Magnetics design & reed switches - lost article

[Rodwell]

Ages ago I had a great .pdf app note type document that covered the
practical implementation of reed switches. It had stuff on biasing with
other magnets & a whole pile of drawings & notes on how sheet metal &
other fabricated parts could be used to control the operation of the
reed. Needless to say I cant find it & google has not been helpful.

Does it ring a bell for anyone? I've got a feeling it was a pretty old
article, it would be great to find it again.




I've got a situation where I need a reed to only operate when the magnet
is directly in front of it & to release when the magnet moves slightly
to either side.


The mechanical arrangement is similar to a pendulum except that the
pendulum is stable at both A and B. It moves from A to B (or B to A)
when an external force is applied. We want the reed to operate as the
pendulum mounted magnet passes the centre, but only right at the centre
+/- 5mm or so.



A<--------------------------------------------->B
|REED|

------OFF------------|-ON-|-------OFF--------- DESIRED OP.

------OFF-----|--------ON--------|----OFF----- ACTUAL OP.



Our problem is that the mechanical parts bounces a bit resulting in
multiple actuations of the reed rather just a single operation.


The reed clears the magnet by ~5mm & the switching "dwell" is quite
sensitive to the gap. We're thinking some simple sheet metal may be
useful to shield the reed either side of centre. Practical?

[Martin Brown]

On 15/05/2013 13:17, Rodwell wrote:

> I've got a situation where I need a reed to only operate when the magnet
> is directly in front of it & to release when the magnet moves slightly
> to either side.
>
>
> The mechanical arrangement is similar to a pendulum except that the
> pendulum is stable at both A and B. It moves from A to B (or B to A)
> when an external force is applied. We want the reed to operate as the
> pendulum mounted magnet passes the centre, but only right at the centre
> +/- 5mm or so.
>
>
>
> A<--------------------------------------------->B
> |REED|
>
> ------OFF------------|-ON-|-------OFF--------- DESIRED OP.
>
> ------OFF-----|--------ON--------|----OFF----- ACTUAL OP.
>
>
>
> Our problem is that the mechanical parts bounces a bit resulting in
> multiple actuations of the reed rather just a single operation.

It probably always will. Mechanical switches do bounce.

> The reed clears the magnet by ~5mm & the switching "dwell" is quite
> sensitive to the gap. We're thinking some simple sheet metal may be
> useful to shield the reed either side of centre. Practical?

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.

--
Regards,
Martin Brown

[dca...@krl.org]

On May 15, 8:17 am, Rodwell <rodw...@non.com> wrote:
> Ages ago I had a great .pdf app note type document that covered the
> practical implementation of reed switches. It had stuff on biasing with
> other magnets & a whole pile of drawings & notes on how sheet metal &
> other fabricated parts could be used to control the operation of the
> reed. Needless to say I cant find it & google has not been helpful.
>
> Does it ring a bell for anyone? I've got a feeling it was a pretty old
> article, it would be great to find it again.
>
>

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


Dan

[asdf]

On Wed, 15 May 2013 20:17:36 +0800, Rodwell wrote:

> Our problem is that the mechanical parts bounces a bit resulting in
> multiple actuations of the reed rather just a single operation.

That one seems the right application for linear hall sensors.

www.allegromicro.com/~/Media/Files/Datasheets/A1301-2-Datasheet.ashx

http://sccatalog.honeywell.com/imc/printfriendly.asp?
FAM=solidstate&PN=SS496B

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

> We want the reed to operate as the pendulum mounted magnet passes
> the centre, but only right at the centre +/- 5mm or so.

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.

[Uwe Hercksen]

Rodwell schrieb:

> The mechanical arrangement is similar to a pendulum except that the
> pendulum is stable at both A and B. It moves from A to B (or B to A)
> when an external force is applied. We want the reed to operate as the
> pendulum mounted magnet passes the centre, but only right at the centre
> +/- 5mm or so.
>
>
>
> A<--------------------------------------------->B
> |REED|
>
> ------OFF------------|-ON-|-------OFF--------- DESIRED OP.
>
> ------OFF-----|--------ON--------|----OFF----- ACTUAL OP.
>

Hello,

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

Bye

[Robert Macy]

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

[John Larkin]

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?


