I imagine there are a lot of problems with this circuit - it's just a conceptual idea right now, that may be way off. Having said that, I can't even get LTspice to run and show me how it fails to work, blows up, whatever. The added component data for the SCR is from here:

http://www.littelfuse.com/technical...technical-center/spice-models.aspx#question_3
SCR SPICE Model Sxx40x
SPICE_Littelfuse_SCR-Sxx40x_A_lib.htm

I've only done a few other LTspice simulations, and none of the others have included any parts that didn't come with LTspice, so I'm willing to bet this is user error on my part - doing something wrong with how I included this part model. The error I get when I try to run a simulation is:


And here's the circuit I'm trying to simulate.


Again I'll say that I was in a rush and just wanted to see if I could get it to run at all for now. I have yet to calculate proper values for most of the resistors. Also, I plan to use a different opto (Fairchild Semiconductor MOC3063SR2M) but couldn't find a model for it either. If I ever get this sim to run, it'll probably kill the opto immediately unless I change the operating voltages. Ignoring component values, etc. any idea why it won't even run?

Thanks in advance for any insights!

 

It is not only the absence of convergence.Your electronic circuitis no good.If the make and include your scheme,it immediately fails.There is no need in four thyristors.Here's how you can build an AC switch:

 

All right, I'll play around some more with circuit ideas. I had expected LTspice to simulate the circuit even if it was a stupid circuit, at which point I would see the results, even if it was a dead short, wild oscillation, etc. Sounds like maybe that's not the case.

As for the 4 thyristors, that's because I want to switch both legs of power, not just one. I'm essentially trying to make a double pole SSR. The few that I've found ready-made are very expensive, more than seems appropriate considering how inexpensive regular SSRs can be. Using two separate SSRs would work of course, but is quite bulky. Back to the drawing board!

Thanks for sharing your circuit. Maybe I can experiment with that as a starting point.

 

The root cause is that there is no ground in the "isolated" part of the circuit.

Trick: connect one of the nodes on the isolated side to ground through a 1GΩ resistor.

 

The root cause is that there is no ground in the "isolated" part of the circuit.

Trick: connect one of the nodes on the isolated side to ground through a 1GΩ resistor.

Thank you so much! I've read about that issue before and totally forgot about it. It's running now! (Of course my circuit still needs all kinds of work, but that's a different problem entirely. When I get to where I need advice on that, I'll be back.)

 

So, I've got a new version of my idea sort of working in LTspice. The goal is to provide SSR style switching to BOTH legs of power to a 240VAC 1/3HP motor. I have absolutely no idea how to simulate the inductance (and therefore the inrush current and back EMF) of this motor. The resistor and inductor shown in my current sim were chosen pretty haphazardly and adjusted until the running current was in the right range, but I have no idea if this is anything close to representing other characteristics of the motor. I'm sure a really good simulation is a nightmarish undertaking, but if there's a simple way to at least get in the ballpark, that would be great. Motor specs, current circuit design, and LTspice files are below.

Thanks!

Motor specs: View attachment 87997
Current circuit:

 

Every thing works one TRIAC.You probably want to increase electrical safety?

 

Every thing works one TRIAC.You probably want to increase View attachment 88277 electrical safety?

Right, I want this to be setup such that, if the motor were chopped off and its power leads were dangling in the air, there would be no shock hazard when the triacs were off. (Obviously in the chopped cord scenario, there would still be a shock hazard when the triacs are activated.)

This circuit is intended to replace double pole mechanical relays, which provide the safety and reliability we want, but allow the motors to create huge amounts of noise when switching on and off. (The noise issues have been greatly reduced with improved grounding and the addition of snubbers, but the noise is still present at undesirable levels, so we'd prefer to eliminate it entirely, which SSRs seem to do.)

So, any advice on how to better simulate the motor that I'm switching with this circuit?

 

In your sim, R4 and R6 are 500m. Did you intend 500Meg?

 

In your sim, R4 and R6 are 500m. Did you intend 500Meg?

Yes!
I must say, though: V1 is a pretty incredible power source, dumping all that current to ground and still no voltage sag!

 

Yes, megaAmps and megaVolts are no sweat for ideal components .

That 'm' is a common gotcha. Another one is using 'F' for Farad, as in '1F' (which turns a supercap into a 1 femto-Farad blob of nothing).