Coolant Expansion Tank - Testing tank

[coupe]

Skoda Fabia Mk1. Many threads on this and the answer on later tanks is to just replace (light flashing when starting - switch on and off when warm  - goes away) as you can't get at the prongs but does anyone know how to test the tank sensor itself? Assume it's not a simple on off but more likely resistance. If I put a meter across the contacts on the tank what can I test for to see if it's working correctly or not. Just curious really.

[sepulchrave]

You can't, its capacitive, but to be certain you can short the two contacts on the plug with a piece of wire and the light should stay off permanently.

[coupe]

Interesting so if the bottle goes from empty to full you get no change in current or resistance? Capacitive would still give a reading somewhere or am I missing it completely. 

 

Thanks for the response. 

[sepulchrave]

1 hour ago, coupe said:

Interesting so if the bottle goes from empty to full you get no change in current or resistance? Capacitive would still give a reading somewhere or am I missing it completely. 

 

Thanks for the response. 

 

Yes.

 

I'm not saying it's impossible but I am saying it's pointless since a new bottle with a new sensor is a tenner.

We're not trying to reinvent the wheel here.

[coupe]

Thanks. Just curious. Must be the third bottle albeit over 80k miles since we had it. New one has sorted it. 

 

Thanks

[KenONeill]

@coupe - That sensor is supposed to detect low coolant level and/or high temperature. Shorting it out doesn't seem like a great idea.

[Tech1e]

Its a make or break connection

 

Two metal prongs are emmersed in the coolant. Coolant is conductive so the circuit is complete. When the level drops the circuit is no longer complete and the light comes on.

 

So disconnect the sensor = light comes on

Short the two terminals together = light goes off.

[TMB]

6 hours ago, KenONeill said:

@coupe - That sensor is supposed to detect low coolant level and/or high temperature. Shorting it out doesn't seem like a great idea.

 

It only detects level not temperature.

[KenONeill]

4 minutes ago, TMB said:

 

It only detects level not temperature.

OK, so why does it throw a flashing red light?

[TMB]

41 minutes ago, KenONeill said:

OK, so why does it throw a flashing red light?

 

The light serves as both low level and temp warning. The temp is detected by the CTS.

[Breezy_Pete]

The prongs in the expansion tank are described in the circuit diagrams as the "coolant shortage indicator sender" (only, not over temperature) and given the circuit symbol of a variable resistor.

At some time in the past I tried to do a bit of reverse engineering on the circuit in the instruments module (clocks) but the only thing I can find or remember is that the last thing in the circuit on the way to the blue/red wire is a capacitor, suggesting an ac rather than dc sense current. Probably to avoid electrolysis effects. The prongs are too far apart and too lacking in area to be a capacitive sensor, I think. The other wire goes straight to earth; "sender earth in engine compartment loom".

 

 I seem to remember experimenting with a variable resistor on the end of the loom plug and establishing a criterion value above which it would trigger a warning, but I can't find that just now either!. Week off next week, I may be able to dig out some stuff if it rains.

 

 

[Vigneshwaran]

Any findings? I can see that the reservoir is common for most of vag cars. if we can find the resistance value we can test before proceeding to replacement. It would be really helpful since I'm also facing the same issue of low coolant warning when my reservoir is in correct level. Any help is appreciated. :)

[Breezy_Pete]

I've just brought home from work a book of 'old school' resistors with wire leads. I will aim to find a number tomorrow. ☺️

[SuperbTWM]

17 minutes ago, Wino said:

I've just brought home from work a book of 'old school' resistors with wire leads. I will aim to find a number tomorrow. ☺️

 

Sounds like a decade box would be ideal for this purpose

[Breezy_Pete]

Have you got one?

I have a vague memory of it being around 5k Ohms, but swapping a few leaded resistors around across the loom connector contacts should be easy enough.

[SuperbTWM]

I have a few at work yea. 

Like you say, if you know the ball park it shouldn't be too much of a faff

[sepulchrave]

I think you need to measure the voltage at the sensor plug, it's likely to be a small AC voltage rather than DC which would suggest the sensor is capacitive rather than resistive.

The sensor doesn't measure coolant level, simply whether the prongs are immersed or not, immersed would read a small capacitance, dry would read open circuit, this is why shorting the plug puts the light off, any condition other than open circuit will put the light out.

[SuperbTWM]

Usually a capacitance probe wouldn't have 2 bare electrodes, one would be have to be insulated from the liquid by having a PTFE or other plastic coating otherwise you wouldn't be able to measure the capacitance. 

 

I only work on industrial applications though so it could be using witchcraft for all I know

[sepulchrave]

Witchcraft it is then.

 

The fact remains that any condition other than open circuit will put the light out so the means of detection is really irrelevant, spend tenbux and replace the tank.

[Breezy_Pete]

I will do the experiment this morning. Expecting different facts.

[sepulchrave]

The resistivity of water is very high but varies depending on solution and will change significantly during the lifetime of the coolant, if all you want to know is if the probes are immersed or not then capacitance makes more sense than resistance.

[Breezy_Pete]

56k* and under - no warning

62k and over - warning

 

Tested on 9N Polo and 6Y Fabia.

 

* abbreviation of kilohm, thousands of ohms, so threshold is somewhere between 56000 and 62000 ohms.

Edited 16 May, 2020 by Wino

[mrgf]

Vigneshwaran re-raised this issue so I kind of wondered, at the age of the vehicle, perhaps the probes in the bottle have just corroded/scaled up over time and as mentioned by Sepulchrave, replacing the bottle is quite an inexpensive option. Not a difficult procedure, either so barely worth thinking of. After ten years, treat it like an oil/air/cabin filter and replace it. Even a used one will not be worth it as often, used is more costly then new and will probably be half way through its life in any case!

 

I noticed Vignashwaran is based in India... Of corse, parts may be costlier there, comparatively but even if you find the root of the cause, you will in all likelihood, need to change the bottle. I also know, from many students, etc that many places such as India, Bangladesh, Pakistan, etc, use plain water in the car. (Both in coolant and screen wash). This will accelerate any wear to many parts of the cooling system so is really a false economy.

[Breezy_Pete]

It seems to me that some people would rather just fire the parts cannon willy-nilly, without any desire to understand the system they are attempting to repair.  From what @Vigneshwaran posted, I'd suggest he isn't in that category.

[sepulchrave]

Since you can't repair the probes in your existing tank if they have corroded and are giving a false reading then your choices are to short the plug by putting in a wire link or replace the tank.

So all this pedantry is really getting us nowhere, incidentally what was the voltage reading across the plug?