-1

I have created the below circuit that simulates the current characteristics for a contactor. However I am getting this huge spike when the switches transitiion.

Can someone help to prevent this spike from happening?

enter image description here

enter image description here

enter image description here

enter image description here

JoeyB
  • 1,804
  • 13
  • 34
  • 3
    One problem is real capacitors and inductors aren't ideal especially as you increase frequency. Electrolytic capacitors designed for 60 Hz often look more like inductors at 40 MHz. Inductors have capacitance between the coils. Thus, improved modeling considers the parasitics of the components. – Perry Webb May 22 '22 at 12:19

1 Answers1

2

If all you want is an exponentially-shaped pulse then you don't need all that contraption. Just use a simple current source (F2 > current) set to be exponential with EXP(0 3.2 0 5m 30m 1.875m) as value. Do you have a reference for that contactor? Or even for the claim that this is the current shape of a contactor?

As for the spike, that's due to the misalignment of the threshold vs the two pulse ramps.

BTW, in your picture the controlling voltage sources have no ground. Behind the scene, LTspice may try to prevent shooting yourself in the foot, but it's not something that should be omitted. Also, I'd avoid non-negative thresholds, especially if this is meant to be a .model of some sort.

a concerned citizen
  • 21,167
  • 1
  • 20
  • 40
  • "As for the spike, that's due to the misalignment of the threshold vs the two pulse ramps." I tried adding a delay, so SW3 turns on after 1m SW3 goes off but this just creates a drop in current for that time. How can I align this? – JoeyB May 22 '22 at 12:45
  • @JoeyB, is your contactor designated as break-before-make (likely) or make-before-break (unlikely, but possible)? You should not align the transitions of the switches in your model, you should make it follow the physical behavior of your contactor. – The Photon May 22 '22 at 14:41
  • @JoeyB I've given you a solution that does not involve anything else than a current source. Do you have a reason why you would want to complicate with your schematic? All the more since, not only it will make "matching" difficult, but it will only achieve the *exact same results* as my solution. Also, I meant `.subckt`, not `.model`, but it's a minor mistake. – a concerned citizen May 22 '22 at 14:52
  • This is not for any particular application as yet. I just wanted to learn how to do it, for future applications – JoeyB May 22 '22 at 23:25
  • 1
    @JoeyB Based on [your previous question](https://electronics.stackexchange.com/questions/619777/model-positive-exponential-current-curve-on-ltspice) and your comments here, I believe we're in [XY Problem](https://xyproblem.info/) territory. Perhaps you should explain what exactly you want to achieve/learn? Learn how to do "what" for future applications? – Ste Kulov May 23 '22 at 02:40