Frangibolt Driver Circuit

[arul]

Hi,

I am looking for frangibolt driver circuit which can be enabled from MSP430FR5969 microcontroller. Now am constructed driver circuit using n-channel mosfet. Before that as well as transistor can be used to drive the mosfet.

Frangibolt Datasheet ->  ftp://apollo.ssl.berkeley.edu/pub/MAVEN/MAVEN%20Parts%20Selection/9_Actuators/TiNi%20Frangibolt%20FC2_16_31SR2.pdf
Microcontroller Datasheet -> http://www.ti.com/lit/ds/symlink/msp430fr5969.pdf
Microcontroller operating voltage: 1.8V to 3.6V, but i designed under operating voltage of 3.3V
The diver circuit input voltage is 28V, because the frangibolt operated at 28V so i choosen.

I have attached what i can done of circuit.

Please suggest to improve my design...

Advances in thanks...

[rstofer]

I don't know how the circuit will respond to startup transients.
I think the input diode is backwards.
Consider what happens when the microcontroller starts up.  Usually, the output pin is floating and that might be fine but I would add a pull-down resistor the the base of the first transistor.  I want to KNOW that it is off, not just sitting there with a dangling base.
When the code starts, it is important to set the pin output to the proper output state before actually configuring the pin as 'output'.
Overall, I wouldn't use the circuit.  I am not at all convinced that it couldn't be induced to misfire.  At a minimum, I might add a physical relay to the 28V supply and control it from a second pin of opposite polarity.  One output would be a '1' to fire and the other would be a '0' to fire.  I might even go further...  I would almost certainly have some kind of toggle switch in the 28V supply.

[ajb]

D57 will prevent your MCU from activating the circuit.  Otherwise your circuit looks like it will work, but it will  work just as well if you get rid of everything except Q50, R421, and D56 and just drive the gate of Q50 from the MCU directly (you still want the external pulldown that R421 provides).   

rstofer's suggestion to add a separate enable signal is a good one.  I would suggest adding a time delay (RC filter) component to the enable signal, so that a momentary glitch can't result in an accidental activation of the frangibolt.  That way you have to supply the enable signal some time before you intend to activate the bolt to 'arm' it, before the 'fire' signal will be effective.  You could also interlock the two signals so that the arm signal is only effective if the fire signal is inactive.