How to daisy chain dissimilar backup batteries together?

This is a request for suggestions / comments - I've a few specific questions but really I'd love to know what y'all think of my thinking. If you get me.
I've been running our domestic lighting and music from a couple of truck batteries in the attic, with 400 watts of solar panels on the back roof feeding through a solar charger to keep them charged.

I desperately need more storage capacity, but I know that you can only connect identical batteries together like this:

You can't connect up dissimilar batteries (whether capacity or age differs) because they'll tend to discharge into each other rather than sitting in happy charged equilibrium. So this is a no-no:

This is a problem, because I don't want to chuck out my existing batteries every time I want to add capacity. I want a solution that'll let me hook up as many old used car batteries I can get hold of.

One way might be diodes:

This'll stop the batteries discharging into each other, but am I right in thinking that this would mean that the weakest / smallest batteries would never get discharged to a lower voltage than the big ones? So, in other words, they'll all flatten evenly (at least as far as voltage goes), so I'd never end up with one battery at 10.5 volts - dangerously flat - until the others have discharged that much as well?

It's a shame to lose a bit of energy warming the diodes both on the way in and the way out, but it's a relatively small amount.

It got me wondering whether it'd be better to have the batteries charge and discharge separately. Instead of using diodes to stop inter-battery shenanigans, I could use relays instead:

Note that there's no need for one on the final battery. With this setup, you can see that the first battery is feeding my house, while the third one is being charged at the moment. No matter what state the relays are in, there's no way for the batteries to end up connected to each other, which seems nice:

But there's the issue of how to control the relays automatically. It occurred to me that I could use a single DPDT relay for each battery, controlled by the battery voltage so that when each one was fully charged it would automatically switch from "charging" to "feeding".

This looks like it could work, though I'd probably need a slightly more sophisticated circuit controlling the relay coils, so that there was a nicely presettable "switch onto CHARGE when less than 12.0 volts, switch off when greater than 13.5 volts" thing going on. All getting a bit complicated now, mind.

Well - that's where I've got so far. Any thoughts? Diodes seem like the simplest way to set up an expandable battery bank, but I'm not sure I've got my head around the way stronger and weaker batteries will interact. Obviously, a key aim of all this is to share the load, so that I don't run any one of the batteries down too low (= battery sulphation death).

Don't want to reinvent the wheel, if there's an established way of doing this…

h

I think you have covered all bases.
One thing though, if you use Shockley diodes you loose less forward volts drop with them.

Hi Grumpy_Mike,

So am I right in thinking [Schottky] diodes will do what I need, no need for relays and control circuits?

Basically I just want to protect against any single battery from being discharged too low while others still have charge to give.

h

So am I right in thinking [Schottky] diodes will do what I need,

Yes they will stop the batteries from charging each other and keep the discharge voltages all the same.

It occurs to me that adding all the diodes will probably cause an issue with the solar charger. When it's first connected, it looks at the battery voltage to decide whether to use 12 volt or 24 volt mode. At night it takes its standby power from the batteries - which it won't be able to do any more. Also, it won't be able to monitor the battery voltage.

Hmm. That's a pain.

So, as a hacky solution, what would be the effect of just not using diodes on the first battery, like this:

Note that the first battery (A) no longer supplies the load directly either, or it would end up doing all the work.

Tell me if I've got any of this wrong (I'm assuming a 0.5 v drop across the diodes):

  • When under charge, all batteries will get charged up, but batteries B~n will always be 0.5 volts less than battery A (so they'll never quite charge fully, hey ho, but at least the charger will cut off too soon rather than too late)
  • When no longer charging, there may be a little bit of leakage from battery A out to the rest of the batteries as they self-discharge, but battery A will never drop to less than half a volt above the voltage of the rest of the bank
  • When discharging - and this is where I'm a bit hazy - batteries B~n will provide almost all the current, but some will come from battery A through both sets of diodes
  • When no longer discharging, battery A will discharge into the other batteries only up to the point where it's voltage is half a volt over them.

