# Switching batteries from series to parallel - technical solution

I'm thinking of developing a pocket 'power bank' that could provide both 6V and 12V - a charger to charge either a phone (with 6V) or a notebook computer (with 12V).

The electrical diagram is simple enough; switch attachment of the terminals like below.

The schematics is not everything though. I'd like the charger to have two sockets, one providing 6V, the other - 12V, and to "sense" what to serve by having a plug in either of the sockets (let's say using both at a time is forbidden; I can place them in such a way that both plugs won't fit at a time, say, a slider that opens either of them but never both).

Best if the "sensing" was done in the simplest, mechanical way (e.g. there are sockets that provide an extra "sensing" contact that shorts to mass if the plug is in), but if that would prove too difficult or mechanically complex, a more complex electronic solution would work too, providing it won't drain the batteries when nothing is plugged in. (and of course the circuitry must accept external charging current - accepting only one of the two, either 6 or 12V is fine).

And of course while just cutting one of batteries off for 6V would be the easiest solution, it would halve the capacity for 6V application and cause parasitic charging issue when switching to 12V after the 6V power has been depleted, so it's not really an acceptable solution.

So, how to build such a circuitry - that provides power from two batteries, in series or in parallel depending on which socket was used?

• What do you mean, "how to build such a circuitry...?" What's the actual question? Do you want to know how to specify the components? How to source them? How to assemble them? Mar 1 '15 at 20:25
• @EnergyNumbers: The core of the question is mechanics. It's far less about the electrical side (for which an example solution is given at the beginning) and far more about 'making it wieldy' - yes, how to assemble them so that the switch appears 'transparent' from the user's point of view.
– SF.
Mar 1 '15 at 21:19
• @SF I've just been arguing that it's an EE question :-). My answer addresses the switching transparency BUT the electrical aspects are what will and should be used to make it work well. You could force a mechanical solution on the requirement but using electronics it will "just work" as you desired. Auto change to 6V or 12V with no extra switches. Minimal 12V drain when 12v not used. Sensible dealing with 6/12 charging. Mar 1 '15 at 21:23
• You mention using a slider to open one socket or the other, but not both. Is there a reason why you can't make that slider your voltage switch?
– Mark
Mar 1 '15 at 22:22
• ... lack of a knowledge of electronics. | What output currents do you require at 12V and at 6V? | You say 'relays are bulky' - what size of device are you expecting - 'and power hungry' what % of total power drawn would be acceptable. Electronics can manage 0.00% steady state draw for control if that's a necessary part of the spec. | It would be useful to have a requirement definition that allowed the suitability of an answer to be determined without having to submit it. Mar 2 '15 at 9:37

You haven't told us the context this device will be used in, so it's hard to evaluate how appropriate different alternatives are. Here is one possibility:

The battery is always 12 V. For one thing, this simplifies charging. The 12 V outlet simply wires to the battery leads, and all is well.

The 6 V outlet is actually driven by a buck converter from the 12 V battery. Again, we don't know the context of this device, so it's hard to say whether the extra electronics is worth the other simplicities. Buck converters from 12 V to 6 V can be small, cheap, and quite efficient. Since according to your own answer you can use a special cable for each outlet with extra pins, you can use this mechanism to enable the buck converter so that it uses no quiescient power when 6 V isn't demanded.

The two main advantages of this approach is that no interlock is needed (mechanical intracacies are often more expensive to develop than a circuit) and charging doesn't have to deal with somehow sensing and adjusting to how the battery is configured. The battery is always charged as a 12 V battery.

You could take this concept further by using a higher voltage battery, like 18 V for example, and a buck converter for both 12 V and 6 V. This has the side advantages that both outputs will be well regulated. Another important issue you may have overlooked is not allowing the external load to drain the batteries to the point they become damaged. with a circuit between the battery and your load, it can manage the battery and disconnect the load when the battery gets too low. It can also do over-current sensing, or current limiting, or otherwise deal with a badly behaved load without something catching on fire.

Your question only asks for something simple, but there may be more issues to think about than are obvious at first glance.

