Solar Battery Plant Question

[ebarker]

We want to connect 6 ES17-12 (the ones that came with the robot) in parallel to create a large capacity power plant.

Do the batteries need to be connect with or'ing diodes or just hook them in parallel ?

This plant will be recharged with a solar PV array.

[Al Skierkiewicz]

Ed,
I have been cautioned by the manufacturer and the Battery Institute that connecting two like batteries in parallel should pose no particular problems. Provided that the batteries are of the same age, type, and life (charge/discharge) history. However, with seven in parallel I would recommend steering diodes or switching of some sort. Depending on the current capability of the array, you might want to switch batteries for charging. i.e. have one on charge while the others are delivering load current. Most RV suppliers carry such diodes for RV dual battery systems. They use Schottky diodes for a lower forward voltage drop in most cases.
It would seem that the problem exists with circulating current between batteries of differing terminal voltage. As one battery sags, the other battery will try to deliver current until the terminal voltages are again equal. Unfortunately, this turns into a domino effect.
If size is a major concern, perhaps single cell batteries in series might better suit your needs. What is your application?

[EricVanWyk]

I'd recommend putting some low value high power resistors in line with each battery when you put them in parallel for the first time. This will slow down the balancing and prevent the initial spike of current. This method is slow but easy. After they have settled, take the resistors out. Be sure to run a few quick numbers to estimate the power your balancing resistors will put out, and how long it will take.

Since our batteries are charged with 2, 4 or 6 amps, I'd multimeter verify that they are all within a volt and then put half ohm or quarter ohm 1 watt resistors in there. Since we have 18 Amp hours, I'd plan on waiting a full day. If you do the math, I'm being too conservative, but why rush?

At this point, you should treat the pack as one unified battery, and avoid taking them out of their parallel configuration for anything. Otherwise, you will need to balance again.

As for diode-oring them:
From your description, I assume you are using the pack as a back up for a solar array. Charge during the day, discharge at night. If this is the case, be very careful about where you put your diodes. Frankly, I wouldn't use them in this case.

If you still want them :
I'm not a huge fan of steering diodes because of their voltage drop. Even a big schottkey will drop a third of a volt. To reduce that, you could use an ideal diode controller like the LTC4352, tps2413 or ISL6144. If your currents were lower, you could use an integrated ideal diode like the ones from picorpower. Since you have 7 of them, you could roll your own like this.

[ebarker]

The application is on our off season project that we are working on again for the next 4 months.

Pictures were posted here last fall.

During the next 2 or 3 weeks we need to update our power scheme with 6 (not seven) batteries, plus a solar charge plant.

We have to get some more power measurements done but rough numbers we may be charging maybe half of our full load power draw. Working from memory we may have a maximum plant draw of 50 amps at 12 volts. We will charge on the run. We may be able to upgrade the power plant to charge at the machine full power draw rate. We will see.

We are not opposed to or'ing diodes if that gives us the safest and most reliable operation. We just want to get it right.

We'd prefer to use 6 robot batteries instead of a single large deep cycle for several reasons but that may not be totally practical.

[MrForbes]

Lead acid batteries seem to be pretty forgiving when you connect them in parallel.

With solar power, you want to make sure you are using as little power as necessary, since power costs a lot to generate. Conservation is the most cost effective approach. How well are your pontoons designed? are they really just rectangular blocks? How fast does it really need to move? Could you reduce speed a little bit and reduce power consumption a lot?

We used solar power for our underwater ROV for the NURC this year....good stuff....

[ebarker]

It is what it is.

For better or worse we are moving forward with the current platform design.

Not the most efficient hulls, not the most efficient motors, etc.

So we will deal the cards that are on the table.

Our hull speeds requirement is low - about walking pace.

If we wound up with 400 watts max power draw and 200 watts off the panel on a sunny day we can live with that.

[MrForbes]

Do you have the solar panels already? If so, then don't worry about it....

Do you have a solar charge controller?

