Hi Guys,

My application requires power be maintained by automatic switching from and to a secondary power source (AC/DC adapter) in the event of failing solar power.

 

First, I didn't like the way it was drawn, so I redrew it. Second it is a really dumb circuit, and has a serious design flaw that would instantly vaporize the transistors. I fixed that problem, and simulated it, and am comparing it to my suggested circuit:



I am plotting the voltage across the load in both circuits [red=V(load1), green=V(load2)]as the simulated solar panel voltage goes downward from 20V to zero. Note that as the solar voltage goes below 12v, the battery supplies the load current [same as blue= I(D2)].

Note the lower parts count, lower power wasted, higher output voltage for the circuit on the right. You could improve it further by using Schottky diodes in place of the Silicon ones.

 

Hi Mike,

I am glad of your audit, and would love the ability to come to the same conclusion, hence my determination to learn.

Only today my transistor learning curve got heavier in that have I downloaded LTspice to perfom simulations and will now spend time learning spice hoping to throw light on how the original circut functions.

Your load 1 solution has revised components & count; I assume to meet application parameters and aviod smoke; however your simulation gave a result so the design "works???"................with what I percive to be 11.2v low battery overtones.

My current application supplies power to a 12v nominal solar charge contol rated 5A and it is the power source to this device that I would subjet to crossover actions.

Your response allows me to build a baseline "known good " simulation and compare performance with any amount of postulations

I regard myself further along having made this request so again many thanks for your input

nth

 

...
Your load 1 solution has revised components & count; I assume to meet application parameters and aviod smoke; however your simulation gave a result so the design "works???"................with what I percive to be 11.2v low battery overtones...

Note that the diodes shown will not stand 5A. The original circuit had R3 in the wrong place, but with my change would "work" as shown, with the higher voltage drop across the transistor added to the diode drop. The diodes do the "switching" by themselves with no help from the transistors...

 

Hi Mike,


In practice I will implement your load 2 solution however I will investigate your load 1 layout with diodes omitted in order to gain better understanding of functionality’s in this circuit.

My confusion stems from observation and gives rise to the following:

The diodes do all the "channeling" - why have the transistors at all if they are redundant

A BPJ with sufficient forward bias enters saturation, functioning as a closed switch i.e. allowing the controlled current to pass, this means current is present when I would have no current

With the solar control “dark” i.e. no power; the switch is open and will not pass current when I would have it do so.

If the above is true -why post the original circuit as a solution to "automatic transfer from multiple power sources".

For completeness my 5A value is a design maximum of the PWM solar control and not the "pass thru' amps" supplied from my 20Watt array.

Again many thanks for your input- today is a spice day and hopefully illuminating

Best regards
nth

 

...
If the above is true -why post the original circuit as a solution to "automatic transfer from multiple power sources".

Because any idiot that knows a bit of html can put up a web page. Just because someone is computer literate enough to create a web page does not mean that they are an electronics guru... In my experience, about 80% of the circuits posted on the web are created by newbies that really cant be trusted...