Hmmm. I can see from the link that the random option would be more versatile. For that option I'm now wondering if perhaps you might need 3 coils to guarantee the movement you want?

Three coils would be prohibitively expensive. I intend to produce quite a few of these units once I have a working prototype so the fewer components the better. Having given it some more thought I prefer the random approach out of the two options, also reversing the polarity of the coil should avoid those pesky dead spots, plus I think I'd only need one coil.

Nick

 

If the magnet happened to land randomly with one pole directly aligned with a single coil, isn't it possible that neither pulse polarity would cause a twitch?

 

Three coils would be prohibitively expensive. I intend to produce quite a few of these units once I have a working prototype so the fewer components the better. Having given it some more thought I prefer the random approach out of the two options, also reversing the polarity of the coil should avoid those pesky dead spots, plus I think I'd only need one coil.

Nick

I agree that a single coil and polarity reversal seems like the best approach for your desired effect. You might be able to lower parts count and assembly labor (although possibly trading those for higher cost) by using an h-bridge driver IC. I'd have to dig deeper to be confident in any specific recommendations, but after a quick search on DigiKey, I thought this one looked promising:

http://www.digikey.com/product-detail/en/TLE5206-2S/TLE5206-2S-ND/1283081

That would handle the power switching, but you'd still need something to tell it when to reverse polarity. I'm thinking you could use a flip-flop for the polarity reversal signal and use the direct signal from the receiver to activate power. That would mean that each button push created a pulse on the coil, and that each button push would be opposite polarity from the one before it. So, most of the time each button push would make the dial swing. In the rare cases when the dial happened to be resting in the dead zone for one polarity, the first push would have little to no effect, but the second push would have a powerful effect. Does that sound ok, or do you need the polarity to switch back and forth quickly and automatically so that the effect never requires two separate button pushes?

 

How long do you want;

1. The first solenoid to be on
2. The pause
3. The second solenoid to be on

I was thinking a 4017 or 4022. The pushbutton would trigger a monostable to drive Clock Inhibit low (enabling the counter) while an astable would drive the counter. It's frequency and a combination of the outputs would create the sequence desired.

Or use an Arduino Pro Mini (or similar small form factor microprocessor or PIC)

 

Are these gadgets going to be hand-held, or static?

 

If the magnet happened to land randomly with one pole directly aligned with a single coil, isn't it possible that neither pulse polarity would cause a twitch?

Seems highly unlikely to me, but maybe possible. If the magnet lands in its null spot for one polarity, the opposite polarity would create a powerful repulsive force on the magnet, making it want to be in any other position more than its current position. It would be like balancing a ball on the point of a pyramid. There's theoretically a position in which it would stay, but in reality very unlikely.

Having said that, the real world scenario could also have imperfections that favor stasis. If the pivot point is relatively stiff, that might increase the very narrow range of angles at which there's too little force to get things moving.

 

If the magnet happened to land randomly with one pole directly aligned with a single coil, isn't it possible that neither pulse polarity would cause a twitch?

Just tried it and reversing the polarity chucks the magnet round on its way so a solution that can switch polarity seems to be favourite at the moment.

Thanks
Nick

 

That would mean that each button push created a pulse on the coil, and that each button push would be opposite polarity from the one before it. So, most of the time each button push would make the dial swing. In the rare cases when the dial happened to be resting in the dead zone for one polarity, the first push would have little to no effect, but the second push would have a powerful effect. Does that sound ok, or do you need the polarity to switch back and forth quickly and automatically so that the effect never requires two separate button pushes?

Hi Ebowolf17, that sounds perfect and is something I'd visualised. I just don't have the technical knowledge at present to design the circuit so any help greatly appreciated.

Cost is not an issue for my projects, space is however. The dial and 'werks' for this project is to fit in a small wooden clamshell type box approx 3.5 x 3.5 x 1.5 inches which I will be making.

Thanks
Nick

 

Are these gadgets going to be hand-held, or static?

Static. During performance, a small wooden box is introduced, opened and left on the table. The sitters then place hands on the table around the box and the seance begins! From that moment it is not touched by the performer - hence the toe switch. Spooky eh? Hope there are no magicians on this board - I'm giving all my secrets away here!

