Multi-way light switching - are there 'electronic' ways to do it?

[tin...@isbd.co.uk]

I know the normal way to provide multi-way light switches is using two
SPCO switches and intermediate switches for however many more ways one
needs.

I want to do the same but using only SP on-off switches because the
shape and size of switch I want to use is only available as single
pole on-off.

Obviously one can do this using a relay controlled by each switch (two
SPCO and the rest DPCOs wired to provide the intermediate switch
action) but I was wondering if there are ICs manufactured for doing
this. If they can control lights (12 volt, an amp or two) directly
then so much the better but I'd be happy enough to add power switching
to a low level device if necessary.

--
Chris Green

[Andrew May]

From a purely electronic point of view if you had a low voltage supply
use a pull-up resistor to the 5v and use the switch to ground other end.
Wire all of these to an exclusive-or gate and use the output to switch a
relay to control whatever load you need.

[NT]

You could just use a digital 4000 or 74LS series IC so that any
operation of any switch changes the output or on/off state. Add a
relay to drive the lights, and a transistor to drive the relay.


NT

[tin...@isbd.co.uk]

I don't think that works does it? Apart from anything else an EOR
gate only has two input, I'm looking for (possibly) three or more.

The EOR approach will work for a two way switched light I agree.

--
Chris Green

[tin...@isbd.co.uk]

What 4000 series or 74LS series chip would provide this logic though?
It can't be done simply for more that two inputs I don't think.

--
Chris Green

[Mike Barnes]

tin...@isbd.co.uk:

You could take a look at Easy Switch (e.g. from TLC). No mains required
at the switch, it operates by battery-powered radio that fits behind
your switch in a deep back box. In "double" mode, changing the position
of any registered switch will toggle the receiver output, which is an
uncommitted SPDT relay. Although it's called "double" mode I imagine it
will work with any number of transmitters.

--
Mike Barnes

[tin...@isbd.co.uk]

It's not *entirely* clear but it does sound as if that would do what I
want - it says the receiver "Learns up to 16 transmitters".

Expensive though!

--
Chris Green

[Andrew Gabriel]

In article <foqh29-...@chris.zbmc.eu>,

tin...@isbd.co.uk writes:
> I know the normal way to provide multi-way light switches is using two
> SPCO switches and intermediate switches for however many more ways one
> needs.
>
> I want to do the same but using only SP on-off switches because the
> shape and size of switch I want to use is only available as single
> pole on-off.

Can you get momentary action types, i.e. with a sprung return?
If so, you can use am impulse relay, which is how it's normally
done in some other countries.

My light switching is mostly done electronicly, so I need lots
of momentary action switches. I often end up inserting the
momentary action springs from other switches, when I find that
I need it in a form that doesn't exist.

--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]

[Harry Bloomfield]

Andrew May formulated on Tuesday :

Much easier with several push buttons instead of a switch - wire in as
many as you want, one pull-up and just one wire around all of the
buttons. First press turns it on, second press (of any button) turns it
off, even use a timer IC to turn it off after a period of time.

--
Regards,
Harry (M1BYT) (L)
http://www.ukradioamateur.co.uk

[jgharston]

tinn wrote:
> I know the normal way to provide multi-way light switches is using two
> SPCO switches and intermediate switches for however many more ways one
> needs.

Points, not ways. Unless you really are asking about controlling a
traffic light.

This is a multi-way switch:

o------
o------
---o----->o------
o------
o------

What you're desribing is multi-/point/ switching, having several
switches as several points (or places). This is done using
*two*-way switches:

o-------------------o
---o----> <----o----
o-------------------o

See. Each switch can only switch two ways. In this example
there also happens to be two points/places. However,

o------o o--------o
---o----> X <----o----
o------o o--------o

is three /position/ switching, it still uses two-/way/
switches. The switches can only switch in two different
ways, A or B. None of the switches can switch to a
third position. One of the two-way switches happens to
be an intermediate switch, but it still can only
switch two different ways.

JGH

[tin...@isbd.co.uk]

That was what I had originally considered but I really can't find any
suitable push button switches for this particular application.

