Do you guys agree with this?

 

No, it is flawed.

You have several parts that do not meet specifications for the application, including the regulator itself.

D1 is not rated for the current, and the reverse voltage could be exceeded during load dumps.

F1 is too large.

The 10R resistor is too large, and too low of a wattage.

D2 would get fried during a voltage transient.

The regulator would then get fried, as the max input voltage would then be exceeded.

 

Try using a Simple Switcher. You can easily get 3A out of one, even with a tiny heatsink (at low currents, no heatsink at all is needed.)

 

Question is: how to have the over voltage clamping protecting with the simple scheme of relying on the voltage drop across a resistor to switch on Q1?

 

Try using a Simple Switcher. You can easily get 3A out of one, even with a tiny heatsink (at low currents, no heatsink at all is needed.)

Ok I try using a simple switch-mode too.

 

Question is: how to have the over voltage clamping protecting with the simple scheme of relying on the voltage drop across a resistor to switch on Q1?

Over-voltage protection uses the zener and fuse.

In reality the zener will get very hot and fail and the fuse will not blow if an over voltage transient occurs. You need a TVS device that is designed for high surge currents.

 

Over-voltage protection uses the zener and fuse.

In reality the zener will get very hot and fail and the fuse will not blow if an over voltage transient occurs. You need a TVS device that is designed for high surge currents.

I thought it used the combination of the zener plus the resistor? Technically termed 'crowbar' protection in textbooks.

For the resistor, the theory is that current of 600mA passing through it will cause a voltage drop of 0.6 across it, therefore forward biasing Q1 into conduction to series pass current into the load.

So a 1Ω resistor?

 

ok i try using a simple switch-mode too.

lm2576-5.0 switcher

If you do use a linear, the pass transistor will be burning over 21 Watts.

 

I thought it used the combination of the zener plus the resistor? Technically termed 'crowbar' protection in textbooks.

For Q1, the theory is that current of 600mA passing through it will cause a voltage drop of 0.6 across it, therefore forward biasing Q1 into conduction to series pass current into the load.

So a 1Ω resistor?

That is crowbar protection. The problem is, a zener won't conduct enough current to blow the fuse unless it's designed to do so. Most are designed to conduct a few mA. You could use a 16-20V TVS device. These are designed like zener diodes but can absorb huge surges without failure.

Q1's purpose there is to provide extra output current; it is controlled by the regulator.

 

Ahh I see. I forget about the fuse for a moment.

 

Agreed?

 

Ahh I see. I forget about the fuse for a moment.

TVS device: http://parts.digikey.com/1/parts/719517-diode-tvs-16v-600-watt-bi-dir-p6ke16ca.html

Capable of dissipating a 600W surge, that's 37.5 amps peak. If the fuse blows in 1 second at 4A, it will blow in about 11.3 ms at 37.5A.

 

TVS device: http://parts.digikey.com/1/parts/719517-diode-tvs-16v-600-watt-bi-dir-p6ke16ca.html

Capable of dissipating a 600W surge, that's 37.5 amps peak. If the fuse blows in 1 second at 4A, it will blow in about 11.3 ms at 37.5A.

I studied this like 20-years ago @ trade level. Have never practically applied it though.

 

I studied this like 20-years ago @ trade level. Have never practically applied it though.

I^2 * t is a good approximation. If 4A blows it in 1 second, it uses 16 "fuse units" of life. (I forget the units!) 37.5A^2 * 0.0113 also uses 16 lives, so the fuse blows as well.

 

lm2576-5.0 switcher

If you do use a linear, the pass transistor will be burning over 21 Watts.

Are you certain it is simple to use?

 

I^2 * t is a good approximation. If 4A blows it in 1 second, it uses 16 "fuse units" of life. (I forget the units!) 37.5A^2 * 0.0113 also uses 16 lives, so the fuse blows as well.

So it has a working life is this what you mean?

 

So it has a working life is this what you mean?

During a failure, yes. But if it's allowed to cool down a fuse will often survive a high current event. Eventually metal fatigue from thermal expansion may get to it.

I've repaired a lot of LCD TVs recently where a specific fuse on the video control board blows. Replacing the fuse fixes it. I'm guessing it just wears out after being used near its rating for a long time.

 

During a failure, yes. But if it's allowed to cool down a fuse will often survive a high current event. Eventually metal fatigue from thermal expansion may get to it.

What have you got a science degree or something? I mean I did 2 to 3 learning all this stuff and we never went this far into it.

 

Forget the LM7805.

Use a LM2576 Simple Switcher SMPS Regulator, uses fewer extra parts than your 7805 design.

3A capacity/5A peak, use automotive rated components for caps and inductor, hot glue parts to board to keep vibration from ripping heavy parts from board, add inline PTC fuse and 0.1uF input-> GND cap as well as a Schottky diode in series with a 1-2 Ohm resistor to input for transient protection.

 

Yeah inductors. Inductors are the only electrical entity that cannot be intergrated into a chip.

Shame about that.