Trouble with DS1307 battery backup

[aratti]

Very likely a wiring problem. Did you place two diodes to isolate the battery from power?

You can have more help if you post the schematic.

Cheers

Al.

[aerostar]

Very likely a wiring problem. Did you place two diodes to isolate the battery from power?


Al.

You do not need your suggested diodes - if you look at the data sheet the backup battery has its own pin, and the battery should be connected to that and ground.

[Amoque]

I was thinking more like stray capacitance issues, or I2C resisters of wrong value (3K3)- something of that nature; maybe battery back-up voltage (3.32V) issue. Wiring the battery could hardly be easier and I've looked, until my eyes crossed, at the holder to be sure there was nothing wrong with it. I think I'll try a different battery, but I've already tried a different 1307 and a different 12.5 pF crystal, verified wiring- exact same result. Circuit is rock solid, just goes "404" on me when the battery is inserted.

[mackrackit]

What happens if you leave the battery inserted? Can you then reprogram the time?

[Amoque]

No... The LCD initializes, then displays "TEST" on startup or reset (both work fine, even with battery inserted), then clears the LCD and displays the time in a simple loop. Function is consistent regardless of battery, except the display remains fixed at 04:04:04 with the battery in place.

Reprogramming sets the time registers to the startup value, but 04:04:04 continues to be displayed until the battery is removed.

In exercising your suggestion, Dave, it appears that the clock continues to function correctly (and accurately) while the display is erroneous; rather it is some anomaly in the I2C that affects the output or display! What is the likelihood that different (lower? Higher?) I2C pull-ups would resolve this?

EDIT: Appending or propending text before or after the time displays as expected as well.

[Amoque]

Ok, mystery solved.

It appears that noise, capacitance, or competing residual signal on the breadboard was causing issue. Running the voltage supply directly to the circuit board resolves all issues. Thanks all - and you particularly you Dave, for putting me on the right track.

[Amoque]

Ok, after further research, it appears related to battery voltage. The "new" battery I purchased reads over 4 volts. Just a head-up to others: something to look at if you're getting wonky readings from your DS RTC. I clipped a high draw LED to it for a bit - works flawlessly now. Sorry for the "Much ado about nothing."