Coolant Temp Gauge Error: Causes & Fixes

Hi,

I'm looking for information on how to bypass the plateau function.
Coolant temperature gauge for the old TRW speedometer in a Volkswagen Golf 3.

I'm annoyed that it doesn't show the actual temperature.
displayed!

If I remember correctly, in modes like 'test mode,' this 'filtering' is applied.
declined - I would have preferred it to be permanent!

Ultimately, I want to display the (correct) oil temperature on the screen.
to obtain. In that case, I would prefer a simple solution or modification.
Prefer a design consisting of a DAC and controller with an almost inverse plateau characteristic.

If anyone has any information or knows anything (for example, about changes to the speedometer's EEPROM), please let me know.
(No problem), I would be happy if he could post it here.

Best regards,
Rudi

Hi,

I would also be interested.

However, I believe the "inverse plateau function" is not feasible because you cannot create an input signal that would cause the needle to remain at a specific value, such as 80 degrees (in a mathematical sense, the inverse is no longer a function because it would have an infinite slope).

Best regards,

January.

Hello Jan,

If that is indeed programmed as a jump, then some values will be difficult to determine.
My hope (for this - unpleasant -) type of implementation would be that the ramps are programmed...
and that there is a voltage associated with every pointer connection.

Before I do anything in that direction, I'm going to run some tests to characterize its behavior.

Best regards,
Rudi

Rudi wrote:

My hope (for this - unpleasant -) type of implementation is that ramps were programmed and that there is a voltage associated with each pointer setting.

I probably have to temper your hopes as well. With the Polo, the gauge barely moves at all during a change of approximately 20,000 kilometers, and it was the same with the AFN Golf, if I recall correctly.
That is to say, at a pointer reading of 90, there is a whole range of possible voltage values...

If you want to eliminate this effect using the input characteristic curve, you need to make a corresponding upward jump in the input voltage when it reaches the "plateau threshold."

If you want to eliminate this effect using the input characteristic curve, you must make a corresponding upward jump in the input voltage when reaching the 'plateau threshold'.

Okay. That's why a controller and a digital-to-analog converter (or...
a programmable potentiometer) is required. This allows you to assign virtually any value to the input signals.
Initial values are assigned.

Best regards,
Rudi

Hi,

If it's an extremely steep ramp with a wide, flat plateau, you might be lucky. But I don't think so, because then the needle would have to jump suddenly to 90 instead of gradually increasing, so you wouldn't really see the movement while driving.

I think it's more likely that they have some internal mechanisms and, in addition, a very strong electronic damping system, which ultimately causes the slow movement of the needle.

Best regards,

January.

Yes, yes... give me some encouragement .

I believe that certain input parameters are associated with fixed output parameters. And the 'magic' happens when, within a temperature range of 80 to 100 degrees, the control signal is set to precisely 90 degrees.
Rudi, what's stopping you from directly controlling the stepper motors and bypassing their internal control mechanisms? If you're planning to install a controller anyway, it's only a relatively small step further. And you'll have more freedom and won't have to perform any complicated maneuvers.

Jan6K wrote:

I think it's more likely that they have some internal inconsistencies and, in addition, a very strong electronic damping system, which ultimately causes the slow movement of the needle.

When I once replaced the coolant temperature sensor in my Polo with a potentiometer, I could clearly see that the relationship between temperature and display reading was "statically skewed."

Around 50°, the needle lifts off the stop (and the display still shows the correct value). From about 60°, it starts to advance more and more, until it reaches the 90 mark at a real temperature of 75°.

It remains stationary there and doesn't really start to increase until 110°C, but then it quickly catches up with a rising actual temperature, only lagging behind by about 2°C at 120°C. Above that, the display is quite accurate all the way up to its maximum value of 130°C.

If you quickly turn the potentiometer, the needle will move across the entire scale in about 3 seconds (?), which means it's definitely not heavily damped.

@lappen:

Steppers... in principle, there's nothing against it, but it would involve more effort, and I already...
Ideally, I'm looking for a solution that doesn't require a controller.

@Ulf:

then it seems like you can adjust all the positions... I'll try connecting a potentiometer myself .


Best regards,
Rudi

In the service manual for the T5, I saw the characteristic curve of the display that was installed there.

It aligns with what Ulf described.

Therefore, it should be possible to compensate for the plateau function using two inverse functions on the linear subintervals and a jump across the plateau.

Okay, please provide the German text you would like me to translate into English. I will only provide the translation.

Hi Martin,

Thank you for the pictures - I haven't seen the T5 SSP before .

Therefore, one should be able to compensate for the plateau function using two inverse functions on the linear subintervals and a jump across the plateau.

Okay, here's the translation:

'Or simply provide a table with the DAC control values and the corresponding sensor values.'
(or what evolved from the ADC), as an index.
(it might cost a little storage space, but it works great ).


Best regards,
Rudi

Hi,

Looking at the curves, it should be possible to create a relatively simple analogue circuit based on the sensor characteristic curve, which could then be simply inserted between the sensor and its wiring.

At least in cars that have a sensor for the instrument cluster (KI) and one for the transmission control module (MSG), this shouldn't have any further negative effects.

Best regards,

January.