--

John Larkin Highland Technology Inc
www.highlandtechnology.com jlarkin at highlandtechnology dot com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators

[amdx]

Other than that, what's wrong with them? :-)
Mikek

[Jeroen Belleman]

On 2013-05-15 14:17, Rodwell wrote:
> [...]

>
> I've got a situation where I need a reed to only operate when the magnet
> is directly in front of it & to release when the magnet moves slightly
> to either side.
>
>
> The mechanical arrangement is similar to a pendulum except that the
> pendulum is stable at both A and B. It moves from A to B (or B to A)
> when an external force is applied. We want the reed to operate as the
> pendulum mounted magnet passes the centre, but only right at the centre
> +/- 5mm or so.
>
>
>
> A<--------------------------------------------->B
> |REED|
>
> ------OFF------------|-ON-|-------OFF--------- DESIRED OP.
>
> ------OFF-----|--------ON--------|----OFF----- ACTUAL OP.
>
>
>
> Our problem is that the mechanical parts bounces a bit resulting in
> multiple actuations of the reed rather just a single operation.
>
>
> The reed clears the magnet by ~5mm & the switching "dwell" is quite
> sensitive to the gap. We're thinking some simple sheet metal may be
> useful to shield the reed either side of centre. Practical?

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

[Syd Rumpo]

On 15/05/2013 13:17, Rodwell wrote:

<snip>

>
> I've got a situation where I need a reed to only operate when the magnet
> is directly in front of it & to release when the magnet moves slightly
> to either side.
>
>
> The mechanical arrangement is similar to a pendulum except that the
> pendulum is stable at both A and B. It moves from A to B (or B to A)
> when an external force is applied. We want the reed to operate as the
> pendulum mounted magnet passes the centre, but only right at the centre
> +/- 5mm or so.
>
>
>
> A<--------------------------------------------->B
> |REED|
>
> ------OFF------------|-ON-|-------OFF--------- DESIRED OP.
>
> ------OFF-----|--------ON--------|----OFF----- ACTUAL OP.
>
>
>
> Our problem is that the mechanical parts bounces a bit resulting in
> multiple actuations of the reed rather just a single operation.
>
>
> The reed clears the magnet by ~5mm & the switching "dwell" is quite
> sensitive to the gap. We're thinking some simple sheet metal may be
> useful to shield the reed either side of centre. Practical?
>

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.


[N###S] Magnet

| | Soft iron or bent legs
---<REED>---



Cheers
--
Syd

[Bill Sloman]

On Wednesday, 15 May 2013 22:17:36 UTC+10, Rodwell wrote:
> Ages ago I had a great .pdf app note type document that covered the
> practical implementation of reed switches. It had stuff on biasing with
> other magnets & a whole pile of drawings & notes on how sheet metal &
> other fabricated parts could be used to control the operation of the
> reed. Needless to say I cant find it & google has not been helpful.
>
> Does it ring a bell for anyone? I've got a feeling it was a pretty old
> article, it would be great to find it again.
>
> I've got a situation where I need a reed to only operate when the magnet
> is directly in front of it & to release when the magnet moves slightly
> to either side.
>
> The mechanical arrangement is similar to a pendulum except that the
> pendulum is stable at both A and B. It moves from A to B (or B to A)
> when an external force is applied. We want the reed to operate as the
> pendulum mounted magnet passes the centre, but only right at the centre
> +/- 5mm or so.

<snip>

> The reed clears the magnet by ~5mm & the switching "dwell" is quite
> sensitive to the gap. We're thinking some simple sheet metal may be
> useful to shield the reed either side of centre. Practical?

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.

