Transient response of semicinductor switch

Discussion in 'The Projects Forum' started by paw1, Feb 6, 2015.

  1. paw1

    Thread Starter Member

    Jan 13, 2015
    32
    0
    I want to simulate the transient response of a high frequency (~100 MHz) sine wave when a semiconductor switch (SPDT) changes its output path.

    I made a switching circuit in LTspice IV, but it resulted in an ideal response (no transients).

    I also tried to make a similar circuit in Simulink, but I couldn't even make it run due to the error:
    "Each physical network must be connected to exactly one Solver Configuration block. There is no Solver Configuration block connected to Physical Network with the following blocks: ...". I included the Solver Configuration block in the schematic, but it wouldn't connect to the circuit since it had the wrong connector.

    Best regards
    -paw
     
  2. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,980
    3,706
    If you share a schematic of what you are thinking about, we might be able to help you.
     
  3. paw1

    Thread Starter Member

    Jan 13, 2015
    32
    0
    It's something like this:
    System.png
    I want to simulate the transient response when the switch is turned to the "On" position.
     
  4. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,980
    3,706
    So, let me understand...
    How fast are you switching the SPDT?

    Why are you running a 100 MHz signal at 10 amps directly into ground (short circuit?) when the SPDT switch is off.

    Where do you live that a circuit like this will be useful or interesting? That is, a 100MHz signal driving a load that draws 10 Amp will cause all kinds of interference in the FM radio band.

    Is the 100MHz a carrier wave and the BP filter passes another frequency or is it simply passing the 100MHz input signal?
     
  5. paw1

    Thread Starter Member

    Jan 13, 2015
    32
    0
    The SPDT will be switched about every half millisecond. The signal will have other frequencies as well, that's why the BP-filter is there.

    It was stupid to draw a ground in the 'Off'-position. What I actually meant was 'Ignore this part of the circuit' as it doesn't matter. I also meant 10 V not 10 A (haven't had any coffee today).

    The only thing I'm interested in is to see the transient response when the switch goes to 'On'-position.
     
  6. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,281
    1,232
    Did you actually build the filter in spice?
     
  7. crutschow

    Expert

    Mar 14, 2008
    12,976
    3,220
    What "transients" are you expecting to see?
     
  8. paw1

    Thread Starter Member

    Jan 13, 2015
    32
    0
    I expect to see transients of higher frequencies (than 100 MHz) just after the switch has changed position, as I've been told is the real world situation with these kinds of switches.
     
  9. paw1

    Thread Starter Member

    Jan 13, 2015
    32
    0
    The filter is being built, but it still needs improvement. The filter is not of any importance in this thread though.
     
  10. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,281
    1,232
    If it is a passive filter I don't think you will see any transient. If it is active there might be a better possibility of some depending on what it looks like.
     
  11. crutschow

    Expert

    Mar 14, 2008
    12,976
    3,220
    That transient is due to real-world parasitic capacitance and inductance in the circuit, which are layout and component dependent. So you need to add an estimated value for those in you circuit if you want to see the high frequency transient.
    Some parasitic capacitance will be introduced if you use a MOSFET for the switch, for example.
     
  12. paw1

    Thread Starter Member

    Jan 13, 2015
    32
    0
    That's the kind of answer I was looking for. Maybe using a MOSFET could get me there simulation-wise. The problem is that it's hard to find good switch-models for LTspice.
     
  13. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    And if the filter is low-pass, or band-pass, it will attenuate any transients caused by the switch. Also note that the magnitude of any transients is greatly dependent on the phase of the RF at the instant the switch closes. If the switch is a device like a PIN diode, then you can also shape the switching signal to limit the rate at which the switch turns on, which is what is done in most VHF transceivers that use PIN diode T/R switching.

    You would only use the LTSpice switch if you are modelling a relay. Exactly what is doing the switching?
     
  14. paw1

    Thread Starter Member

    Jan 13, 2015
    32
    0
    One of the functions of the BP-filter is to attenuate these transients, along with harmonic signals. I need to simulate (document) this because it's going to be part of a report for a school project. If possible, I want to simulate the transient response at several points through one cycle.

    The switching will be done by a semiconductor device. Most probably a SPDT-switch.

    Thanks.
     
  15. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,766
    1,101
    Have you tried searching semiconductor-SPDT-switch-manufacturers' sites to see if they have a Spice model available for download?
     
  16. paw1

    Thread Starter Member

    Jan 13, 2015
    32
    0
    Didn't know manufacturers did that. Thank you for telling me.
     
  17. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    I'm puzzled about the semiconductor switch? Do you have a part number or data sheet?
     
  18. paw1

    Thread Starter Member

    Jan 13, 2015
    32
    0
  19. crutschow

    Expert

    Mar 14, 2008
    12,976
    3,220
    That part has a very high frequency response so I would expect that any significant switch transients will be the result of the external parasitic inductance and capacitance, which is largely layout dependent.

    Are you going to use microstrip or stripline (transmission line) design of the circuit board?
    That would significantly reduce any transients.
     
  20. paw1

    Thread Starter Member

    Jan 13, 2015
    32
    0
    I'm not sure if stripline will be possible, so I guess microstrip is more probable. Although, I must admit, I haven't really thought about interconnecting possibilities yet.
     
Loading...