Mosfet question.

Discussion in 'General Electronics Chat' started by montner, Oct 21, 2010.

  1. montner

    Thread Starter Member

    Mar 12, 2009
    10
    0
    I have the following circuit on a backplane. It controls 2 electromagnetic counters when pulsed by the main board. The board can detect if the counters are plugged in, if not, it gives an error message. How can I remove the counters and mosfet and make the board think they are still there. I don't know enough about mosfets or electronic design to figure it out myself. Can I show potential 12V at the signal to trick it? I hope someone can explain the complete circuit, as well as give me the answer, because I always want to learn, not just get an answer quickly.

    My 4 year old grandson did the schematic. Well maybe.
     
  2. montner

    Thread Starter Member

    Mar 12, 2009
    10
    0
    Here is the datasheet on the mosfet. With these specs maybe someone can answer how the board detects the 12v through the mosfet.
     
  3. montner

    Thread Starter Member

    Mar 12, 2009
    10
    0
    Someone PLEASE help me. I think the following will work but don't want to ruin my board. I think the board was sensing the induced voltage across the semiconductor, so adding a capacitor to simulate it and adding a diode to stop the higher voltage coming from the board it might work. I would like someone smarter than me to help.
     
  4. timrobbins

    Active Member

    Aug 29, 2009
    318
    16
    I suggest just using the 10k ohm resistor to ground will do the trick. The mosfet gate is effectively an open circuit - and doubt that any detection technique would use the mosfet per se. The 10k is probably used in a pull-down sense for a logic sense.

    Ciao, Tim
     
  5. montner

    Thread Starter Member

    Mar 12, 2009
    10
    0
    Thanks for the reply Tim. If I unplug the electromagnetic coil it senses that and comes up with an error. The only way I can see it doing that is if it is sensing the 12V through the mosfet, but since it is physically separated it must be sensing the capacitance through? If i unplug the coil, that is the same thing as what you are suggesting?
     
  6. mbohuntr

    Active Member

    Apr 6, 2009
    413
    32
    Sorry for not replying,I did read it. The schematic was well beyond my comprehension,we all seem to have areas of expertise. Mine is very small, and very narrow... read... "still learning"... :rolleyes:
     
  7. timrobbins

    Active Member

    Aug 29, 2009
    318
    16
    If the uP is giving you an error message the it is geeting another form of feedback to tell it the counter is not in circuit. Perhaps there is a current sense signal being measured - does the 0V to the mosfet come directly from system ground - or likewise does the +12V line come directly from system +12V - or do either go through a sense resistor or some other circuitry?
     
  8. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Are there more layers to the circuit board where other means of sensing the presence of the counters could exist?

    What is the uP? There isn't really any way for the I/O lines to "see" the coils as you have it drawn, without some nifty magic in the software, anyway.
     
  9. montner

    Thread Starter Member

    Mar 12, 2009
    10
    0
    Thanks for the replies. The 12V comes straight from the source and the 10K resistor and source off the mosfet go directly to ground. There are no other connections to this circuit as I have traced every pin out to the board. It is exactly as I have drawn it (badly I might say). I am not sure why the diode is in parallel with the electromagnetic coil. But nevertheless there is always 12V at the drain of the mosfet when the coil is plugged in and missing when it is removed. And the board senses that.

    Is it possible that the voltage induced across the silicon dioxide raises the gate to a high level compared to O Volts the same way communication is done in other circuits such as serial communication?
     
  10. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    The diodes are there to snub a voltage spike when the current stops flowing through the coil. Diodes across coils are basically essential when working with any inductors/coils that will be switched on/off.

    The 10k from gate to Source/Ground is to hold the line low when no signal is present. Internally, a MOSFET doesn't have a connection between Gate and either source or Drain. The drain is connected to the coils to the power, so switching the MOSFET "ON" will energize the coil for a period of time, then turn off the MOSFET.

    I guess how the uC is determining what is "long enough" for the counter to work may have something to do with the reason it fails when the counter is absent.
     
  11. montner

    Thread Starter Member

    Mar 12, 2009
    10
    0
    OK, I understand about the diodes and their function now. Thank You! And there is no physical connection between gate and drain, but there is reverse drain capacitance Crss = Cgd of 71pf according to the datasheet. Will the 10K resistor null the effect of this?
     
  12. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718

    There is a gate charge that takes a while to build up before the MOSFET is ON, this is shown in the datasheet. It is conceivable that the processor could do some fancy footwork in software to guess when current is flowing based on the time for the MOSFET to turn "on", but that would be sort of odd.

    Is there perhaps a hall effect sensor next to or under the coils? Typically look like a transistor and respond to magnetic fields to give feedback. I'd rank this solution as more likely than attempting to toggle between input, tristate, and high on the uC's output pins to essentially measure gate capacitance.

    I am assuming there are only the two coil connections to each counter, and no other electrical connections.
     
  13. montner

    Thread Starter Member

    Mar 12, 2009
    10
    0
    There are no hall effect sensors or other connectors to anything. I am stumped. I guess I will remove the mosfet from my existing board and try my solution to see if it works as well as some others before I put it on my new board design and have them made.
     
  14. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Just make sure you don't put over 5V on the line to the uC. Possibly even over 3.3V depending on the logic Power Supply. Accidentally putting the 12V on an I/O pin would pretty much fry the entire circuit.
     
  15. montner

    Thread Starter Member

    Mar 12, 2009
    10
    0
    If I put a capacitor and diode between 12V and the gate line, like I drew before, will it show too much voltage and hurt the uC?

    Sorry I am so ignorant, but I learn fast.
    Thank You for your help.
     
  16. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    The capacitor will block DC, except at power on. No real use for a standard diode as shown in 2nd schematic above.

    In the current configuration, the uC never sees the 12V, at most, it can only detect how long it takes for the MOSFET to turn on, so there is no need to have the 12V line, just add the capacitor in parallel with the resistor from gate to ground.

    I'm not sure that the wild guess is how the circuit is detecting the coil, though, so don't make any permanent changes until you have a solution. Somebody else may have an idea or something obvious that I've completely forgotten about.
     
  17. timrobbins

    Active Member

    Aug 29, 2009
    318
    16
    Montner,
    You could try temporarily insering a resistor in place of the coil of the counter to see if that fools the uP - it may assist in indicating the feedback mechanism.

    Are you sure there is only a single coil connection to the counter - no other contacts or ?
     
Loading...