--
Bill Sloman, Sydney

[John Larkin]

Horrible thermal emf's.

[Robert Baer]

John Larkin wrote:
> On Wed, 15 May 2013 20:17:36 +0800, Rodwell<rod...@non.com> wrote:
>
>> Ages ago I had a great .pdf app note type document that covered the
>> practical implementation of reed switches. It had stuff on biasing with
>> other magnets& a whole pile of drawings& notes on how sheet metal&
>> other fabricated parts could be used to control the operation of the

>> reed. Needless to say I cant find it& google has not been helpful.

>>
>> Does it ring a bell for anyone? I've got a feeling it was a pretty old
>> article, it would be great to find it again.
>>
>>
>>
>>
>> I've got a situation where I need a reed to only operate when the magnet

>> is directly in front of it& to release when the magnet moves slightly

>> to either side.
>>
>>
>> The mechanical arrangement is similar to a pendulum except that the
>> pendulum is stable at both A and B. It moves from A to B (or B to A)
>> when an external force is applied. We want the reed to operate as the
>> pendulum mounted magnet passes the centre, but only right at the centre
>> +/- 5mm or so.
>>
>>
>>
>> A<--------------------------------------------->B
>> |REED|
>>
>> ------OFF------------|-ON-|-------OFF--------- DESIRED OP.
>>
>> ------OFF-----|--------ON--------|----OFF----- ACTUAL OP.
>>
>>
>>
>> Our problem is that the mechanical parts bounces a bit resulting in
>> multiple actuations of the reed rather just a single operation.
>>
>>

>> The reed clears the magnet by ~5mm& the switching "dwell" is quite

>> sensitive to the gap. We're thinking some simple sheet metal may be
>> useful to shield the reed either side of centre. Practical?
>>
>>
>>
>
> 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?
>
>

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

[Rodwell]

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.

[Rodwell]

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

[Rodwell]

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

[Rodwell]

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

[Rodwell]

Legacy design unfortunately.

[Rodwell]

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.

[James DePauw]

> reinforcement for the idea. Hopefully we'll try it out today.- Hide quoted text -
>
> - Show quoted 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.

[James DePauw]

> I hope you succeed.- Hide quoted 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."

[Syd Rumpo]

On 16/05/2013 23:22, James DePauw wrote:

<snipped>

> 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.

Sorry, James, that's not even worth arguing with. You don't understand
magnets.

Cheers
--
Syd

[josephkk]

Why is there so much wanting to fix really bad (legacy) designs, but do
not want the changes that a real fix would entail?

?-)

[Bill Sloman]

On Sunday, 19 May 2013 19:16:50 UTC+10, josephkk wrote:
> On Thu, 16 May 2013 08:41:27 +0800, Rodwell <rod...@non.com> wrote:
> >On 15/05/2013 09:55, John Larkin wrote:
> >> On Wed, 15 May 2013 20:17:36 +0800, Rodwell <rod...@non.com> wrote:

<snip>

> >> 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?
> >>
> >Legacy issue in an existing design....... :)
>
> Why is there so much wanting to fix really bad (legacy) designs, but do
> not want the changes that a real fix would entail?

It's a lot less work to document a band-aid fix than "real" fix. Sensible people do both - the band-aid fix keeps the production line running until you can find the effort to document the real fix.

--
Bill Sloman, Sydney

[Robert Macy]

On May 19, 2:16 am, josephkk <joseph_barr...@sbcglobal.net> wrote:

> On Thu, 16 May 2013 08:41:27 +0800, Rodwell <rodw...@non.com> wrote:
> >On 15/05/2013 09:55, John Larkin wrote:

Because, we march 'backwards' into the future.

Or, because the user base is too large to ignore.

[Tom Hoehler]

"Bill Sloman" <bill....@ieee.org> wrote in message
news:ba02dc7a-e6c4-4506...@googlegroups.com...

How about a shielded reed switch triggering a 555? :)
Tom