Need a magnetic key switch

Discussion in 'General Electronics Chat' started by MrSoftware, Dec 21, 2015.

  1. MrSoftware

    Thread Starter Active Member

    Oct 29, 2013
    504
    124
    Can anyone point me to a magnetic key switch? So far my googling has been fruitless.

    This is for elderly disabled people with limited dexterity. The idea is they walk up to the device they're turning on, stick the magnet on and that turns it on. Basically a magnetic reed switch would be fine, but it needs to have a strong enough magnet to hold the key there until the person removes it. It needs to handle 12v DC, and maybe 0.5A of current (enough to operate a relay), though more robust is fine too. One key can work for all switches, it's not a security device. I can make one, but I'm really hoping to find something off the shelf as we're going to need a bunch of them. Has anyone seen something like this?
     
  2. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,570
    2,381
    Last edited: Dec 21, 2015
    AnalogKid likes this.
  3. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,570
    2,381
    One other alternative using the SS400 devices is to use a bi-polar magnetic latch type, this way there is no need for the magnet to stay on the sensor, all you need to do is make a magnet actuator for the users, and mark one side ON the other OFF and it just needs a momentary activation to self latch.
    Max.
     
  4. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,550
    I wouldn't use a reed to drive a relay, but maybe rather a hall effect sensor (is that what a SS400 is, Max?) activating a transistor or a mosfet, since an ordinary reed is not capable of handling the large currents needed to drive a relay's coil.

    Once you've designed the right circuit, all you'd need to do is place a piece of metal beside or around the sensor so as to make your magnet stick until someone removes it.
     
  5. MrSoftware

    Thread Starter Active Member

    Oct 29, 2013
    504
    124
    Thanks for the ideas. Is there maybe a magnet detector circuit that I could use to drive a FET or a relay? Or maybe I could drive a MOSFET gates with a reed switch, and run the relay with the MOSFET?

    I think the bipolar idea would work for average people, but this crowd would likely forget to turn it off... But maybe turning it on would be enough if it automatically turned off, hmm.. Cool idea I'll think about that more, thanks.
     
  6. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,550
    I suggest you forget about the reed switch altogether... it's fragile, sensitive to vibrations, and will not withstand the test of time, IMHO. A hall sensor is a far better element. And besides, you will need to use a mosfet or a transistor to drive the relay coil anyway, whether you use the reed switch or not.
     
  7. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,570
    2,381
    Unless you use a SSR you will need a device to drive the relay such as a 2n7000.
    The SS400 come in Unipolar, Bi-polar and Bi-polar latch types, so you have a choice of latch or non-latch type.
    The Hamlin, now Littlefuse reed sw comes in a flange mount and are pretty rugged.
    I have used them for limit switches.
    What kind of devices are we talking?
    Max.
     
  8. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,546
    1,252
    We don't have much detail, but my guess is that a "device" being operated by an elderly person probably is not a high shock and vib environment. Plus, reed switches are hermetically sealed and good for tens if not hundreds of thousands of operations. And, their sensitivity does not decrease with age. It might need a current amplifier to drive a power relay, but so would a Hall device.

    Race car legend Jimmy Clark did not use the engine to decelerate into curves. Unlike almost all of his contemporaries, he used the brakes. And unlike *all* of his contemporaries, he still is the rated the #1 Formula One driver of all time - 40 years after his death. His idea was that something designed to accelerate sucked at decelerating; use the brakes to brake. Same thing here - for long term reliability, use a sensor to sense and an amplifier to amplify.

    ak
     
  9. MrSoftware

    Thread Starter Active Member

    Oct 29, 2013
    504
    124
    I've given my word that I wouldn't divulge too much about the project until it's done, but what I can say is this is to turn a device of sorts on and off on a car for handicap people. They will roll up to the car in their wheelchair, stick the magnet on and the device will power up. Currently a key switch is used, but these people lack dexterity in their hands and the key switch is a real challenge, and we think this will be an improvement. So this will be mounted inside a plastic box on the outside of a car. It doesn't have to be "secure", but there needs to be a component that can be taken away (a key of some type) so kids and such don't play with it. We feel a magnet will be fine.

    Ultimately the switch needs to control about 30A @14v. I like solid state parts so the hall effect sensor is appealing. Assuming I use a big MOSFET in place of a mechanical relay, will a hall effect sensor be enough to drive a MOSFET this size into saturation, or will I need another transistor or driver in between? Typical full current duty cycle will be less than a minute.

    I'm still hopeful to find an off the shelf solution, a switch capable of 30A @14v that is turned on with a magnet, but it's starting to look like I'll need to make something myself.
     
