Post by thread:Question regarding start up latch...

>I have a question regarding the latch up state where Vcc goes below specs >and the MCLR line remains high. I understand this can cause a latch up of >the micro. > >My question is : Will the watchdog then reset the micro as soon as things >are back to normal? My early experiences with the 16c71 showed that allowing the supply voltage to drop to about 0.7Vdc, then raising the supply back to normal would cause the PIC to lock up. The external reset input was NOT able to reset the PIC. If MCLR is asserted before the supply drops to the twilight zone, the PIC does not lock up. Based upon that experience, I now use the Xicor X25043 power up reset / power supply supervisor / watchdog in all my high end designs. As a bonus, the xicor part includes 512 bytes of eeprom. Battery powered designs using the 12c5xx and 12c67x parts don't seem to have a problem - I try to ensure that the supply collapses completely (below 0.3Vdc) when the battery is removed; VCC rise time is nice and fast when the battery is inserted and the PIC powers up reliably. AC powered designs with the 8 pin parts either use a simple reset controller (1 transitor) *OR* use a 2 transistor power switch to switch the logic supply. The 2 transistor power switch is used for designs where external MCLR is not available. Most of my inexpensive AC powered designs use a zener regulated supply. The 1 transistor reset controller uses the E-B junction as part of the regulator (4v3 zener) - the chip is held in reset with a pulldown resistor as long as the zener is not conducting. An extra cap from V+ to MLCR ensures that momentary glitches don't reset the PIC. Total added cost of the reset controller: about 0.15 US. The 2 transistor power switch follows either a zener regulator or a traditional 3 terminal regulator. It is intended for designs that don't have the external MCLR available. The internal reset requires a quick rise time on VCC (easy to do) *AND* I try to ensure that the supply never lives in the twilight zone (harder to do). The power switch keeps power OFF until the unregulated supply is above a certain threshold, then keeps VCC applied until the supply drops below a lower threshold. The hysterisis is picked to be substantially more than the maximum ripple voltage on the unregulated supply. Total added cost of the switch to the power supply: less than 0.20 US. I should point out that I have not charachterised the 8 pin parts in regard to their behaviour while VCC is in that twilight zone - I don't know if they behave the same way as the 16c71. I got bit once and have simply ensured that I won't get bit again by the same problem. Higher end designs can afford the cost of the Xicor part, lower end designs can certainly afford the 15 or 20 cents extra. dwayne Dwayne Reid <.....dwaynerKILLspam@spam@planet.eon.net> Trinity Electronics Systems Ltd Edmonton, AB, CANADA (780) 489-3199 voice (780) 487-6397 fax Celebrating 16 years of Engineering Innovation (1984 - 2000) * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Do NOT send unsolicited commercial email to this email address. This message neither grants consent to receive unsolicited commercial email nor is intended to solicit commercial email.