3
\$\begingroup\$

I bought a PCB module including a DS1307 real-time clock to evaluate if that RTC is right for the system I'm designing.

Here's part of the schematic:

The DS1307 works with a battery voltage between 2 V and 3.5 V. I use the module with a 3 V coin cell battery.

Resistors R37 and R38 make a voltage divider that turn the 3 V of the battery in about 2.5 V, and feed such voltage to the VBAT pin of the DS1307 real-time clock.

What's the use of such voltage divider? Why not connect the 3 V battery directly to the VBAT pin?

EDIT: I added the 5 V flag to the voltage live on the top of the picture.

\$\endgroup\$
7
  • \$\begingroup\$ I'm not sure. Why did you put the divider there? \$\endgroup\$
    – pipe
    Apr 27, 2016 at 10:02
  • \$\begingroup\$ Is the schematics correct, I think it isn't. \$\endgroup\$ Apr 27, 2016 at 10:04
  • \$\begingroup\$ @pipe: That is the schematic of a purchased module using theDS1307. It does something strange that isn't mentioned in the datasheet (at least, not on a quick scan through.) \$\endgroup\$
    – JRE
    Apr 27, 2016 at 10:07
  • \$\begingroup\$ @Marko Buršič: I checked, it is the schematic that the seller of the modules releases to clients. \$\endgroup\$ Apr 27, 2016 at 10:09
  • \$\begingroup\$ Could be helpful if you mention the name of the module, or publish the complete schematic. If, for example, the voltage bus is 5 volt, the divider is needed so the input does not exceed 3.5 volt while on normal power. \$\endgroup\$
    – pipe
    Apr 27, 2016 at 10:13

3 Answers 3

9
\$\begingroup\$

If the system is powered from a 5 volt supply, the divider is necessary to limit the voltage at VBAT. The voltage after the diode D4 will be somewhere between 4 and 4.5 volts depending on the current and diode, which is above the 3.5 volt maximum.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks dude. I originally forgot to mention the 5 V line, but you got it right anyway. \$\endgroup\$ Apr 27, 2016 at 10:29
14
\$\begingroup\$

That module design with the voltage divider is TERRIBLE.

  1. It puts a permanent load on the battery when the 5V is off shortening the battery life significantly.
  2. In addition it reduces the usable voltage range of the battery again shortening its life span.
  3. And lastly the circuit forces a charge current into the battery when the 5V is on. Almost all coin cells are NEVER supposed to be charged.

Here is how the circuit should be changed:

enter image description here

\$\endgroup\$
6
  • \$\begingroup\$ I see your point, thank you for the brilliant design hint. \$\endgroup\$ Apr 27, 2016 at 13:30
  • 2
    \$\begingroup\$ @ElectricalArchitect - Well it is not quite that brilliant. There is really no need for the voltage divider and the diodes when you use a DS1307 or other similar RTC chips. All you really need is to connect up the battery to the VBAT pin of the RTC chip and to GND and be done with it. The RTC chip takes care of internally switching the battery in and out when the chip VCC goes on and off. \$\endgroup\$ Apr 27, 2016 at 13:37
  • 1
    \$\begingroup\$ @ElectricalArchitect - There are some RTC chips that do not have a separate VBAT pin and then a circuit similar to above would be required to drive the chip's VCC pin. Another application where a charging off the 5V may be applicable is when the battery is not used but instead replaced with a Super Cap. \$\endgroup\$ Apr 27, 2016 at 13:39
  • 2
    \$\begingroup\$ Great answer. Too bad it doesn't answer the question at all. :) \$\endgroup\$
    – pipe
    Apr 27, 2016 at 15:14
  • 2
    \$\begingroup\$ "Diodes in series between the battery and the VBAT pin may prevent proper operation. " quoting the DS1307 datasheet \$\endgroup\$ Apr 28, 2016 at 1:13
0
\$\begingroup\$

This board design is indeed full of silliness -- in fact, the DS1307 RTC chip has a UL recognition to not reverse-charge lithium coin cells that is rendered void by this circuit!

For a standard, non-rechargeable coin cell, simply remove R38, R39, and D4, and replace R37 with a 0-ohm jumper.

As to why the circuit was added? I suspect the board was originally designed to work with supercapacitors or rechargeable coin cells, and the DS1307 lacks on-chip support for charging these. Still, it is a poor design -- a better solution would have been to use a fixed 2V shunt ref such as an LM4050-N-2.048 as if it were a low-current Zener (Zeners and forward-biased diodes are both notoriously sloppy at low currents) with a Schottky in series with it to prevent the battery from trying to power up the rest of the board, as well as a 1K resistor for supercap applications to limit inrush currents.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.