Side effects:

  • The charger will go into 'float' mode half a volt too early for most of the bank - which I may be able to get around by telling the charger it's connected to a Gel battery instead of SLAs (Gels, apparently have a float voltage 0.5v higher than wet SLAs)
  • The load will see half a volt less than it's used to

It seems to me that this could work. Does it seem like a bad idea? Am I right in thinking Battery A could actually be a tiny battery - it's acting more like a sorta voltage follower than as a store of power?

You seem to have a good grip on this.
I am not sure how your generator "sees" the battery voltage but one way of making it see better is to include a highish resistor value across one or more diodes. Then any cross charging is an absolute minimum but the generator should still see the voltage as it is not taking any significant current from the batteries.

With batteries connected in parallel, even if they are identical, you cannot assume they will all charge equally,
ie the charging current will equally split between the multiple batteries.
Most standalone Solar Installs have the batteries in series, which guarantees equal charging currents
through all the batteries.
Either way, there is no foolproof method of ensuring that they are all being charged equally.
With lead acid batteries, the only foolproof method of determining the battery charge is to measure
the SG of the battery acid, which cant be easily done electronically.

Hello Friends,

A generator that runs off of natural gas or propane with an automatic transfer switch is what you need for piece of mind. Prices vary depending on what you want to run in a power outage situation. Do a google search for Guardian generators, check out the guardian web site. I checked locally here in Northern Ohio, and I can have the 10k not gas with transfer switch installed. I recommend you to purchase the battery with the optional AC charger. The Universal AC Charger has adjustable output voltage, you can set its output voltage same as or close to your device input voltage. In this case, you can leave the battery connected to your device when the battery is connected to the AC power adapter.

Best Regards,
Anthony Martello :slight_smile:

Thanks for the advice, everyone -

@Grumpy_Mike - I've ordered a load of Schottky diodes to play with. Managed to find some on Farnell that have a forward drop of around 0.25 volts which sounds nice and efficient. The resistor-across-the-diode(s) trick sounds good - it'd never have occurred to me. Seems so logical, thanks!

@Mauried - points noted. Even if the batteries started off with the same open-circuit voltage each, when you take a reasonable load from them their voltage drops / sags, and then once the load is removed they take a little while for their voltage to bounce back. I'm guessing that batteries in different conditions (age / capacity) will sag and bounce back at different rates, meaning they'll recharge at different rates too.

The main thing is that even if they don't charge equally, I should still get an increase in usable capacity even if it isn't as much as you'd get with a more sophisticated per-battery monitoring and charging system. With SLA batteries, it seems the key thing is to spread your load demands over as many batteries as possible, so if adding another old one - even if it hasn't much capacity - reduces the amount the others will discharge by even a small amount, it'll be helping.

I know that I'd get a more efficient system by running things at a higher voltage, but the convenience of having the whole setup (panels, batteries, lights etc) running at 12 volts also translates into savings - no need to keep converting 24 or 48 volts down into something usable.

@AnthonyMartello - a generator would be a great backup to have but I'm trying to build a system that doesn't rely on fuel supplies. I figure that if we do have an energy crisis the solar panels will keep us lit and entertained and we'll hook an alternator up to an old push-bike as a way to supplement it if needed. I like the idea of a bike-powered TV :slight_smile:

My aim isn't to have a perfect system, just a cheap and safe one, and one I can replicate and explain easily for friends and neighbours. If I can make use of old spare car batteries just by adding a couple of diodes to them and connecting them up to the bank then it'll cost me next to nothing and it'll save the batteries from being thrown out. And so far, I've spent less than £800 on solar panels, charger and fuses, and the system's given us free lighting, music systems in most rooms, iphone and iPad chargers everywhere. It's a bit hacky but it's worked perfectly for over a year now. Bring on the econopocalypse ...

I'll report back once I've done some experiments. Everything hinges on whether we can fool the charger into thinking everything is working normally.

Again, many thanks folks, your time and expertise is hugely appreciated.