I'd like to use LTspice to simulate how a mechanical switch produces a spike in voltage when it is opened while driving an inductive load.
Can it be done?

 

I'd like to use LTspice to simulate how a mechanical switch produces a spike in voltage when it is opened while driving an inductive load.
Can it be done?

I haven't used it myself, but you might try the sw (voltage-controlled switch) component, driven by a PWL or PULSE voltage source.

 

LTspice has a voltage controlled switch model (sw) that will certainly generate a spike with an inductive load.
You need the sw model discription for the switch to work (look up "sw" in the help file).
You add it as a Spice directive (hit the S key while in the schematic window) .

 

Here is something I did some time ago...

 

Very, very interesting. Thank you both.

I just found the Vswitch.asc example included in LTspice and am finding it quite useful.
My only issue with it is that it's a voltage controlled switch... is there no other kind of switch that could be timed during the simulation? But maybe that's not important since I can always configure a voltage source to drive it.

Thanks again!

 

Here is something I did some time ago...

View attachment 84056

Excellent file Mike! Thanks for posting.

 

...
My only issue with it is that it's a voltage controlled switch... is there no other kind of switch that could be timed during the simulation? But maybe that's not important since I can always configure a voltage source to drive it.

It is easy to make the switch state a function of time, a voltage, a current, etc...

 

It is easy to make the switch state a function of time, a voltage, a current, etc...

Would you please share with me your definition of easy? I'm a neophyte in most LTspice-related things... need to find an example of making a switch a function of time... or maybe I just need more experience with the spice-directive instruction.

 

The sim I just posted makes the switch state a function of time because the controlling voltage V(pulse) is a function of time. By default, the switch is closed when the input voltage is >0V, and is off if the input control voltage is <=0V. If you prefer it to have a different Threshold voltage, you can set it in the .model SW statement. You can model a switch with hysteresis, too.

 

The sim I just posted makes the switch state a function of time because the controlling voltage is a function of time. By default, the switch is closed when the input voltage is >0V, and is off if the input control voltage is <=0V. If you prefer it to have a different Threshold voltage, you can set it in the .model SW statement. You can model a switch with hysteresis, too.

Yeah... that's more or less what I meant on post #5 ... but you've given me even more information now, thanks.

 

Hello all... I just found the time function in LTspice, and I'm using it to change resistor or pot values (or any value for that matter) during the simulation. In the case of a resistor, for instance, one only has to change its value to:

R=if(time<10m, 5, 15)

Alternative:
R=table(time,0,5, 10m,5,10.01m,15, 20m,15, 20.01m,25)

You can make nested if() of course.
R=if(time<10m, 5, if(time<20m,15, 25))

EDIT: will next try it with the switch

 

I'd like to use LTspice to simulate how a mechanical switch produces a spike in voltage when it is opened while driving an inductive load.
Can it be done?

Hi cmartinez,

I'm trying to make the same thing with LTSpice, could you please tell me how you did it? Or even better, could you send me the .asc file? It would help me a lot!

 

Hi cmartinez,

I'm trying to make the same thing with LTSpice, could you please tell me how you did it? Or even better, could you send me the .asc file? It would help me a lot!

I'm afraid this thread is quite old, and I can't remember if I solved this problem or not. Besides I can't find an asc file in my small collection related to this subject.

But I know a bit more about LTSpice by now. There are several ways of approaching this problem. One would be to use a standard voltage controlled switch, which is already a part of the LTSpice library of components.

​

And I'd use a voltage source to control that switch, with its state defined through values in a time table. Another method would be to use an ordinary resistor, with its value defined as a function of time (as explained in post #11), using an extremely low value (like 0.01 ohms) when it's closed, and a very large one (like 100 mega ohms) when it's open.

 

Hello lettopina,

Fortunately, we still have this.

 

I use the volt controlled switch in LTspice all the time for mechanical switches. I usually set the model something like

.model switch SW(Ron=10m Roff=1G Vt=0.5)

That gives 10mΩ contact resistance, 1GΩ isolation resistance, and a voltage trip point of 0.5V. Then for the voltage control I use the PWL type of voltage source to go from 0 to 1 at whatever time I want to switch it, like

PWL(0 0 1m 0 1.001m 1)

That's a transition from 0V to 1V over 1μs, switching at 1ms of simulation time. If I want to simulate switch bounce, I either use a multi-transition PWL voltage source (for psuedo-random bouncing), or you can use a pulse voltage source with the delay set to the time you want it to come on and a finite number of cycles. If your mechanical switch has bounce time specified, you can set the frequency and number of cycles from that. Sometimes they're not specified, so I set it somewhere around 5-8 bounces in a 5ms time, based on specs I've seen before for specific switches.

 

Be aware this is a resuscitated 3-year old thread.

 

One curiosity

Be aware this is a resuscitated 3-year old thread.

Yes, but it was by the TS.

 

Hello lettopina,

Fortunately, we still have this.

Sadly he was last seen over 7 months ago

One curiosity
Yes, but it was by the TS.

I don't think so. It was @Seamus_One who brought it back to life. @cmartinez was the TS

 

I don't think so. It was @Seamus_One who brought it back to life.

Right.
I look at the wrong thread.

 

Right.
I look at the wrong thread.

That's OK. It's happened to me on more than one occasion, and you're entitled to an occasional pass.