I've been thinking about the connections/layout for those who are going to be using 'server rack' batteries such as Pylontech or Seplos Mason, which have their own BMS inside each enclosure.

If you connect such batteries to a single inverter then I agree with the standard layout.

However, if you want to have multiple inverters in order to increase the output power and/or connect more solar panels, then I can foresee a problem. A fault on one of the inverters (a near short-circuit) will blow all fuses within the battery enclosures.

I'm placing the following diagram here in order to assist in discussion of this issue.

My suggestion/recommendation is that it should be the cabling and the inverters which are each protected by separate fuses, rather than the battery units.

@transparent ordered a small adjustable desktop power supply - 30v and 10a - for my top balancing. Should be here tomorrow.

That's good @batalto

The first step is to first decide on what voltage you are going to use as the upper limit. This will typically be just above 3.6v but do read some of the plentiful comments on internet blogs and YT videos.

You set the power supply to that voltage before you connect it across the 16 cells in parallel. Never touch that adjustment knob again until the balancing is completed. You must never try to adjust the voltage with the cells connected.

You need to use copper wire with as large a cross-sectional area as you can find. Some old 2.5mm cable from a household power circuit will do fine. It's theoretically rated at around 27A, depending on how an electrician fixes it.

You won't have enough bus-bars to connect across all the cells, so you'll need to use this wire and some M6 nuts to make do for the remaining connection posts.Maximum torque for Eve cells is only 8Nm. Take care.

You'll get a considerable voltage drop across the cells. For that reason I always connect the power-supply in the centre of the run... feeding 8-cells on each side.

Set the current of the PSU to the maximum 10A and away you go.

Resist the temptation to nudge the voltage knob higher however low the volts appear to be on the end cells of the line. After (many) days the current will start to drop away and the entire set of cells really will reach the voltage you set in the first place!

To be safe, I usually cover the set of cells with a piece of cardboard during this days-long operation. If you accidentally drop a metal tool or piece of cutlery across the terminals, it will quickly weld itself in place, using around 40kA of juice!

@transparent just waiting on my power supply to arrive. Here is my setup. It's a touch "ghetto"

@transparent I hooked up the power supply this evening, but I think I've done something wrong. I set it to 3.6v and 10amp, but when I connected it the voltage said 0 and the amps started at 20a and slowly dropped to 18, but still with the voltage at 0. I've disconnect, but not sure if it was working correctly or not. Thoughts?

Low voltage and high current is typically a short-circuit.

Have you got one or more of the cells reversed? The positive terminals should all be connected to the + output of the power supply.

If the connections look good, then try it one cell at a time for 10s or so.

Put some masking tape on the top of each cell with a letter to tell them apart. Leave space to write in pencil what the voltage & current readings are.

Apparently the PSU has 2 modes. It's in constant current at the moment. But it shows no voltage? I checked and the cells obviously have a charge and volts

You require Constant Voltage mode. You're bringing all cells up to the same 3.6v (or whatever you choose).

Use a multimeter to see what's actually being delivered to the cells.

Those red & black flexi leads look awfully thin to me. I'd suggest you use the screw-terminal action of the PSU output terminals to secure a thick copper wire. Banana-plugs will present a significant resistance for a top-balancing operation.

On your earlier photos you had the flat-plate copper busbars connecting to the positive terminals of the cells, which were closest to the front edge of the work-bench.

On the recent photo, the black lead seems to be trailed along that front edge. I'm worried that you've got the wires the wrong way around. Please check carefully!

20a could be the cells discharging themselves through your psu

@transparent you are right. I'm an idiot. 3.606v going in on the multimeter

No... you're not an idiot. You did exactly the right thing by posting here.

We need to keep the above dialogue intact here so that future DIY installers can find it. We all learn from each other.

I hope your PSU has survived. There may be an internal fuse which has blown. The best time to make this sort of mistake is now... whilst you have a small power supply which limits the damage.

Imagine the joules that would wreak havoc on an inverter if you connected a battery to it the wrong way around! 😲 

@transparent the fuse is fine and everything is ok. It wasn't connected very long. And it is now charging, or at least it seems to be as power is coming out at 3.6v. I'll have to check the fuse