Detecting when an attached mains device is switched-on

Jan 09, 2012 20 Replies

The discussion on detecting the presence of mains made me think again about ways to detect when a mains device has been switched-on. An application would be controlling a dust extract system, for example, when a router is switched-on. Ideas thus far include sensing across a pair of back-to-back diodes in the neutral, winding a few turns around a reed switch, or making a current transformer. Has anyone tried anything else?


Following events over the Xmas break, I'd certainly like to know when my cess-pit pump is running.

And (a different problem) when the level in the tank reaches, er, significance.

on 09/01/2012, NoSpam supposed :

I've managed to detect current flowing in a mains cable, by winding a few turns of one of the supply cable around a transformer and rectifying the output - that is easy enough, but there are 13amp adaptors sold for computers, where if you turn the computer on, it also switches its other sockets on. Made so the computer can power up other devices such as printers etc.. They are not expensive and are designed to save energy.

And they're crap. Well, the one I bought was - similar problems reported by another poster here - it steadily got less and less sensitive until it no longer worked at all. I threw it away.

Thats probably reasonably good..especially if you wand a few turns of the neutral the other way round as well.

The problem with the off-the-peg gizzmos is that the ones I've seen don't have a run-on, and for woodworking extract it would be good to have the slave output keep going for 30-60 secs after the control device was switched-off.

A current transformer is probably the simplest way to do it (with the output constantly re-triggering a 30-60 second 555 timer that controls a triac or relay).

Maybe a faulty one - I have had two working here fine for several years.

This reminds me of the chap who drowned in a cess pit - the funeral was quite quick because he had already been interred.

Detecting water levels is easy, particulary dirty water levels. Justa couple of probes fixed at the required height and detect when a current strarts to flow between them. Built this sort of thing with justa couple of transistors when I was ald but these days it seems to need a PIC or at the very least some form of IC. B-)

That is quite complicated though whats wrong with a float switch like you find on a submersiable pumps?

formatting link
long cable, mains rated. Bound to be other similar things about.

When I was involved in the water industry, lots of years ago, we used a device which was just a sensitive low voltage relay in series with probes.

That reminds me...

When I was a kid and we used to go caravanning, I made a water level detector for the "aquaroll" water tank. This had a pipe straightener to inset the hose into - basically a vertical bit of plastic pipe with a cap on to that screwed into the filler spout.

Detectors were 5 bits of thin stainless rod bent into rings to be a tight fit on the pipe at even spacings along the pipe. Plus an extra one at the bottom as a common electrode.

Hacksawed a slot in the end of each ring to take a bit of fine stainess wire that was sleeved. Wire was crimped into the slot by crushing with a big vice.

Other end silver soldered to multicore copper wire and joint potted inside bit of plastic pipe with loads of araldite. DIN plug on far end to plug into socket under floor.

Sensing circuits were a couple of resistors per electrode and a gate in a Schmitt trigger (hex package, one IC) which had enough oomph to drive an led in a bar stack directly.

IIRC sensing current was a nominal 0.1mA in typical tap water. Proved utterly reliable and no calibration needed.

After years of occasional use, there was no discernable corosion on the stainless electrodes.

easy way to remote sense tank stuff is an oil watchman.

Google it. Ultrasonic should do the job.

I had to address this problem with a pond I have - it has a very slow leak and requires a gallon or so (in 1000) every couple of days. I tried the simple dc twin probe solution and found I got plating on one of them. I then did an ac solution and that had too many volts on it and killed frogs !

I now have an electronic one that does use one simple IC in a remote box but allows me the bonus of a period of over run to allow a small overflow into a damp garden area.

Ping me if you want more details..

Rob

Usually those are ac devices, as any lengthy period of immersion with dc erodes one contact.

SteveW

I installed such when I wired in the new pump at Xmas.

Oh, nothing. That's what I intend to do. I just need the Round Tuit.

Low ohm shunt, current sensing chip and galvanic isolated output (e.g. opto-coupler). It could all be powered from the mains.

This is probably overkill:

formatting link

By "current sensing chip" do you mean a voltage comparator across a shunt? If so, why would a series resistor be better than a couple of back-to-back diodes?

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required