From DIDEAS Wiki
- Datalogger supply voltage is restricted to the 2.7 to 3.3V range. Outside of this range communication with both the DS1339 and SD/MMC card become problematic.
- Keep fresh batteries in the unit - however there is no advance way of know that a low voltage will occur shortly soon - and thus the batteries usage needs to be further restricted.
- Break the jumper trace connecting BAT+ to VCC. Install a Schottky diode between BAT+ and VCC. Install a 2nd Schottky diode between VCC.SD and VCC. The result is that the maximum of BAT+ SD.VCC are used to power the CPU and DS1339. The voltage differential between the SD card and CPU will be small to permit communication. The CPU will operate from the BAT+ voltage under most conditions, however when DS1339 communication is required (only at reset), the voltage booster will be started. The voltage booster is also normally powered during SD communication.
- DL6F PCB crystal pads are covered in solder mask part of this should be exposed. See note below.
- remove portion of solder mask exposed with flat screwdriver
- Not possible to test battery voltage as the voltage booster needs to be enabled to enable testing, and thus the boosted voltage is tested
- No low voltage reset
- Watchdog reset software protection can fail if the supply voltage is too low to maintain WDR oscillator - but not low enough to cause reset. Could happen when batteries are near depletion or removed.
- (2-22-06) Not a datalogger problem - but an system integration issue : Datalogger connects to power "downstream" of a 10 ohm resistor in the host. The logger operates fine, however upon powering the SD card, current surges drop the system power below the 2V cut off causing the host to reset.
- (2-23-06) Time of day clocks have excessive drift. The cause is probably not the result of temperature changes and is most likly the result excessive (stray) capacitance on the time-of-day crystal.
- (2-23-06) Crystal packages should NOT be soldered to circuit board.