Hello all,

I am working on an artwork that takes the form of a pendulum operated by a 12 dc volt solenoid.

Solenoid Link here:
http://www.hobbytronics.co.uk/solenoid-12-36v?utm_source=google&utm_medium=googleshopping&utm_campaign=googlebase&gclid=COWFwcPk2LwCFYUIwwodRgYAUg

a reed switch and magnet on a moving paddle attached to the pendulum arm (see attached image)....

All works well for about 20 mins then the the reed switch contacts fuse together and the pendulum stops. Blue sparking starts to occur prior to the fusing together suggesting an overload of some sort. I am not too knowledgeable about electronics so does anyone know of an simple elegant way of making this thing work without fail? A simple circuit i can build.

I am not sure about the value of the reed switch but it is about and inch and a half long and i think, but am not totally sure is rated at about an amp??

I am waiting for a video of the pendulum to upload to give you a better idea of how the pice works, will post presently..

here it is : http://vimeo.com/97652817

all the best



Derek Ogbourne

 

... and i think, but am not totally sure is rated at about an amp??

Well I think you've answered that question.

Do you have any automotive relays handy? You could use your reed switch to switch a relay instead of the solenoid, which would likely take fare less current and not burn your reed switch, and then use the relay to switch the solenoid.

 

Thank you for my first reply ever on this forum

I don't have a automotive relay handy. i am kind of set on the solenoid idea as it gives the pendulum a good kick, i would prefer to have another type of switch rather than a reed type.

 

Someone few weeks ago suggested an interrupter? but these seem to only trigger smaller currents...

 

I had a similar stuck-contact problem with a reed of unknown make. The contacts were carrying well below their rated current and had not burned, but they had become permanently magnetised by the passing magnet! Try a different reed brand.

 

Maybe you are right a different reed switch might work, a giant one would work

 

i did try two reed switches together but this didn't work

 

Thank you for my first reply ever on this forum

I don't have a automotive relay handy. i am kind of set on the solenoid idea as it gives the pendulum a good kick, i would prefer to have another type of switch rather than a reed type.

I wasn't suggesting any change to the solenoid, only that you insert a 12V relay in between the reed and the solenoid. This would reduce the current being switched by the reed. Any junkyard would have tons of relays to choose from.

Another switch choice is called a "roller switch". Just search that term and you'll find lots of examples. It requires contact, so you may not like that, but the ones I've seen need very little pressure to activate.

There are many solutions if you can live with a small circuit instead of relying on the switch to do it all. Photo cells, hall effect sensors, all sorts of things. But all of these separate the sensing function from the switching function and are a bit more complex.

 

Ah i see , good, thanks, a small circuit would do me as well ... will look into the roller switch idea and the automotive relay one in the meantime. Found a Reed switch company:

http://www.meder.com/switch_ksk-1a54_uk1.html?&tx_jppageteaser_pi1[backId]=2866

maybe this reed switch would withstand the fusing ?

thank you

 

ah.. just had a look at the roller switch, and have tried one of those already does slow down the pendulum to a halt, it needs very little friction to keep the the device swinging.

 

Would this really work Wayneh?

http://www.maplin.co.uk/p/12v-30a-dc-spdt-automotive-relay-n00aw

 

I have a suggestion to make this design more efficient.
Instead of triggering the solenoid when the pendulum gets within range, you should trigger the solenoid when the pendulum goes OUT of range. In other words, trigger the solenoid as soon as the pendulum begins to swing backwards. That way you're not stopping and then reversing the pendulums swing with each cycle. The only problem being that, because it's more efficient and always adds the same amount of energy each cycle, you might have a hard time calibrating the pendulum just right, so that it swings at the same height each time without any sort of feedback loop.
If you use a microcontroller, you could apply feedback by timing the delay between each cycle.
Just an idea. Perhaps too much work for this project.

 

Yes, it would need only about 150mA instead of the larger current needed by the solenoid. It's overkill for switching the solenoid, but it would work.

You know, the magnet actually causes a bit of drag when it moves past the reed switch (or any conductor). It's probably tiny but it's there. An optical interrupt would be zero drag. You could use something like this. (Not an endorsement, just an example of what I'm picturing.)

I think it's possible you might find such a thing that's rated for enough current to switch your solenoid with no other parts. Not sure.

 

Yes it is a difficult one to get the timing right Austin, i normally lift the pendulum as high as it will go then adjust the distance of the paddle. i have to many options now , hmm, i think the micro controller would be a little complicated for me. I have thought of paying someone to sort this out, this is the first of 3 pieces that i am building, the other two i think area little bit more complicated than this one. I tried to find someone but ended up here which is brilliant.

 

Wayneh, i might try the relay and then the interrupter although i have no idea which one to buy, the specs confuse me.

 

Do you think this circuit will work?

 

All works well for about 20 mins then the the reed switch contacts fuse together and the pendulum stops. Blue sparking starts to occur prior to the fusing together suggesting an overload of some sort. I am not too knowledgeable about electronics so does anyone know of an simple elegant way of making this thing work without fail? A simple circuit i can build.

It sounds like you could benefit from a catch diode across the solenoid to suppress the kickback voltage (causing the arc and burning of the reed contacts). A 1N4001 or higher would be sufficient. Put it across the coil of the solenoid: banded end towards the + terminal of the battery. That will help reduce the wear on the reed from arcing when the switch opens.

You'll still need to ensure that the reed can handle the 650ma load of the coil. If it won't, the driving the coil with an external relay or transistor will be called for BUT, you'll still need the diode. If you use a relay to drive the other relay, you should put a diode across its coil as well.

Good luck.

EDIT: the hot setup is a diode and zener as described in the attachment but the diode alone is sufficient for what you are doing.

 

That was very useful info John, lets see how it goes with a diode then, be back either way until i get it sorted. Thanks

 

I might also suggest a diode across the solenoid in case the kick from the solenoid when the reed switch opens is causing arcing and damage the contacts thereby causing failure. But keeping the current within the reed switch spec is very important without a doubt.

 

Thanks for reafirming the use of a diode in the circuit, will give it a go tomorrow