• it would be 'useful' to know desired max Iout. Also cost, size and shade of Octarine. I doubt that you or I are going to be able to determine the shade. Mar 2 '15 at 9:44
• @Russell: Yes, and the specs (such as they are) keep changing too. His own answer doesn't actually answer the question as asked. Without the OP stepping back and telling us the parts that matter and not wasting our time with imagined solutions, this question will be impossible to answer other than by shooting in the dark. Mar 2 '15 at 13:59

You have addressed many of the main points - it's really just "a matter of engineering".

If you take 6V out from the top of the right hand battery and 12V from the 2nd contact down (12V output of DPDT switch) as shown on the diagram below, then you never get incorrect voltages at the output. The changeover to 12V mode can be triggered by inserting a plug in the 12V socket, as you suggest. If you wanted this to happen only when there was a need for 12V current draw you could connect a high resistance sense electronics powered by +6V that detected an actual load present. With some thought you could fully automate this so a 12V load connected to the 12V socket would auto-trigger 12V mode when a 12V load was detected and terminate once there was no load current.

In most cases a physical relay is still hard to better in terms of value for money and low voltage drop.

For 12V charging you can trigger the 6V/12V changeover based on charger voltage or a dedicated switch or with separate charging jacks.

Battery balance should be reasonable for identical batteries which are equally charged initially. As you note, you must never draw current from only one battery or charge one battery - changing from 6V to 12V with imbalanced batteries is probably OK but changing from 12V to 6V will tend to destroy the switch with imbalance currents. A very small current limiting resistor in each battery lead would be wise - this can probably be just the wiring resistance with connections being made in such a way as to maximise interbattery resistance with the existing wiring. A resistance that limits real world maximum battery to battery imbalance current to say 5 to 10x relay rated current for very occasional unexpected circumstances is probably OK. Resistance in the battery leads is of course undesirable in normal operation and should be minimised where possible.

• Relays are both bulky and usually quite power-hungry (and would probably need to be powered from a single 6V, which would unbalance the load); I'd much rather do without them. Power-sensing circuit sounds like an intuitive solution but if you get down to implementation details it becomes quite convoluted - unless you know a way to make it trivially simple, which would be quite welcome. The socket-switch is an interesting idea but it usually provides just a single switch connected to negative contact, not double and isolated from the rest of the circuitry - I'm not sure how to use it here.
– SF.
Mar 1 '15 at 21:29

I think I came up with an idea which should be trivial to implement and work without any active electronics. Since the design is to employ dedicated charging cables and I can take liberties with the sockets, I can use the plugs as "keys" that route the power as I like: I'd use 4-terminal socket (just a single one) and a "6V cable" and "12V cable" (obviously differing with plugs on the "recipient" end), which would connect to the terminals in a way I need. No "sensing circuitry" and the process of switching voltage is integrated into necessary activity of the user of the device.

I'm still leaving the question open - maybe someone comes up with something even better?

• Your mechanical lockout had better be solid. Plugging in both cables will short both batteries. Also, charging will now be a problem because the charger won't know what voltage to put across the "battery". Mar 1 '15 at 23:25
• @OlinLathrop: plugging in won't be a problem as there's just one socket, I'm doing away with two separate sockets entirely. The charger will use the same cable as 'recipient', again, routing voltage as needed through the plug, so I can use both 6 and 12V, just each with appropriate cable. The only concern is choosing a socket non-standard enough so that nothing else than my 'key plugs' fits (at first I pondered USB socket, then I realized I risk routing 12V through data lines if someone plugs a common data cable in).
– SF.
Mar 1 '15 at 23:54

I made a fast drawing with a bit more proper electronic:

You cannot put two electrical sources in parallel, this will break your system imediatelly or on the long term.

I am using Not gates because they are the most common, also the simplest to build.

Connect the green with the green ($12V$) and orange with orange ($6V$) at the moment the circuit is on $12V$, press the switch and it will work on $6V$ (you need to work a bit the switch so what is not connected is then connected to the ground)

I also indicated that you can charge your batteries in series please