My RTCC can't keep its time when VDD shutdown

[BeeElectronic]

[temtronic]

Since 99% of my projects have an LCD module, the 1st line of code is a delay_ms(500);. This is to ensure the LCD module has time to 'powerup' properly and while the LCD_init() may do it, I've got into the habit of manually coding the delay.
BTW, just because a PSU is 'certified' it may not work 100% with a PIC ( or other micro...). I'd like to see the certification document to read the rise time, fall time, ripple amount and power limits.
I see you're using an LM317T, do you have a 1N4001 between In and Out terminals ? Bypass caps on all 3 terminals ? What value filter caps on In and Out ?

Jay

[BeeElectronic]

I think I found something:

I have 2 ways to power supply my board:
- 1 by the normal way: 2 faston +12V/GND (come from laboratory PSU)
- 2 by my debugging board which include FT232 / 12VDC PSU from a wall mount AC-DC power adapter / ICD-U80 / 10 pin ribbon to my board with +12V/GND/PGC/PGD/MCLR/RX/TX pin

If I use the first way: I disconnect the +12V or GND faston pin on my board to power off. With this method the RTCC don't lost the time.
-> I did about 40 restarts and not one of them reset the time.

If I use the second way: I disconnect the +12V jack to my debugging board (without +12V or GND form laboratory PSU, the debugging board supply my board).
-> I did about 40 restarts and not one of them reset the time.

If I use the second way: I disconnect the ribbon between my board and my debugging board (without +12V or GND from laboratory PSU, the debugging board supply my board).
-> I did about 20 restarts and 5 reset the time. This is the method I use since the beginning because its better access for me.

So if I remove the +12V PSU from the faston pin or from my debugging board everything works fine. But If I remove the +12V PSU by disconnecting the ribbon cable between my debugging
board and my board I have this RTCC trouble...

[gaugeguy]

When you disconnect a ribbon cable, not all pins disconnect at the exact same time. You may momentarily end up with data lines back feeding into the processor with power removed.
You may briefly end up with +12V and data lines but no ground or ground and data lines but no +12V.
This can cause unexpected operation as a data pin ends up either above the processor power pin or below the processor ground pin.

[temtronic]

What gaugeguy says is very true ! To power down ANY computer, you need a controlled, known method... just 'yanking out' a ribbon cable is NOT a good idea !!

Jay

[BeeElectronic]

This ribbon cable is not only for debugging but also for industrial programming and testing for us and the customers. The connector is from Wurth and
it is quite similar to the TAG CONNECT in its use. In this case the RTCC is very sensitive to that and I don't want reprogram the time of one board out of
three by hand. So I keep my DS1307 with this PIC16F19176.

Thanks everybody for your help.

[Ttelmah]

Key thing is that connectors that are designed for hot connection
are actually designed so the pins are made in specific sequences.
So if you look at the USB connector, the ground pin makes before
the other pins. The Tag connect connector does this. Does yours?.