--
Chris Green

[tin...@isbd.co.uk]

Possibly, though I'm now tending towards using relays. DPCO 3 amp
relays which would be perfectly adequate for what I want to do can be
bought for quite a lot less than £1, stick two or three on a piece of
veroboard and hide it away somewhere and the job's done really. It
also allows me to use ridiculously thin wire to the switches which
might be quite an advantage.

--
Chris Green

[tin...@isbd.co.uk]

jgharston <j...@arcade.demon.co.uk> wrote:
> tinn wrote:
> > I know the normal way to provide multi-way light switches is using two
> > SPCO switches and intermediate switches for however many more ways one
> > needs.
>
> Points, not ways. Unless you really are asking about controlling a
> traffic light.
>

Yes, I know that's strictly correct, but "two way switch" in a domestic
setting means what you call multi-point.

--
Chris Green

[Andrew Gabriel]

In article <fabi29-...@chris.zbmc.eu>,

In the US, a "two way switch" is a "three way switch", because it's
got 3 terminals.

[Mike Barnes]

tin...@isbd.co.uk:

FWIW I have a couple used in "single" mode with several receivers each.
Perfectly reliable, probably over 5 years and still on the original
batteries.

--
Mike Barnes

[Frank Erskine]

On Tue, 6 Mar 2012 22:56:48 +0000 (UTC), and...@cucumber.demon.co.uk
(Andrew Gabriel) wrote:

>In article <fabi29-...@chris.zbmc.eu>,
> tin...@isbd.co.uk writes:
>> jgharston <j...@arcade.demon.co.uk> wrote:
>>> tinn wrote:
>>> > I know the normal way to provide multi-way light switches is using two
>>> > SPCO switches and intermediate switches for however many more ways one
>>> > needs.
>>>
>>> Points, not ways. Unless you really are asking about controlling a
>>> traffic light.
>>>
>> Yes, I know that's strictly correct, but "two way switch" in a domestic
>> setting means what you call multi-point.
>
>In the US, a "two way switch" is a "three way switch", because it's
>got 3 terminals.

Presumably a DPDT switch is known in the US as a six way switch... :-)

They really haven't much of a clue.

--
Frank Erskine

[Andrew May]

True, I am not aware of a chip that has a multiple input XOR but
something like a 7486 has four two-input XOR gates that can be cascaded.
Two inputs into the first gate, another two inputs into the second. The
outputs of gates one and two into the inputs of gate three. The output
of gate three will be high if there are an odd number of the four inputs
high which satisfies the OPs requirements.

Output of three could be connected to one of the inputs of four and the
other input connected to another switch if five inputs where required.

[Andrew May]

But doesn't that defeat the original requirement of using a specific
design of switch which is only available as a single pole?

[Mike Barnes]

Andrew May <andrew...@hotmail.com>:

No, it's a two-wire connection. A signal is transmitted on make and on
break.

--
Mike Barnes

[Andrew May]

On 07/03/2012 09:42, Mike Barnes wrote:
> Andrew May<andrew...@hotmail.com>:

>> But doesn't that defeat the original requirement of using a specific
>> design of switch which is only available as a single pole?
>
> No, it's a two-wire connection. A signal is transmitted on make and on
> break.
>

Apologies. I had missed the fact that it can be made to work with any
switch.

[tin...@isbd.co.uk]

Yes, it hadn't initially sunk in (to me) that extra cascaded XORs will
do what I want to any required level. Thanks.

--
Chris Green

[John Rumm]

On 06/03/2012 16:31, tin...@isbd.co.uk wrote:

A pile of bistables...

Use one on each switch to convert it from edge to pulse operation. Then
"or" all of those together and feed a T type bistable to toggle its
output on each pulse.

You would need some debouncing on each switch though, or you will get
loads of edges per switch operation!


--
Cheers,

John.

/=================================================================\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\=================================================================/

[Andrew Gabriel]

In article <9romd9...@mid.individual.net>,

A parity generator chip is simply a multiple-input exclusive OR gate,
e.g. the 74280 9-bit Odd/Even Parity Generator is basically a 9-input
XOR gate (with an additional second inverting output).

[Phil]

Is there not a way to do this using X10?

[Andrew Gabriel]

In article <jjajmp$l9u$1...@dont-email.me>,

Actually, just to clarify, a parity generator is not simply a
multiple-input exclusive OR, but it is what you are after for
this logic case.

> Is there not a way to do this using X10?