[ebarker]

we do not have any solar equipment yet, so keep talking......looking for recommendations

[MrForbes]

Ok....hmmmm...

Keep in mind that solar panel power ratings are a bit funny, they are rated at some voltage and at some current, but the voltage is at no load, and the current is into a short circuit, so you do NOT get the max rated voltage and current at the same time! A 75w panel might put out only 5 amps when charging at 12v in full sun.

Think carefully about how long you will want to run the device, and how long you can have it charging. Usually it's more cost effective to buy more batteries and leave it charging as long as possible, than to buy more solar panels.

You probably want a charge controller that is rated for a bit more current than your panels can generate.

[ebarker]

I'm thinking we want to implement this in phases.

1st we install 6 batteries, with or'ing circuit.

2nd install a big enough charge controller to handle maybe 400 or 500 watts.

3rd install a single panel on the craft, maybe 200 watts.

4th install another panel on the craft, or 'apu' style on the ground to get us to 400+ watts.

So we have to do some planning.

It is a solar powered lake rake....

[MrForbes]

Have you measured the current draw while at cruising speed? or loaded down (I'm not sure what a "lake rake" is)

[ebarker]

we have to do some instrumentation work and collect some numbers. we had gotten some quick lab numbers with stalled motors and such but it isn't really realistic.

so back to the lake.

a 'lake rake' is a term we invented. sort of like a robotic lawn mower, it skims the lake for 'flotsam' - floating debris.

eventually it should be able to run autonomously we might think......

In Melbourne Australia at Parks Victoria there is a static litter trap shown here

This is a robotic mobile trap.

[MrForbes]

neat!

I suggest putting a motor in a big sink/bathtub and measure current draw at full speed. Motors and propellers do interesting things....you might want to set up a bollard pull tester if you can get up the motivation....

[Al Skierkiewicz]

Ed,
Your power output on PV arrays is limited by size and sunlight. In Georgia during the summer, you should be able to get plenty of sunlight provided you can keep the craft in the sun and not under trees. In order to get an array capable of 500 watts, you may have to increase the size of your craft. A quick check found 125 watt units are 25 x 60 inches but the cost is in the $300-$400 range. Current output would be around max charge current for one battery, 6 amps@16 volts in full sunlight. I am guessing you would mount the array flat to provide less wind load so you have to fudge the current and voltage output a little. Without some form of switching, it is difficult to prevent the array from providing some load current which takes away from the charge current. Even at 6 amps, our batteries will take 2-4 hours to fully charge, each.
As to parallel operation, there are varying opinions on this subject. Many people will tell you it is OK to do it, some will tell you never do it. The Battery Institute of America gave me this advice...If the batteries are of the same type, age, temperature, and charge/discharge cycle life then it can be assumed that the only difference lies in manufacturing tolerances. If any of these factors are different, then the individual battery chemistry is sufficiently different that terminal voltage will vary from battery to battery. Millivolt differences will cause current flow from the higher voltage to the lower voltage battery until the difference is equalized. However, the charge state, internal temperature and internal resistance of the higher voltage battery will have changed to the point that it's terminal voltage has now changed. The result is that both batteries will self discharge over some period of time as one feeds current to the other. It is expected that used batteries with our type of load history will have greater than a few millivolts terminal voltage differences due to changes internal to the battery, and the experts expect self discharge to occur in perhaps a day with just two batteries. (Diodes prevent this circulating current.) AGM batteries are not very tolerant of deep discharge and may be damaged by this kind of use. Different battery types (chemistry and construction) differ in terminal voltage variance so blanket statements cannot be applied to all battery types used in parallel. One application that used batteries in series/parallel operation was electric submarines. Remember that these were large, wet cell batteries where the specific gravity of the cell electrolyte was checked several times a day.

[MrForbes]

I think the application may make a difference as to whether or not you'd want to run the batteries in parallel or not. I think it's likely that this particular application is something that would not be left alone for long...it will be turned off and have the batteries all disconnected at the end of each work day. Also, it's working time will be limited.