Normally closed semiconductor switch

Discussion in 'General Electronics Chat' started by Marion Haggis, Aug 1, 2012.

  1. Marion Haggis

    Thread Starter New Member

    Aug 1, 2012
    3
    0
    Hi All,

    I want to design a semiconductor based lamp switch circuit that defaults to normally closed i.e the lamp is on. I have a very limited power source (a coin cell battery) that will power a microcontroller to switch the light on/off. If the battery fails I want the lamp to default to being turned on, not off. Triacs etc need a gate voltage to turn them on, I want something that defaults to on (with no gate/trigger voltage) and is held off by applying a voltage (DC, ~1V?) to the gate. As I have a very limited power source, I want the gate current to be micro amps (around 100uA). I think this rules out solid state relays. In an ideal world I'd like the microcontroller to be able to dim the light with this circuit too. Any ideas?

    Thanks
    Marion.
     
  2. #12

    Expert

    Nov 30, 2010
    16,294
    6,804
    Jfets are normally "on" and this one requires less than a volt to turn it off. It's only a start but, if you understand jfets, you can work with this (or similar).
     
  3. crutschow

    Expert

    Mar 14, 2008
    13,014
    3,234
    The problem with a jfet is that it requires a negative gate-source voltage for turn off. That makes it difficult to shut off when you only have one supply voltage.
     
  4. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Last edited: Aug 1, 2012
  5. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    "If the battery fails I want the lamp to default to being turned on, not off."

    Where does the power supply for the lamp come from? Does it have the same reference as the coin cell? What's it's voltage?
     
  6. #12

    Expert

    Nov 30, 2010
    16,294
    6,804
    My point exactly. So much informatin missing that we can't tell if there is room to set up a voltage divider on a 1 nanoamp gate, what the processor is going to do, etc.
     
  7. Marion Haggis

    Thread Starter New Member

    Aug 1, 2012
    3
    0
    Sorry everyone, I thought my problem was clearly described but realised I missed out a key point - the lamp is AC powered and I'm after a digital control circuit to switch an AC lamp on/off, defaulting to ON if the power in the digital circuit is removed.

    The micro is currently a Cypress psoc, using their first touch demo HW with RF. (I wan't to make a remotely controlled lamp. )

    The coin cell is a 3V CR2032 and the psoc runs from the cell but is mostly sitting in sleep mode with a low duty cycle wake up to check for any RF transmissions.

    I want the PSOC to control the light but I don't want the PSOC/coin cell to have to supply large currents (relative to what the PSOC requires) to energise a relay (to turn the light on), or trigger a solid state relay etc. Additionally, I want the circuit to default to normally closed so if the PSOC battery dies (so there's no power at all in the control circuit), the lamp will default to on, not off.

    The issue is that I need a lower power solution (>1V gate/trigger voltage, 100uA typ current) for keeping the lamp turned _off_ and for it to default to normally closed when there is no power.

    Any ideas?

    Thanks,

    Marion.
     
  8. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Unless we find a component that already has the AC power parts all built-in you will need to built that AC part for yourself. Discussion on that may possibly not be allowed on AAC. For the mods to decide.

    example:http://www.discovercircuits.com/PDF-FILES/lowssr2.pdf

    The circuit needs to be adapted in order to work as normally closed by changing the position of the optocoupler.
     
  9. MrChips

    Moderator

    Oct 2, 2009
    12,442
    3,361
    Use a photo-coupler (opto-isolator) whose emitter is powered from the AC supply to turn on the photo-receiver which turns on a TRIAC.

    Your digital control is used to disable the photo-coupler.
     
  10. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Can you explain further? The OP wants a control current between nothing and maybe a few uA.
     
  11. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    So I guess the uC part and the AC part are not necessarily required to be galvanically isolated?

    If it does need to be isolated the why use a battery and not a AC#DC converter? There are some tiny modules or you can build one yourself.

    If you really want to go with the battery AND it needs to be isolated you may have to use something similiar to what I posted before and maybe use a depletion MOSFET to trigger the TRIAC. If you use pulses to charge its gate you may get away with less current consumption on the optocoupler's control side.
    Anyway, I don't see why use a battery if a big AC power source is available... You would never need to worry about replacing a battery.

    I recently built an RF remotely controlled lamp, but it cost me about $30 I think, and then I saw they sell them on ebay for much less... Now I think of how to make it much cheaper...
     
  12. MrChips

    Moderator

    Oct 2, 2009
    12,442
    3,361
  13. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
  14. Marion Haggis

    Thread Starter New Member

    Aug 1, 2012
    3
    0

    Hi praondevou,

    The battery is needed as there is no current flowing in the AC wires when the lamp is turned off. The whole circuit will be in a box so I don't need the uC to be isolated.

    Thanks
    Marion
     
  15. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Something I don't understand. If there is no AC voltage available if the lamp is turned off (via RF and the uC) then how will you turn on the lamp???

    If the uC controls the lamp then there must always be AC present. If AC is always present you don't need a battery.

    What am I not seeing here? Can you describe the logic of that circuit in detail?
     
Loading...