You could get X10 into the solution if you wanted to.
Would it simplify it? Probably not - I don't believe
there's an X10 "toggle" command, so it won't help this
any.

[dennis@home]

"Andrew Gabriel" <and...@cucumber.demon.co.uk> wrote in message
news:jjai88$bqs$2...@dont-email.me...

> A parity generator chip is simply a multiple-input exclusive OR gate,
> e.g. the 74280 9-bit Odd/Even Parity Generator is basically a 9-input
> XOR gate (with an additional second inverting output).

XOR AIUI, output high when any one input is high but not when more than one
is high (its not very exclusive if it gives a high output with say three
high inputs).
I.e. its not a parity generator except for 2 bits.

I think you can make a parity generator by cascading multiple two input
XORs.

[Rod Speed]

Phil wrote
> Andrew Gabriel wrote
>> Andrew May<andrew...@hotmail.com> wrotes

Yes, it does it that way automatically, you can have
as many switches as you like and they are all identical.

Quite a bit more expensive then simple mechanical switches tho.

But you do save on the wiring, but that doesnt help if its there already.

[Rod Speed]

Andrew Gabriel wrote
> Phil <philip_barto...@yahoo.com> wrote

>> On 08/03/2012 15:12, Andrew Gabriel wrote:

>>> Andrew May<andrew...@hotmail.com> write

>>>> tin...@isbd.co.uk wrote

>>>>> I don't think that works does it? Apart from anything else an EOR
>>>>> gate only has two input, I'm looking for (possibly) three or more.

>>>>> The EOR approach will work for a two way switched light I agree.

>>>> True, I am not aware of a chip that has a multiple input XOR but
>>>> something like a 7486 has four two-input XOR gates that can be
>>>> cascaded. Two inputs into the first gate, another two inputs into
>>>> the second. The outputs of gates one and two into the inputs of
>>>> gate three. The output of gate three will be high if there are an
>>>> odd number of the four inputs high which satisfies the OPs
>>>> requirements.

>>>> Output of three could be connected to one of the inputs of four
>>>> and the other input connected to another switch if five inputs
>>>> where required.

>>> A parity generator chip is simply a multiple-input exclusive OR
>>> gate, e.g. the 74280 9-bit Odd/Even Parity Generator is basically a
>>> 9-input XOR gate (with an additional second inverting output).

> Actually, just to clarify, a parity generator is not simply a
> multiple-input exclusive OR, but it is what you are after for
> this logic case.

>> Is there not a way to do this using X10?

> You could get X10 into the solution if you wanted to.

In fact any X10 light can have as many switches as you like with no special treatment.

> Would it simplify it? Probably not - I don't believe there's an
> X10 "toggle" command, so it won't help this any.

You dont need a toggle command, each switch can turn the light on or orr.

You can kick yourself now.

[Martin Bonner]

On Thursday, 8 March 2012 17:08:34 UTC, dennis@home wrote:
> "Andrew Gabriel" <and...@cucumber.demon.co.uk> wrote in message
> news:jjai88$bqs$2...@dont-email.me...
>
> > A parity generator chip is simply a multiple-input exclusive OR gate,
> > e.g. the 74280 9-bit Odd/Even Parity Generator is basically a 9-input
> > XOR gate (with an additional second inverting output).
>
>
> XOR AIUI, output high when any one input is high but not when more than one
> is high (its not very exclusive if it gives a high output with say three
> high inputs).
> I.e. its not a parity generator except for 2 bits.

I don't think your understanding is correct. Whenever I have heard multi-input XOR referred to, it has been (A XOR (B XOR C)) == ( B XOR (C XOR A)) == (C XOR (A XOR B)) in other words TRUE if an odd number of inputs are true. ... which is just what a parity generator gives.

> I think you can make a parity generator by cascading multiple two input
> XORs.

Well of course.

[dennis@home]

"Martin Bonner" <martin...@yahoo.co.uk> wrote in message
news:27987529.77.1331300106982.JavaMail.geo-discussion-forums@vbhv6...

Well its probably age, but a gate which produces a 1 when an odd number of
inputs are 1 is a parity generator.
Why would you need to call it an exclusive OR.
What do you call a gate that produces a 1 when any one input is 1 but not
more than one if not an exclusive OR.