 

Seems highly unlikely to me, but maybe possible. If the magnet lands in its null spot for one polarity, the opposite polarity would create a powerful repulsive force on the magnet, making it want to be in any other position more than its current position. It would be like balancing a ball on the point of a pyramid. There's theoretically a position in which it would stay, but in reality very unlikely.

Having said that, the real world scenario could also have imperfections that favor stasis. If the pivot point is relatively stiff, that might increase the very narrow range of angles at which there's too little force to get things moving.

As you say - very rare. If it does happen I'd simply pick up the box and place it in a different position and in doing so dislodge the magnet from its null position. Not a problem!

 

Spooky eh? Hope there are no magicians on this board - I'm giving all my secrets away here!

Apparently, you are merely an assistant with no powers of clairvoyancey.

 

Apparently, you are merely an assistant with no powers of clairvoyancey.

Correct. Clairvoyancy is a fallacy, sorry to burst your bubble. LOL

 

No bubble burst. See, I knew you were going to say that. I used to be a seer to earn extra cash.

 

Another idea now that what your doing is known. Instead of a magnet wheel and solenoids, how about an RC servo with a rotary switch to select appropriate resistor values ? The servo works by different resistance values giving different amounts of movement. The 'dial' could then be labeled with, the yes - no - answer unknown, etc. By selecting the switch to different resistor values the pointer would then point to the different "answers".

 

Was there something wrong with the 2 pulse idea?

 

Hi Ebowolf17, that sounds perfect and is something I'd visualised. I just don't have the technical knowledge at present to design the circuit so any help greatly appreciated.

Cost is not an issue for my projects, space is however. The dial and 'werks' for this project is to fit in a small wooden clamshell type box approx 3.5 x 3.5 x 1.5 inches which I will be making.

Thanks
Nick

Maybe someone else on these forums can help sort this out. I thought I had an idea for how to sim this and make it work, but I can't quite get it right.

It mostly works if I leave out the "delay for enable" components on the schematic below. However, depending on the rise time of the switch signal, sometimes the flip-flop polarity reversal wouldn't trigger until after current was already flowing through the coil, which would mean trying to reverse coil current while it's flowing, which is a really bad idea.

That's why I tried adding the simple delay circuit so that current flow through the coil wouldn't start until distinctly after the flip-flop had time to set the correct polarity. But in the sim (LTspice) the flip-flop never changes output states with the delay circuit added. I can't understand it at all. From where I sit, there shouldn't be any interaction between the two.

I'd like to think that the concept laid out below is valid, and that it's just a matter of component selection and fine tuning component values, but I'm really not sure. Any advice is welcome!

 

Was there something wrong with the 2 pulse idea?

Hi Ron_V, nothing at all, in fact I was going to order the components. I was just interested to see if it could be done any easier just using one coil.

Nick

 

Maybe someone else on these forums can help sort this out. I thought I had an idea for how to sim this and make it work, but I can't quite get it right.

It mostly works if I leave out the "delay for enable" components on the schematic below. However, depending on the rise time of the switch signal, sometimes the flip-flop polarity reversal wouldn't trigger until after current was already flowing through the coil, which would mean trying to reverse coil current while it's flowing, which is a really bad idea.

That's why I tried adding the simple delay circuit so that current flow through the coil wouldn't start until distinctly after the flip-flop had time to set the correct polarity. But in the sim (LTspice) the flip-flop never changes output states with the delay circuit added. I can't understand it at all. From where I sit, there shouldn't be any interaction between the two.

I'd like to think that the concept laid out below is valid, and that it's just a matter of component selection and fine tuning component values, but I'm really not sure. Any advice is welcome!
View attachment 100253

Hi ebeowulf17, really appreciate your input on this. It's beyond my knowledge to comment so hopefully someone else will chime in.

Many thanks
Nick

 

Can you post your asc file, ebeowulf?
Nick, can you determine the approx minimum pulse duration needed to twitch the magnet when the coil is driven with 3V?

 

Can you post your asc file, ebeowulf?
Nick, can you determine the approx minimum pulse duration needed to twitch the magnet when the coil is driven with 3V?

Hi Alec_t, I would say 500ms but I'm basing that on a single press of a microswitch on my breadboard. Hope that helps.

Thanks
Nick