Non-contact coolant alarm - false alarm?

[oibaf]

Weird one here folks...

I have a non-contact coolant alarm fitted to the header tank. It's one of the ones which sticks to the outside of the tank, no drilling (I'd thought it was a Haydyn one but it's not). I used a bit of fingernail glue to stick it to the tank (didn't have superglue to hand). It'd been fine for over a week.

I was driving yesterday and the alarm went off. I ws able to stop the engine immediately because I was in an industrial estate so not in traffic. Fearing the worst, I got out and looked under the van but there's no puddle of coolant. No trace of a leak at all.

I left it for a while, then put the key in the ignition to move the van 50 yards to a parking space. Coolant alarm chirped when I turned the van on (self test) but didn't actually go off.

Leaving the van for a couple of hours, it was cool enough for me to carefully open the header tank and dip the dipstick in. The coolant level was perfect. Same as it's been when checked at home a day before.

I then decided to drive home but the alarm went off again. Couldn't see any issues so decided it was a false alarm and indeed the alarm turned off after a minute and didn't come back on. At no point during any of this time did the temperature gauge actually rise past the usual 11 o'clock position.

Surely a false alarm? But how?

I have this alarm tapped into the 12v feed for the cigarette lighter/door mirrors/radio (fuse . I didn't use an inline fuse.

Checked starter battery after a few hours at home - 12.7 volts. Leisure battery was just 12v though.

Any thoughts? Coolant seems fine so really am thinking it's a bad alarm or bad voltage?

[g8dhe]

A few years ago I had a play with both ultrasonic and capacitive measurement methods of level detection, capacitive weren't good at all possibly due to coatings in the tank, ultrasonic type was better but reliability still wasn't good, I actually used some data loggers to record the echoes over a couple of weeks of use and had some very strange effects that made me discount the sensors / method as not being ideal, hence I took it no further. I see there is now a design out that appears to be catching on but also several people on FB have mentioned problems with it, possibly due to the effects I saw which I think was caused by turbulence in the tank when the coolant was being pushed around. So not convinced myself that its any improvement on Haydn's approach with the screw sensors and measuring the current thru the coolant.

[oibaf]

So you'd reckon it was a false alarm too? I tested the sensor before fitting it, without facing a liquid it would beep furiously, when held to a glass of water the beep would stop as expected.

I don't mind the odd false alarm as long as it works when I really need it to but if there's somethingi can do to mitigate false alarms,all the better.

[g8dhe]

Its difficult to be certain, but there is normally a requirement to "blend" between the materials (impedance matching) between the source and any materials/liquids to get a good return signal from the final bounce return, otherwise you end up with reflections from each change of surface, the very same effect your relying on to detect the coolant level, but which you don't want to interfere with that signal. That said its an assumption, maybe the sensors have improved or there is a matching layer built into those particular sensors which solve that problem, I'm not in a position to tell so I won't criticises the design, merely my own experience from about 5 years or so ago. I was also looking at using them for measuring the LPG level in the gas tanks but again it needed a lot more work than I wanted to put in for what would be a minor gain when working out how much gas was left in a tank!

[oibaf]

That makes sense. I never thought about the LPG level in a tank and how that's measured. Interesting to think about!

[haydn callow]

Don’t be put off from fitting a sensor screw to the tank. 10s of thousand have now been fitted and not a single problem….also the alarms have proved to be extremely reliable over many years..

[teenmal]

So you'd reckon it was a false alarm too? I tested the sensor before fitting it, without facing a liquid it would beep furiously, when held to a glass of water the beep would stop as expected.

I don't mind the odd false alarm as long as it works when I really need it to but if there's somethingi can do to mitigate false alarms,all the better.

If you give Peter a shout he is very helpful ..... https://www.ebay.co.uk/itm/124325809550 ... Sw50xfafGI

[oibaf]

So you'd reckon it was a false alarm too? I tested the sensor before fitting it, without facing a liquid it would beep furiously, when held to a glass of water the beep would stop as expected.

I don't mind the odd false alarm as long as it works when I really need it to but if there's somethingi can do to mitigate false alarms,all the better.

If you give Peter a shout he is very helpful ..... https://www.ebay.co.uk/itm/124325809550 ... Sw50xfafGI

Thanks for that. I actually did and be replied suggesting to change the sensitivity of the sensor.

I have to admit, I've not checked the earth wire I ran nor the crimp joint I did into the live feed.

Funny story - when I first bought that alarm, I thought it was one of Mr Haydyn's ones! I didn't think anyone else did them. I was new to all of this!