Confusion about rotary switches.

[Skywave]

I'm sure my understanding about the descriptions of multi-pole, multi-way rotary switches is correct, but it seems that in some quarters, confusion reigns. As examples, my understanding is as follows, (where x, y and z are substitutes for numbers):

1. One wafer, one rotating contact, x selectable positions.
That is a one-pole, x-way, rotary switch.

2. One wafer, two rotating contacts, (not electrically connected), y selectable positions per rotating contact.
That is a two-pole, y-way, rotary switch.

3. Three wafers, one rotating contact per wafer, z selectable positions.
That is a three-pole, z-way, rotary switch.

I have deliberately omitted reference to MBB or BBM since here, those are not relevant.

Your comments are invited, please: thanks,

Al.

[Biggles]

Sound reasonable to me Al.

[ronbryan]

Al

Your example 3 does not include the number of wafers in the final description. Could you entertain the possibility that example 3 could be called a 'three bank 1 pole z way rotary switch'?

I mention this possibility only because the number of wafers is not otherwise defined in the switch description.

Ron

[kalee20]

Generally OK. But be aware that there are other wafer types than a single wiping contact - such as all contacts shorted except one, the one which ISN'T is selected by shaft position.

[Craig Sawyers]

The last one is three wafer, each wafer being a single pole z way switch.

But there are always exceptions, for example where there are two sets of contacts on the same wafer - one set on one side and one set on the other. And the kind that kalee20 mentions.

Craig

[emeritus]

Further to #3 I have some ELMA rotary switches that have three single-pole wafers and an adjustable stop that allows you to select anything between 2 and 12 positions. So it could be configured as a 4 way switch, when it would be a three wafer, three pole, 4 way switch. I have other switches that have three poles on a single wafer, that would be one wafer, three pole, 4 way switches. Admittedly it would be unusual to need to have a three pole, 4 way, switch with one pole per wafer, but I suppose it might be necessary in particular circumstances.

[Diabolical Artificer]

I'll admit to get completely confused about switch nomencleture,. Thank's Al for shedding some light on the matter. So a pole is the contact, input as it were and a way is the output. That makes life easier.

Trying to work out complex switching is a mind bending exercise. A recent example is a four wafer sw I tried to use to switch four guitar pickups. The sw had 8 positions, each wafer had the upper on position's 1-4, the lower 4-8. Each PU has two wires screened, one must be grounded.

The neck PU (PU 1 & 2) is usually configured as a humbucker so they're connected in series, an easy one to switch. The difficulty arose when trying to work out how to switch PU's out of phase, as singles etc. Which means selecting and cutting, grounding or making positive each of the 8 wires. I never did quite manage it and had an exercise book full of diagrams charts.

Andy.

[kalee20]

Originally Posted by Dr Wobble

I'll admit to get completely confused about switch nomencleture,. Thank's Al for shedding some light on the matter. So a pole is the contact, input as it were and a way is the output.

Sort of, the pole number is effectively the number of independent channels on ths switch. And the way number is the number of positions.

There is the irregularity that if there are just two ways, they are usually renamed throws. So a 2 pole 2 way switch is usually called double-pole double-throw, or DPDT. Similarly, DPST, SPST, SPDT. More usually applied to toggle switches, but can be applied to rotaries too.

[Skywave]

Originally Posted by ronbryan

Your example 3 does not include the number of wafers in the final description. Could you entertain the possibility that example 3 could be called a 'three bank 1 pole z way rotary switch'?

I mention this possibility only because the number of wafers is not otherwise defined in the switch description.

Yes, that reference could be added to give a fuller description of the switch. Although I realise that that can be a useful piece of additional info., my concern here is the choice of words and phrasing which defines the logical operations only of a given switch.

[Skywave]

Originally Posted by kalee20

Generally OK. But be aware that there are other wafer types than a single wiping contact - such as all contacts shorted except one, the one which ISN'T is selected by shaft position.

Indeed: that type I am well aware of. On a multi-wafer switch intended for coil switching, it is common to find at least one wafer that does that - shorting out unused coils.

Al.

[Skywave]

Originally Posted by Craig Sawyers

The last one is three wafer, each wafer being a single pole z way switch.
But there are always exceptions, for example where there are two sets of contacts on the same wafer - one set on one side and one set on the other.

Which makes 2-poles per wafer. So if there are N wafers all the same, then it's a 2*N pole switch. As I said earlier, the number of wafers can be a useful piece of supplementary info., but the switching operations are usually not affected by the number of wafers.

Al.

[Skywave]

Originally Posted by Dr Wobble

Thanks Al for shedding some light on this matter.

No trouble at all: I'm here to contribute as much as I am here to learn.

Re: your post #7:
I always regard the moving contact of a rotary switch as the 'pole'; what that pole connects to as the 'way'. So, yes: 'pole' = 'input'; 'way' = output . . . usually!

However, the possible arrangements for a multi-wafer, multi-pole, multi-way switch are almost limitless! Most of the time, the notation of M-poles, N-ways (with possibly W-wafers stated) is adequate. But as Kalee20 points out, there are ones that are a 'logical not' type. And then there are ones which are totally bespoke: different switching arrangements on each wafer; different number of poles on each wafer; some 'ways' not present, etc. Obviously, the simple nomenclature stated above cannot be used for such arrangements. I recently had to scrap-for-salvage a Marconi TF 1041 (horrid thing: a design disaster, IMHO), and the two rotary switches are of that bespoke type. They might come in useful one day: possibly removing wafers to build a custom switch for myself.

But taking an overview, 'switchery' can indeed get complicated - as you report. Many are the hours I've spent with pencil and paper trying to optimise such for a given need. Often, resorting to Boolean logic and Karnaugh maps helps.

Al.