The 5V switcher failed recently and it was replaced with an available LM317 linear stuff last week. After that the set up has become very unstable - the VFD display and the MCU just freeze on power ON and at times.

-I remember, that some 89C51 micros don't like it when power-on lasts too long! Vcc should reach its final value within a few milliseconds. May be there's too much storage capacitance in front of LM317, so that power-on time is degraded?

-Keep in mind, that slew rate from 4.75V to 4.25V at power-down or brown-out situations must be greater than 300µsec for the DS1232! May be this is violated in certain situations?

-Why did the switcher fail? Could be an indication that there's a hardware bug in your circuit...

-May be oscillator needs too long for turn-on, as result of a poor crystal or unproper burden capacitors? What happens if you choose maximum reset time?

Kai