Why H-bridge should be set to off-state (high-impedance) in a fault case?

I read a thesis on H-bridge design which said : All four power transistors are switched into off-state because of a fault (short circuit, under-voltage, over-temperature).
Similiar opinion can be found in another book: Smart Power ICs - Technologies and Applications (By B. Murari, F.Bertoti, G.A. Vignola)

Why all the four transistors should be switched into off-state once a fault case detected? Why not swtich off the high-side power transitors and switch on the low-side power transistors for current recirculation when a fault detected?

Thanks in advance.

zhongxg said:

I read a thesis on H-bridge design which said : All four power transistors are switched into off-state because of a fault (short circuit, under-voltage, over-temperature).
Similiar opinion can be found in another book: Smart Power ICs - Technologies and Applications (By B. Murari, F.Bertoti, G.A. Vignola)

Why all the four transistors should be switched into off-state once a fault case detected? Why not swtich off the high-side power transitors and switch on the low-side power transistors for current recirculation when a fault detected?

Thanks in advance.

Click to expand...

What if the fault is due to a shorted high-side transistor? Turning on the low-side transistor would cause very bad things to happen.

zhongxg said:

Why not swtich off the high-side power transitors and switch on the low-side power transistors for current recirculation when a fault detected?

Thanks in advance.

Click to expand...

In addition to what Barry mentioned; if your load is a DC motor, doing so will apply a hard brake to the motor.
If the load has significant amounts of mechanical inertia, such hard braking can be very damaging.

Guys, Thanks a lot!

Can we just switch on the low-side transistors several miliseconds for free-wheeling, once the current is zero then shut all the four transistors off? The H-bridge we used is for driving a small cooling fan, the maximum working current is about 500mA and the working voltage is 12V.

If you are trying to address fault conditions, you might be
best off to enumerate and prioritize (like DFMEA). You will
not likely be able to deal with every scenario that can be
imagined, but you must deal with those that are flowed
down through your customer or any certifying body.

For example, short-to-ground is a plausible outcome; short-
to-power, less so; an armature short or open is plausible,
a drag induced or loss-of-lubrication overcurrent is too.
Output overvoltage, reverse current, those are unlikely.

Anyway, put some order to it and you can focus on the
signature and appropriate response for each, and check
off the "musts" at least.

Barry, Schmitt Trigger and Dick, thanks a lot for the help.
As Dick suggested, I'll try to list out the "musts" according to my current design.