  10. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,550
    Most mosfets work with a minimum voltage at their gates of 10V (usually, the recommended level is 12 to 15V, since 20V is tops for them). So it all would depend on the working voltage of the hall effect sensor of your choice. If your sensor is capable of handling 14V, then I see no reason why it couldn't drive a mosfet directly, since those are voltage-driven devices, as opposed to normal transistors, which are current-driven.
    Your circuit would need to include tvs diodes strategically placed to protect the mosfet from transients.
     
  11. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,546
    1,252
    Yes.

    Many Hall effect sensors have an open collector output that can sink a few mA, enough to drive the LED input of a purchased SSR, and have a voltage rating of at least 12 V, enough to fully enhance a power MOSFET. So electrically this is looking pretty simple.

    To drive a power MOSFET directly, a reed switch would be controlling less than 1 mA, a current so small that the switch should last a looooong time. A non-Hall solution would be 1 reed switch, 1 10K resistor, and 1 MOSFET. OK, throw in one 0.1 uF cap for switch debouncing. No bells and whistles, K.I.S.S.

    ak
     
    cmartinez likes this.
  12. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,550
    Here are four hall effect sensor candidates for your consideration:

    http://www.digikey.com/product-detail/en/OH090U/365-1001-ND/374779
    http://www.digikey.com/product-detail/en/SS443A/480-2002-ND/701357
    http://www.digikey.com/product-detail/en/OHN3020U/365-1569-ND/1636533
    http://www.digikey.com/product-detail/en/OH180U/365-1567-ND/1636530

    @AnalogKid, I agree that a reed switch is a viable candidate for this application too, my only concern now would be vibration, since the device would be used in an automobile. But I have to admit that I don't have any previous experience with reed switches used in that field.
     
  13. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,546
    1,252
    How is the timing of the 30 A current related to the timing of the switch:

    1. 30 A comes on immediately, stays on as long as the switch is engaged, stops the moment the switch is opened.

    2. 30 A comes on immediately, and stays on a fixed amount of time no matter how long the switch is engaged. After its timed period it shuts off, and does not come on again until the switch is opened and then closed again.

    3. Other.

    Also, since this is an automotive or other vehicle application, mounting the reed switch vertically will decrease substantially its sensitivity to vertical shocks and vibration. Also, a 2nd resistor would create a turn-on time constant that can de-bounce the occasional pothole.

    ak
     
  14. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,570
    2,381
    I have used reed switches in limit switches and rotary tool changer switches in CNC machine applications, where they are exposed to quite a bit of vibration.
    Max.
     
    cmartinez likes this.
  15. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,550
    You'd probably need to add a couple of limit switches to the circuit too.
     
  16. MrSoftware

    Thread Starter Active Member

    Oct 29, 2013
    504
    124
    Thanks again guys, I think we're getting somewhere.

    This switch will actually be the master on/off switch to power the rest of the system. No power will flow until the user operates a secondary control. So time to full turn-on for the MOSFET can be quite long, even a full second or more would likely be fine. I think a typical use case would be this master switch is turned on, 20 seconds later the user operates a secondary control causing 30A current to flow for about 30 seconds, probably a 30 second rest period (no power flowing) followed by 30 seconds more of 30A current, then the key is removed and the system powered off for a relatively long period of time.

    I've got some FZ44VZ MOSFETS in my parts bin, maybe this would do the trick. What size pull down resistor should I use on the gate, would a 10k do the job?

    http://www.irf.com/product-info/datasheets/data/irfz44vz.pdf
    https://www.digikey.com/product-detail/en/IRFZ44VZPBF/IRFZ44VZPBF-ND/666098
     
  17. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,570
    2,381
    The Mosfet should work, and 10k also.
    Max.
     
  18. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,546
    1,252
    First pass at a magnetic key switch. Note that these are parts in my design library. Your MOSFET should be fine, there is no need for 1% parts, etc. Pretty generic stuff. As with all automotive circuits, there could be some transient protection and power cleanup, but this is enough to discuss the operational details.

    ak
    CarDeviceSwitch-1-c.gif
     
    cmartinez likes this.
  19. MrSoftware

    Thread Starter Active Member

    Oct 29, 2013
    504
    124
    Wow awesome thanks! I might have equivalents for all these parts in my bin, I'll make one up in the next couple of days and post back the results.

    Are R2 and C1 for noise control, and D? for protection against inductive spikes from the load?
     
  20. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,546
    1,252
    Yes and yes. R1-C1 prevent accidental turn-ons caused by mechanical shock to the reed switch (it takes a lot, but still...). R2 is there to turn off the FET when the switch opens. Its time constant with C1 is not needed for anything; it is a consequence of C1 being there to filter turn-ons. Both time constants slow down the FET transition from off to on and back, increasing its power dissipation if there is any significant current through it at those times. Both times are less than 0.1 second, so that should not be a problem. More precise control comes with an opamp or comparator as the input circuit.

    ak
     
Loading...