Forums » CircuitLab Feature Requests

Relays should feature 'on voltage', 'hysteresis voltage' and 'switching time' parameters.

+1. Lack of switching time parameters is causing my modeled circuits to significantly differ from reality.

If you're really dependent on the accuracy of the relay model then you might like to look at:

http://www.intusoft.com/articles/relaycd4.pdf

With some thought you can implement everything they are discussing there, in CL.

Note that although CL says in:

https://www.circuitlab.com/docs/circuit-elements/#laplace

that 1/s in the Laplace block creates an integrator, in practice a small constant is also required for the simulation to work.

See the use of the Laplace block to make an integrator in:

https://www.circuitlab.com/circuit/f73b76/modulated-sources-01/

Hmmm, um ...

A relay, as simple as it seems to be, is a very complex electromechanical device, thus I’m not sure if you ever could simulate it using CL. I guess it would require different models for the various existing types of relays / circuit breakers.

Also the behavior depends on the application so you could probably model a “standard” (test) situation for one relay but not the truth (the reality) in your application.

However, what I understand from the feature suggestion would be a slight improvement of the just too simple model in CL, a kind of delay between energizing, break, make and back to off (could be the same delay to keep it simple, just two parameters).

One could adjust these parameters from measuring the real target device or just guess a few ms, this would be much better than CL’s zero - model.

Relay theory may be endless but accuracy is not the concern here in CL (just my feeling).

@Signality: Regarding your Laplace example I have some questions which I’ll try to post tomorrow.

Regards, Sancho

Sancho,

As you say, I am not suggesting or interested in (however other people may be) properly modelling a relay.

This would be impractical anyway since you would need to know contactor masses, spring constants etc. which I doubt one would find in a data sheet :)

A data sheet however does [normally] define a 'switching time', 'on voltage' and 'hysteresis voltage' etc. Which I would assume would be fairly simple to replicate in CL (not that I know how it is programmed), since we only need to 'simulate' the relay rather than model one.

Also, adding an inductance parameter would be nice too. Really, we only need to have the relay look to CL like a series RL combination.

Regards, Josef

Scrub my comment about inductance, I wasn't paying attention :)

Wow this is an old thread, but I really could use CL more if the relays could be customized. I'm not interested in detailed physical attributes like speed, but I can't model a real world circuit because I can't customize at what voltage or current the relay would trigger or release.