Hello, I'm looking for a (physical) switch arrangement that achieves a very particular kind of behaviour:
-- It's a rotary switch, or some arrangement that allows traversing its three states in repeating fashion without hitting a limit and coming back (state 1, state 2, state 3, state 1, ....). Think multimeter dial, except only 3 states.
-- There are 3 inputs (X, Y and Z) and 3 outputs (A, B, and C), and the three states are:
-- State 1: X--A Y--B Z--C
-- State 2: X--C Y--A Z--B
-- State 3: X--B Y--C Z--A ( so you can see it's very much rotating through the (non-reversed) possible permutations )
-- 40V, 3A (as if this wasn't hard enough already)
-- Break-before-make, naturally
-- Not at all fussed about how it looks/how it's mounted.
-- EDIT: I only need one of these; this is a one-off project.
I've seen teardowns of test gear on EEVBlog that have little switches stacked together, all operated by a single shaft running through the whole stack, this might be a promising approach but I'm struggling to find the relevant category on element14 -- and I'm not sure those are "high-power" parts. Any ideas would be greatly appreciated.
Can you use MOSFETs or similar to take the high current off the switch?
Can you use MOSFETs or similar to take the high current off the switch?
That's the fallback option I'm trying to avoid -- MOSFETs would require power to run, and I'm not quite sure how to drive MOSFETs to link one power signal (which may be positive or negative) to another. I guess some arrangement of BJTs and optocouplers could do the trick too, but these are all much more complex than the clearly manufacturable (but maybe not marketable) physical switch I describe.
Many rotary switches have physical "stops" in them to stop you rotating them continuously. If you don't need a huge number of these I would look at modifying some rotary switches.
Many rotary switches have physical "stops" in them to stop you rotating them continuously. If you don't need a huge number of these I would look at modifying some rotary switches.
Interesting idea. To clarify, this is a one-off project, so modification is definitely an option.
I've seen teardowns of test gear on EEVBlog that have little switches stacked together, all operated by a single shaft running through the whole stack, this might be a promising approach but I'm struggling to find the relevant category on element14 -- and I'm not sure those are "high-power" parts. Any ideas would be greatly appreciated.
Complex switches are not cheap brand new from proper resellers.
http://nz.element14.com/electroswitch/d4g0303n/switch-rotary-3p3t-1-5a-115v/dp/1383598Search for 3 position 3 pole.
But yeah, 3A is probably pushing things.
You might find it easier to use a relay or two and a lower current switch
Does your switch need to make/break active 3A power feeds or will switching be done with power off (so it only has to carry 3A, not switch it).
If the latter you can get away with a lower current switch.
-- It's a rotary switch, or some arrangement that allows traversing its three states in repeating fashion without hitting a limit and coming back (state 1, state 2, state 3, state 1, ....).
I doubt you will find anything with 3 positions and 120 degree spacing it isn't ergonomic which prompts the question what are you really trying to do.
If the switch is operated by some mechanism then a cam switch would be the usual solution. 3 stacks of 3 micro switches with 120 degree spacing operated by a single lobe cam would work, or a stack of 9 micro switches and 3 single lobe cams.
I doubt you will find anything with 3 positions and 120 degree spacing it isn't ergonomic which prompts the question what are you really trying to do.
If the switch is operated by some mechanism then a cam switch would be the usual solution. 3 stacks of 3 micro switches with 120 degree spacing operated by a single lobe cam would work, or a stack of 9 micro switches and 3 single lobe cams.
I dont think he needs 120deg spacing, at least he doesn't say that. Just that it needs to be 3 position. So a rotary switch with one platform containing 3 polls and 3 positions per poll should work. It would have 9 detents when rotated by would cycle through only 3 switch combinations. The 3 poll terminals would be the 3 inputs and the 3 sets of 3 position terminals could be shorted together to create the 3 output patterns he wants.
At least, thats how i interpret his requirements.
I dont think he needs 120deg spacing, at least he doesn't say that. Just that it needs to be 3 position. So a rotary switch with one platform containing 3 polls and 3 positions per poll should work. It would have 9 detents when rotated by would cycle through only 3 switch combinations. The 3 poll terminals would be the 3 inputs and the 3 sets of 3 position terminals could be shorted together to create the 3 output patterns he wants.
At least, thats how i interpret his requirements.
Correct.
I doubt you will find anything with 3 positions and 120 degree spacing it isn't ergonomic which prompts the question what are you really trying to do.
If the switch is operated by some mechanism then a cam switch would be the usual solution. 3 stacks of 3 micro switches with 120 degree spacing operated by a single lobe cam would work, or a stack of 9 micro switches and 3 single lobe cams.
This is a great suggestion, can you buy cams that are designed with this end-use in mind? Or is this more of a DIY job? Either way, I'm going to explore this.
It's about time I explained what I'm "really trying to do", because it is a bit silly.
The long term goal is to make a circuit (most likely with FETs or solid state relays or something) that can sit in between a 3-phase (3-winding? Y winding?) BLDC motor and its controller, and somehow "overclock" it. So if the controller puts out 60 rpm, the circuit chops that up and makes it 90 or 120 rpm. Now the short-term goal, where the switch comes in, is in figuring out how fussy the controller is, and whether it'll get upset by getting EMF spikes back when it's least expecting it. What I was hoping to do was fit this on one wheel of a two-wheel robot, and by spinning this switch around while it thinks it's driving straight forward, make it veer off course. If, instead, it goes into something's-wrong-I-give-up mode, then I know not to bother trying! Worth it, and kinda fun, if I can slap something together for a little bit of money, but not essential.
This is actually quite an easy problem to solve. Lorlin plastic enclosed switches will do what you want. Here is how to do it:
1. Unscrew the mounting nut and remove the end stop washer (it has a protruding stop that goes into the switch and sets the number of postions). The rotary switch is now 360 Degrees with no stops.
2. The switch selected should be a 3 pole 4 way (with these switches they always add up to 12) Connect pin 1 to pin 4, then to pin 7, then to 10. Pin 2 should be connected to pin 5 then to pin 8 and then pin 11. Pin 3 should be connected to pin 6, then to pin 9, then pin 12.
I believe this will give you what you wanr i.e a 3 pole 3 way switch with no end stop.
I have used Lorlin switches for 25 years and found them to be reliable. A good choice.
http://www.lorlin.co.uk/rotary_switch_products.phpAurora
Yes but they're limited to 150mA. Is that enough?
There are three versions shown CK, CKF & CKS. CK is stated as 5A non switching (?) current and the other two are just stated as 1A @ 24V. The 150mA rating is stated as when working at 250V ac/dc. For info I have run them at 3A with no issues at all.
Much will depend upon the application and how often the full load will be switched. Also whether the installation is designed for a certain level of reliability.... i.e. a mission critical situation where the switch cannot be a weak link.
Aurora
If you intend to rotate the rotary switch using a motor then the switch contacts need to be rated for the speed of rotation. Usually for rotating like this you would use an optical/magnetic sensor to pickup the rotation and a semiconductor to do the switching.
However if you're just doing slow speed testing it may work ok for a short time.
High current and fast moving mechanical switches don't mix well
