Temperature Sender AJ16

[davidr]

So how to test and compare them? Well they work by varying their resistance across the temperature range so I thought I'd test them using a digital thermometer and a multimeter.
So this is the reading on the thermometer

And this was the resistance reading for the original sensor of 1343 ohms

And an identical temperature reading for the new sensor

And the resistance was only 916 ohms

So the result of my experiment was........................failure!!
Anyone like to buy a slightly used temperature sensor for an X300?

[Tintin]

This link gives the possible explanation?

https://www.xj40.com/viewtopic.php?t=996579

Robbie

[davidr]

Too cold for gardening today (or so I told 'er indoors) so a bit of messing about in the workshop was permitted.
I took the original temperature sensor from my car and the DAC2583 I bought in. I wired them both up to my multimeter set to measure resistance and then heated up the sensor using a hot air gun.
Using a digital thermometer I then charted the resistance of each sensor as it cooled from 100 to 40 degrees C. This was repeated 4 times for each sensor and the results averaged to get a more accurate reading. The chart below is the average reading for each sensor.

The resistance is up the side (in ohms) and the temperature along the bottom (in degrees Celcius).
As you can see there is a steady relationship between the 2 sensors with the DAC2583 having a consistently lower resistance reading.
I did some sums and the difference within the key working temperatures (ie 60-100 degrees) between the averages of my readings is between 44 and 46 ohms at each reading.
So I have invested £2.74 in a 45 ohm resistor from Hong Kong via eBay and then I'll try making up a link lead to see if that gives the right readings on my car.
By the way, nothing wrong with my sensor - just curious!

[RichardSEL]

I've changed both my CTS sensors for new from main dealer (Denso made in Japan so must be OK ) on both my XJ8 -- both gave consistently better MPG after the change >30MPG mixed mode, light foot, long run
So interesting results you're getting -- seems from what you're observing that it's not that old sensors get "tired" and don't change their resistance. But time taken to change resistivity for a given temperature change increases?

Is it possible, now that you've got a jig set up, to see whether that's true -- old sensor vs new?

[davidr]

Richard -what this shows is the relationship between the 2 sensors.
The original DAC11079 was taken from my car (and it works fine) just for the sake of this experiment. As it is NLA and I get frequent questions I obtained the DAC2583 to see if that could be made to work.
The chart in my post has the resistance values for the 2 sensors across the working temperature range. Original sensor (DAC11079) in orange and the substitute (DAC2583) in blue.
The resistance of the DAC2583 is consistently around 45ohms less than the original. This is just because they are made to a different spec and not because the older one has degraded. The original behaves exactly as I would expect and the needle sits just below N on the gauge. The DAC2583 would make the gauge read hot when the engine is perfectly fine.
I'm not sure what you're getting at with time to change resistivity? Are you wondering how quickly the resistance changes with temperature? As temperature change would be much more gradual than my test (ie the coolant warms up much slower than me putting a hot air gun on the sensor) I think I'd say that I didn't notice any difference between their reactions although I can't think how to test that. They both warmed up and cooled down much quicker than they would in the car and neither appeared to have any lag in the readings.

[RichardSEL]

That's not to say that my style of CTS would be anything like what you're testing -- looks exactly like yours though. Just curious to know in what way my old CTSs had deteriorated over the lifetime of the vehicles compared to the new that've now both been installed (Sport now sold).
I still have one old one left in my box of "may come in handy" bits

[poprivet]

Interesting results and useful info. I imagine David's sender is for the temperature gauge? not the one that sends info to the engine computer, just wondering if Richard is describing that one.

Regarding deteriation, these are standard thermisters packaged conveniently. I would think they fail either open or closed circuit, I doubt that the resistance range would change, happy to be proved wrong though.

[davidr]

Yes - these are temperature senders for the instrument pack.
The sensor for the engine management sits alongside it in the thermostat housing and is a different part number and thankfully still available.

[davidr]

Update on my little experiment.
I got a 45ohm resistor and made up a link lead as below. The idea of this is to put it between the sensor and the original connector to adjust the resistance and so give the right signal to the gauge.

I replaced my original, correct and working sensor with the CAC2583 and let the car come up to temperature. There is no fault with my cooling system and the gauge normally comes up to just the start of N on the dial which is just as it should be.
Here is the gauge with just the DAC2583 and no link lead.

As you can see, it reads too high making it look like the car is too hot. In fact it was spot which was confirmed by my infra-red thermometer.
I then put my link lead in.

Now the gauge shows too cool. Again I stress, the car was actually spot on correct temperature.

Conclusions? I'm on the right path but I haven't got the resistance right yet! So a couple more orders have been placed for resistors from my eBay friends in China!

[davidr]

I have to admit that I'd forgotten all about this thread and my little experiment until I was poking about in my workshop a few days ago and discovered the package of resistors that I'd ordered from eBay months ago!
I made up leads with various resistance values and eventually settled on one of 20ohms which has the needle just below the N when the car is fully warmed up. Just right!
Here's the resistor and lead made up ready to hide under the injection rail cover.

and here it is in place

Here's the temperature gauge without my link lead reading too high

and here it is with my link lead reading just right!

[poprivet]

Some resistors can change their value due to ambient temperature, I wonder if you may need to mount it somewhere less affected by the engine temperature?

The power resistor you have is designed to disipate heat but could also suffer from heat gain. I would think a simple metal film resistor would do the same job and could be hidden within shrink wrap easily.

[ptjs1]

David,

Tks for sharing this. I'm still slightly confused as to why a 20ohm resistor brings the gauge to original reading, when the difference between the two sensors seemed to be 40-46ohms.

As I'll eventually need to change my sensor one day, I've been intrigued by all the various discussions on these sensors on all the Jaguar forums.

Cheers

Paul

[davidr]

Hi Paul - I wondered that too but I put it down to the crude nature of my experiment. I was heating up the sensors quite quickly using a hot air gun and they were cooling down quickly too. I was measuring the external temperature of the sensor which wouldn't be the same as inside.
If I had the patience, I would have been slowly heating the sensor in water and letting it cool down slowly so the heat distribution was even.
Lack of good laboratory practice I'm afraid!
Al - we'll see how it performs in practice. It's still in the car and will have a run up to Kemble at the weekend