Uninterrupted Supply With Lipo Battery

Hi Allen,

You just need the primary supply (the larger battery) to hold off the secondary supply, until there is absence of the primary supply.

There are ICs that do this (see Linear Tech), look for 'ideal diode' and 'power supply OR'ing circuits' but if your requirements are very well defined (e.g. if you know precisely what the span of voltage will be for the primary and secondary sources) you could probably put something together with MOSFETs, or MOSFETs and diodes if you don't mind the volt drop. A more general purpose solution will use the ICs as mentioned above, but even they will have some restrictions based on the encountered voltages for the primary and secondary supplies.

As you say, things get complicated fast if the voltages change, and there's no single simple solution.

Depending on how much effort you want to put into solving the problem, you could even use comparators and some logic, or even use part of a microcontroller if you've already got one in your design.

How about using a super capacitor for the backup while changing out the large Li-Po battery. Certainly short work if the circuit is low powered anyway.

Clem

Hi Allen,

I am old so I tend to think more mechanical. For simplicity I would use a Make before break DPDT switch that could switch between the Main and Backup battery. That way you would only have to switch to backup while you replace the main battery and then switch back after the replacement is made. Likewise when the switch is on Main the backup battery could be replaced. The Make before break switch would have both batteries in parallel for a brief instant while the switch is passing from one position to the other. If you did not want to use a make before break switch I suspect that a Capacitor could be used to bridge the milli second while a standard DPDT switch transitioned.

John

Hi Allen,

I have typically used Schottky didoes (low forward voltage drop) to steer either the battery voltage or the charging voltage into the voltage regulator.  That way, when charging the battery, the charger voltage will be greater that the battery voltage and therefore power the regulator.  This is an example of one of the techniques that I use for powering the device while charging the battery. The Vbat (sw) is either the battery output (switched) or the output of a boost circuit (switched).  The Vcharge voltage is the charger input.

Good luck,

Gene

Thanks for the response and yes things got complicated quickly.  I originally thought I had a solution working but since I was using a power supply I didn’t see the problem.

Originally I intended to use a pfet where the gate was tied to the primary supply and a Schottky diode to prevent back feeding the gate from the secondary source.  I will post a schematic later when I’m at the computer.  The idea being that when the primary supply was disconnected the gate would go from 1 to 0 (digitally speaking) allowing the pfet to open allowing second supply to conduct and shutting back off when the power was reconnected to the primary source.  This seemed to work until I put batteries in play.  I realized that I can’t guarantee that the primary source is at a higher voltage and therefor the Vsd of the pfet in the second supply will start conducting because Vgs might be lower.  Since the battery voltages slowly decrease this will cause the secondary supply to draw current which is what I am trying to avoid. 
    I have tried a low power comparator with some success to switch supplies using an analog switch but there are still times I can make it glitch. 
   Once I get this switching to work then I need to also put in a circuit to recharge the secondary battery from the primary. 


Thanks for taking the time to look at this.

  Allen

I like the idea, but I’m really trying to make this automatic, with no need to throw a switch.

Allen

The problem is that the circuit at times is not low powered, it periodically draws 300ma in order to power some LEDs, also I am limited in board dimensions. 

I have actually ordered some ORing ics yesterday, waiting for them to come in, I will take a look at what you have suggested above,

Thanks for taking the time to look at this.

     Allen

Hi John,

The mechanical method you describe can work with a relay too. Using the primary supply to hold off the relay, and having a capacitor to bridge that short time period as you say. I did that for a home router, where the primary supply was a DC power brick, and the secondary was AA cells (non-rechargeable) to handle any power blackouts for a few hours. The primary supply energised a relay so that it was normally activated. For the secondary supply, I wasn't concerned with charging the cells, since blackouts here are uncommon, and I could just replace the batteries every couple of years. Depending on how fancy someone wanted to get : ) it could be possible to have a latching indicator, e.g. a very low current LED would be good enough for such a router backup system.

I also sketched the design below at the time, but I didn't assemble this one (so it might not work), since the relay method worked fine for me. The design below is a bit of a hack, since it relies on specific voltages, and involves messing about with particular volt drops, but it's a quick circuit for a specific purpose as mentioned.

I looked at the lineup from analog last night and I’m thinking this is the best solution.  Thanks for the input, much appreciated.

Allen