already dead.

[UrS4boy]

Well it would answer the question I posed in the first post!

When I posted just above, in my mind, my spare AAN had a MFTS sitting there on an empty water manifold. I just went out to the garage to remove the MFTS and bring it inside to test. With 29 mm socket in hand, imagine my surprise when I found no water manifold or MFTS. (*NOW* I remember, I didn't take that when I got the engine 5 years ago)

Sorry.

[twoqu]

Thanks to the help of another member I can report the following.
Basically it does seem possible to test the MFTS BUT the resistances obtained do not seem to accurately mirror what we have deduced. This could be caused by the age of the sensor, the measuring equipment, the mad scientist or indeed the testing conditions. So what did I find?
As expected the sensor put out a resistance that varied according to the temperature change. This was measured between the T terminal and the actual body of the sensor. It did not appear to matter where the other probe was touching the sensor as long as it was on the brass body. I tried a few times with different results:

90C 145 Ohms (90C 140 Ohms)
80C 175 Ohms (80C 175 Ohms)
70C 195 Ohms (70C 235 Ohms)
60C 270 Ohms (60C 285 Ohms)
50C 345 Ohms (50C 355 Ohms)

I used a homebrew type thermometer and a good quality MM. Pics below illustrate a room temperature measurement and a hotter one (@60C+) before I attempted to measure temperature more accurately!
So at least that does prove that there should be some kind of reading. Will need to replace the sensor with this one to see what happen. Will be more careful this time using a socket and torque wrench!

[UrS4boy]

90C 145 Ohms (90C 140 Ohms)
80C 175 Ohms (80C 175 Ohms)
70C 195 Ohms (70C 235 Ohms)
60C 270 Ohms (60C 285 Ohms)
50C 345 Ohms (50C 355 Ohms)

This is what your results look like (fairly consistent except around 70 C)

[twoqu]

Just tried something again. It clearly takes a while for the reading to stabilise due the heat sink effect through the sensor housing to the NTC resistor. For example I have watched the reading drop (indicating a temp increase) from 250 Ohms to 220 Ohms over a period whilst the water temperature would have been slowly decreasing!
So not really a definitive test, unlike say a switch opening at a certain temperature, but perhaps more an indication that the MFTS is not completely dead.
Both of the original VDO 4 pins one have not shown any signs of life although they were already suspect!
Would have been happier to see readings nearer 50 ohms at 90C although at the lower end of the scale its not that far out assuming 288 Ohms is an indicated 50C.

[UrS4boy]

A proper test would have say 6 good MFTSs, one each with its sensor tip in a controlled temp water bath, e.g. one at 50 , 60 , 70 , 80, 90, 100 C and so on. The sensors would be left touching the water bath for say an hour to stabilize the temp in the sensor. Then, while the tip was still emersed, the resistance would be measured.

Alternatively, six dry ovens at those temps could be used.

Doing it with one sensor in a rising (or falling) temp water bath doesn't reflect that fact that the body of the MFTS is retaining (or sucking) heat and the actual resistance is not 100% reflective of he water temp at the moment of the resistance measurement.

[twoqu]

Best I could do in the kitchen!
It would be interesting to make a comparsion with a new one to see how much a used one has degraded over the years. This one must be at least 17 years old. If this one does not function that might be on the cards!
The factory calibration at the top end of the scale would seem to equate to the 110C temperature on the gauge. Its not so clear at the bottom due to the gradiants being bunched together. If its nearer 60C than 50C then actually its not that far out.

[UrS4boy]

Best I could do in the kitchen!

I wasn't being critical, just trying to find a way to explain the different results of the two tests. It is more info than we had before so "good on you".

[twoqu]

If the MFTS sensor works, the real world test would be on the vehicle. As long as I could piggyback of the T terminal and earth then it should be possible to use either vagcom for the engine temp or the climate sensor to get within a degree. That would rule out most of the variables other than the accuracy of those two sensors.
But really I was just trying to determine if this type of testing proves the temperature function of the MFTS is ok out of the vehicle. We will see!

[twoqu]

I carefully fitted this sensor today and measured the resistance between the T terminal and the housing the sensor is screwed into. I watched as the resistance gradually got lower until the vehicle reached operating temperature which coreresponded to a reading of 130 Ohms. I then hooked the wiring up to the MFTS and waited for the gauge to respond - nothing!
Bit confused as I knew the gauge circuit was good and the suspect MFTS had been replaced with what appeared to be a functioning one. The one I removed does appear to be dead. On retesting the resistance I used another part of the engine as the earth and noticed that the resistance was more than double than that measured at the housing. I temporarily used a jump lead connected to the MFTS body and the engine and was rewarded with the gauge indicating just below the 90C mark. I cross referenced with the climate Ch 51 to see it at 88C. So I do now have a functioning accurate temperature gauge but only when its earthed directly to the engine with a jump lead!
Somehow the alloy housing the MFTS is in (this a C4) is not earthed adequately to the engine. The only thing I can think of is that the bolts securing it must be coated in a sealant which is unnecessary as the bolt holes are blind I think. When earthed with the lead you can get the same resistance between the T terminal and the housing as anywhere else on the engine. I was also unable to piggyback the signal off the T terminal when connected to the wiring which seemed strange.
So all I can say is that I measured a reading of 130 Ohms at the T terminal and the housing at 88C with the gauge needle just below the 90C segment.

[UrS4boy]

With your permission, I think I want to add all your info to my MFTS post for future reference for here and other C4 UrS and S2/RS2 forums.

Is that okay?

[twoqu]

Sure!
Just cherry pick the facts rather than my rambling.
Its evident this vehicle has an unusual problem preventing the gauge from reading accurately. The reading of 130 ohms is close to that observed with the kitchen method, allowing for a bit of heat sink. I was expecting something closer to 50 ohms for 90C.
Anyway thats as scientific as I can make it for the time being. Might treat it to a new sensor if I can cure the earthing issue!