I am burning up servos with this buck circuit and don't see the problem. When the servo goes the buck chip caught fire
see attached.

 

The info you provided isn't enough. We need to know what the power supply is that is powering the buck circuit, and we need to know what the load is that the servos represent. We need specs for both sides of the circuit because the fact that it is burning up means it's overheating- too much current. Exceeding specs of circuit.

 

Pins 7 and 8 of the NCP3163 sense the current and shut it down if there is an overload. Unfortunately, you have not shown how they are connected.

 

Here is the full circuit. Voltage in is 24Vdc, Each RCservo can pull about 1.5amps

 

this thing has thermo protection and current limiting at 3.33amps. It actually caught fire

 

It's also a rehashed 78S40 from the 1970s with a drop-out voltage of 1.5V. That means at 3.33A it's dissipating 5W (although that will be reduced proportional to the duty cycle) - I woudn't fancy its chances. I'd suggest that you try something more modern with a MOSFET output stage.

 

It's also a rehashed 78S40 from the 1970s with a drop-out voltage of 1.5V. That means at 3.33A it's dissipating 5W (although that will be reduced proportional to the duty cycle) - I woudn't fancy its chances. I'd suggest that you try something more modern with a MOSFET output stage.

Lol love hearing about the old days ! Thank you for suggestion

 

Of the primeval switched-mode controllers, TL494, TL594, 78S40 and SG3524 are well passed their sell-by dates. SG3525 is still just hanging on because it can drive MOSFETs and has both inputs of its error amplifier accessible.