Battery based tank circuit (tested)
Someone might have already done this before but anyway I havent seen it on this board so I thought Ill post it :)
Alot of poeple here seems to have problem fully charging the tank battery. The standard tank circuit uses a diode typically the MBR6045. Because of the voltage drop of the diode, the charging voltage is always 0.6-0.7V lower than it should be...13.8V. With this standard setup the backup battery never get fully charge and may not function properly during crank.
On the diagram I have added a limiting resistor and an extra diode. The limiting resistor is directly connected to the main battery, this ensure that the backup battery get the full voltage from the main battery.
The typical charging current for a sealed lead acid should be 1/10th of the Ah rating. It is not neccessary to have this limiting resistor in a normal setup since the battery will find its own charging current. Having read the datasheet from Yuasa it says that the charging current can be limited to 1/4 of the Ah rating. Using a 0.8Ah battery a 68 Ohms 3 Watt is found to be suitable.
This resistor is very important in this circuit :
-To limit the charging current to 1/4 of the Ah rating.
-To unsure that the back battery get the full voltage of the main battery.
-To limit the current being drawn from the back up battery during cranking.
The additional diode is there so that the PC PSU cannot drain the charging current going into the backup battery therefore maintaining the full charging voltage.
The diode on the top ensure that the backup battery never get drained during cranking...but only through the limiting resistor. Since the current is limited a heavy drain cannot occur on the backup battery during cranking.
I have tested the circuit and it does work as expected :)
Compaq Evo N400C
Targus DC-DC PSU
Engine off = 12.1V
Engine cranking 10.1V
Engine running 13.5V
PSU cut off = 11.0V
PSU/Laptop curent drain = 2.7A
The circuit was tested with a 1N5822, this diode have a drop of 0.525 at 3A. It is better to use 1N5820 as labelled on the diagram and this only drop by 0.475V. So an even better results can be expected with the 1N5820.
With the test, hardly any heat is generated on any components. The only time alot of heat can be expected is when the backup battery is fully drained but then the 3W resistor should cope with it.
The backup battery was delibrately drained down such as the PC die completely. This cut out happened when the backup is supplying 11.2V. The backup battery is then charged for 3 mins by running the engine. A recrank is made, and yup again it survive a crank after only being recharged up for 3 mins :)
I havent tested how long the backup battery will last if it was charged up for 1 hour of driving as I didnt have the time. Im expecting like 3-4 mins if the battery was fully charged up. But 15 seconds of crank survival should be more than enough.
The maximum current can be drain with this circuit is only 3A as limited by the diode. I dont have a more powerful diode to test it out :( A bigger backup battery can also be used together with a recalculated limiting resistor value and suitable wattage.
For some reason the standard tank circuit uses MBR6045, its better to use MBR3045 or something with a lower voltage drop. MBR6045 drop by 0.7V while the MBR3045 drop by 0.5V
Even with this improved setup I dont think its good enough for ITPS. The only way around the ITPS is to add a relay that short out the diode so no voltage is ever lost. Even with that zero voltage drop circuit, the ITPS minimum operating voltage of 13.3V is still pushing it to the limit.
Ill use a higher rating diode next time on a bigger PC...If anyone wanna try first please do so and report back :)