12v lead acid backup circuit

[desmondlai]

when i construct this circuit in pcb i cant get 14.1v and 13.4v at the point as i stimulate this circuit .... the difference of the pcb and the stimulate is the 7805 and 7812 , i replace with 78t05 and 78t12 that 3A ...
how can i get back the 14.1v and 13.4 v without changing the both 78t05 and 78t12????????

 

[Hero999]

This is why you should never rely on silly simulation programs.

There are a few things wrong with that circuit:

U2 will go into drop out, you won't get 12V on the output but 10V at full load.

C1 is probably not large enough to avoid ripple.

You need to use a higher voltage transformer, I'd suggest 15V. The simulator doesn't simulate the transformer's series resistance or regulation factor so the output voltage will be higher with no load than when fully loaded. The sharp peaks drawn by the filter capacitor charging can cause the voltage to drop below the 16V, the minimum voltage required to keep the regulator regulating properly.

 

[crutschow]

Why would you call a simulation program "silly"? Besides the human brain, It's the most useful tool in designing circuits there is. Of course you have to understand their limitations and recognize if it is producing unrealistic (silly) results, but I would never build a circuit without first simulating it. Simulation can catch many subtle errors in the design that otherwise you wouldn't find until after the circuit is built. It's also a great tool for just learning about circuits.

Simulation of the transformer regulation just requires adding a resistor in series with windings whose value can be obtained from the transformer data sheet or measuring the winding resistances with an ohmmeter.

 

[Hero999]

But he hasn't done that.

Unless there's a hidden property on the AC supply he's set for series resistance and hasn't told us about?

I partly take back what I've said: simulation is useful but just blindly relying on it, is bound to get you into trouble.

 

[MikeMl]

...
I partly take back what I've said: simulation is useful but just blindly relying on it, is bound to get you into trouble.

Just because you can make a circuit work during simulation is not a guarantee that it will work when you physically build it, usually due to not accounting for real-world behaviors and parasitic effects (i.e. model is not accurate enough).

OTOH, if simulation predicts that a given circuit will not work as intended, then there is not much to be gained by actually building it!

 

[crutschow]

But he hasn't done that.

Unless there's a hidden property on the AC supply he's set for series resistance and hasn't told us about?

I partly take back what I've said: simulation is useful but just blindly relying on it, is bound to get you into trouble.

Well, certainly any tool can get you in trouble if you blindly rely on it. That's where the learning part comes in. The preference is to learn how to use the tool properly and recognize when it is giving bum information otherwise. I believe the purpose of this forum is to help people use such tools correctly, not imply that they can not be relied on.

I have designed and simulated many analog circuits with SPICE and it almost always has given a close simulation to the actual circuit operation. It's particularly usefully when trying to determine the effect of component tolerances on circuit operation. The differences between simulation and actual are often due to parasitics, but part of the learning curve is to know when and where these need to be added to the simulation.

I know he didn't include any resistive characteristics of the transformer which is why I suggested a way to do so.

 

[Hero999]

OTOH, if simulation predicts that a given circuit will not work as intended, then there is not much to be gained by actually building it!

There are circuits that work in real life but not simulation: for example try the reverse bias oscillator in most simulation programs and you'll find it won't work.

 

[MikeMl]

There are circuits that work in real life but not simulation: for example try the reverse bias oscillator in most simulation programs and you'll find it won't work.

The reason that it will not work in simulation is because the standard semiconductor device models do not account for some unusual (un-natural?) behaviors which occur in the reverse-bias region. This is not an indictment of simulation or a particular simulator; rather it is a consequence of exploiting behaviors that are not supported, characterized, or tested by the semiconductor device manufacturer. Spice Models supplied by the manufactures do not model this behavior because they do not test for it...

Try designing a circuit that exploits these second-order behaviors, and then building it in large quantities, where you might have to buy semiconductors from multiple batches or even multiple vendors which second-source standard part numbers. Now, when the circuit doesn't work due to a new batch of devices, try going to the vendor to complain about their devices