Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Problem with Pspice simulation

  1. Sep 8, 2015 #1
    Hi,

    I have a strange problem with Pspice, it always shows after successful attempts, the simulation program stops working and all voltages markers turns to gray and cant see any results any more Pspice capture.PNG Sche.PNG
     
  2. jcsd
  3. Sep 8, 2015 #2

    berkeman

    User Avatar

    Staff: Mentor

    What happens if you simplify the output circuit to just a single stage of the multiplier?

    Also, what does it mean on the left where it says L1=L1, L2=L2, L3=L3?
     
  4. Sep 8, 2015 #3
    What happens if you simplify the output circuit to just a single stage of the multiplier?

    it works, but as soon as i copy past some diodes (i change the name and the number of of course) i get this problem,

    Also, what does it mean on the left where it says L1=L1, L2=L2, L3=L3?[/QUOTE]

    that's an alternative way how a transformer is built in pspice
     
  5. Sep 9, 2015 #4

    Baluncore

    User Avatar
    Science Advisor

    Can you attach the ascii node source file to a post. Then I can run it without having to start from scratch.
    Pspice has some bugs. Have you tried to run it on LTspice?
     
  6. Sep 9, 2015 #5
    no i did not :/
     

    Attached Files:

  7. Sep 9, 2015 #6

    Baluncore

    User Avatar
    Science Advisor

    There is no circuit in that file.
    It should have one line per component that specifies nodes component name and value/model. For example; C3 is connected between ground node 0 and another node, maybe 3, it is called C3 and has a value of 1nF.
    0 3 C3 1n
     
  8. Sep 9, 2015 #7
    sorry
     

    Attached Files:

  9. Sep 9, 2015 #8

    Baluncore

    User Avatar
    Science Advisor

    Not useful. Please look in the file before you post it to verify it is the raw spice node code.
     
  10. Sep 9, 2015 #9
    sorry but i found nothing with that format
     
  11. Sep 9, 2015 #10
     

    Attached Files:

    • z.PNG
      z.PNG
      File size:
      8.9 KB
      Views:
      112
  12. Sep 9, 2015 #11

    berkeman

    User Avatar

    Staff: Mentor

    In PSPICE, is there an option to Export a SPICE Deck? That would be the ASCII file that we can try to run ourselves.
     
  13. Sep 9, 2015 #12
    Try shorting the caps with a high-ish value resistor - like 1 meg or 10 meg.
    SPICE goes a little unstable sometimes when you have a long capacitively coupled circuit like that.
     
  14. Sep 9, 2015 #13

    berkeman

    User Avatar

    Staff: Mentor

    I just noticed that C3 and C13 appear to be shorted in the schematic. That may not be the cause of your problems, but you should fix that connectivity at some point. :smile:
     
  15. Sep 10, 2015 #14
    i have a trial version of Pspice, i cant :(
     
  16. Sep 10, 2015 #15
    ok thanks :smile:
     
  17. Sep 10, 2015 #16
    ok , im gonna try it
     
  18. Sep 10, 2015 #17

    Baluncore

    User Avatar
    Science Advisor

    I used LTspice.
    Schematic.png Traces.png
     
  19. Sep 11, 2015 #18
  20. Sep 11, 2015 #19

    Baluncore

    User Avatar
    Science Advisor

    I think it might depend on how you define the driving voltage.
    For sine 12V, LTspice generates; V = 12 * sine(wt) i.e. 24 Vpp = 8.5 Vrms. What does Pspice do ?

    You run at 500 kHz, I run at 1kHz. Neither has any load so we are both looking at electrostatic voltage.

    You show 8 testpoints on your schematic, but you only show 7 traces on the plot.
    Your highest output voltage is not being displayed. Has the EHT shorted another of your capacitors or diodes ?
     
  21. Sep 15, 2015 #20
    hi,

    im using Ltspice now :), im trying to use C19 as Dummy load ( from the german word Ersatzlast) but it does not work
    Multi.PNG
     

    Attached Files:

  22. Sep 15, 2015 #21

    Baluncore

    User Avatar
    Science Advisor

    Attached is my LTspice Vmult3.asc file. Remove the .txt suffix to run it.
    You can attach your .asc by adding a .txt suffix.
     

    Attached Files:

  23. Sep 15, 2015 #22

    Baluncore

    User Avatar
    Science Advisor

    It will take 500,000,000 cycles of your oscillator to charge the C19 = 500uF through 1pF multiplier capacitors.
    At 1.5MHz that will take 333.4 seconds, well over 5 minutes.
    Your simulation runs for 1 millisecond which is why it does not appear to work.
     
  24. Sep 21, 2015 #23
    thanks :), i need C19 to be charged in less than 15 sec , if i put a 35 MHz it schould works right ?
     
  25. Sep 21, 2015 #24

    Baluncore

    User Avatar
    Science Advisor

    Probably NOT.
    It will be difficult to simulate that with any version of spice. 15 seconds at 35MHz is 525 million cycles. It will take hours to simulate.

    You have not specified the design constraints. What output voltage do you need ?
    You have not explained why your transformer ratio is what you have selected.

    It was a wise move to convert to a single phase multiplier, as it needs only 50% the number of diodes and uses 33% less caps, and it only requires twice the frequency.

    In your previous circuit you showed one diode only as an MMSD4148. What diodes are you really using ?
    Are you building an IC ? That is the only way I can see you using 1pF capacitors.
    Your 1pF capacitors have a similar capacitance to 1N4148 diodes. That is really not good.

    1nF capacitors are available that are both small and cheap.
    The bigger the pump capacitors the lower the frequency needed to pump the charge.

    You keep changing the circuit without justification.
    If you have an LTspice.asc file please post it so I can run and modify it without a waste of time.

    The Villard Cascade, voltage multiplier is probably not now the best choice for HV generation.
    What is your application. Are you building a taser or a cattle prod ?
     
  26. Sep 21, 2015 #25
     

    Attached Files:

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook