Magnetics design & reed switches - lost article

It probably always will. Mechanical switches do bounce.

Better would be a small piece of soft iron to transfer the magnetic field to the reed and a slightly higher magnet position.

Does it have to be magnetic? Breaking a light beam would be cleaner.

Do a search on reed switch application notes. Lots to choose from. I think Hamlin had a nice application note many years ago.

Dan

That one seems the right application for linear hall sensors.

Those are the most widely available and cheap I'm aware of.

It should be easy to accomplish through the above kind of sensor coupled to a comparator, or some microcontroller analog inputs if you are reading an array of sensors.

Rodwell schrieb:

Hello,

I would try some other sensors instead, a photoelectric sensor for reflected light or a proximity sensor.

Bye

I just tried the key words "reed switch" at and found 9,380 hits with several of the first listing looking like what you wanted. does the name MEDER ring a bell?

Reeds are inherently nasty, bouncy, twangy, unreliable little horrors. Can't you find a better way to do it? Capacitance, AC magnetics, optical, Hall effect?

Do you realize that a reed switch closes when there is enough field *parallel* with the reeds? If the magnet is oriented at a right angle to the reed switch, I would *expect* it to be actuated twice on each pass. And there would be contact bounce on top of that.

You'd probably be better off with a Hall effect switch.

Jeroen Belleman

I would use soft iron to form a U shape with a gap about the same length as the magnet with the reed switch forming the bottom of the U. Possibly the reed switch connections themselves can be bent and used.

This will form a flux guide which will reduce the horizontal distance over which the magnet acts. A smaller and/or weaker magnet might help too with a smaller gap.

Of course, the magnet is attracted to the iron, so if it's a pendulum which is free to swing then the swing may be unduly affected.

| | Soft iron or bent legs

------

Cheers

It's a pretty horrible idea. Reeds normally bounce and chatter.

Mercury-wetted reeds don't bounce - the mercury film damps the oscillations of the reeds when they collide and come apart.

But even in mercury-wetted reed relays the reeds are still ferro-magnetic and thus attract the magnet.

As has been mentioned, Hall-effect sensors don't bounce and aren't usually ferro-magnetic.

Horrible thermal emf's.

you

There WAS an improvement that has been made banned and now illegal: mercury wetted reeds.

Thanks Robert, no it's not by Meder. It was a "cookbook" style article. I've found hundreds of general reed app notes but none cover steering / blocking of the magnet field.

you

Legacy issue in an existing design....... :)

It's only actuating once on a normal pass. It's mechanical bounce causing the problems.

the reeds when they collide and come apart.

thus attract the magnet.

ferro-magnetic.

Contact bounce isn't an issue, a cap sorts that out. The normal reed operating period is >2 seconds.

Legacy design unfortunately.

Thanks Syd - that's what I'd been imagining may work - good to have some reinforcement for the idea. Hopefully we'll try it out today.

On May 15, 5:49 pm, Rodwell wrote:

et

d text -

Rodwell,

It might be that simply changing the orientation of the magnet would solve your problem. The reed responds to magnetic field strength. The reed is insensitive to the polarity of the magnetic field; it would respond in the same manner to the south pole of a magnet as the north pole of the magnet if the field strength at those poles was the same. The symbol [N###S] which Syd Rumpo used to represent the magnet suggests that the magnet is a bar magnet suspended horizontally over the reed switch. The field in the vicinity of a magnet is strongest at each of its two poles, north and south. The field away from either pole would be weaker than it would be nearer to either pole. If the magnet and its orientation are as shown by Syd, then as the magnet swings over the reed, it is conceivable that the reed is being closed and opened twice; once by each pole. First one pole passes close enough to the reed to close it, then that pole passes by and the reed opens; then the other pole passes the reed causing it to close and open again. If that is the case, then simply changing the orientation of the magnet such that only one of its poles, either the north one or the south one, passes close enough to the reed to cause it to close it. Varying the distance from the magnet pole to the reed would give you some control over the pendulum arc through which the reed remains closed. Putting the magnet farther away from the reed would shorten the arc through which the reed remains closed; putting the magnet closer to the reed would lengthen the arc through which the reed remains closed.

Beyond that, attaching a piece of magnetic steel (i. e., "soft" steel) to each end of the magnet might also solve your problem. Each piece of metal would form a new magnet pole at its unattached end over which you would have some control by changing the size and shape of the metal pieces. By adjusting the size and shape of the metal pieces, you could bring the two new magnet poles close together such that they would no longer cause separate actuations of the reed.

I hope you succeed.

gnet

ly

tre

gth

bly

elp

ted text -

In my previous message the incomplete sentence that reads, "If that is the case, then simply changing the orientation of the magnet such that only one of its poles, either the north one or the south one, passes close enough to the reed to cause it to close it." should have been, "If that is the case, then simply changing the orientation of the magnet such that only one of its poles, either the north one or the south one, passes close enough to the reed to cause it to close might solve your problem."