Idiot's guide to a self build battery

[Oldgreybeard]

I've never had any issues just paralleling new cells together to get them all to the same terminal voltage initially, TBH. Although the initial balancing current can look high the power is tiny, as any potential difference from cell to cell is just a few tens of millivolts. Likewise any heating is very low, because of the low internal resistance of the cells and because power is given by I²R.

As an extreme example, say you had two of those BASEN 302Ah LiFePO4 cells with an internal resistance of 0.18mΩ (from their spec sheet) and a voltage difference of 100mV and you connected them directly in parallel. Ignoring the interconnection resistance, the initial current would be given by the sum of the internal resistances (0.36mΩ) and the voltage difference, using Ohms Law, so I = 0.1V / 0.00036Ω = 277A, which looks a bit scary. However, the power dissipated in each cell is given by I²R, so is 277A² * 0.00018Ω = 13.8W per cell. That's not enough to get the things even lukewarm.

Also worth noting that the maximum charge rate for these cells is 2C for 60s, 1C continuous, so they are never going to be the slightest bit bothered by an initial surge of only 0.92C.

[Joeboy]

I've never had any issues just paralleling new cells together to get them all to the same terminal voltage initially, TBH. Although the initial balancing current can look high the power is tiny, as any potential difference from cell to cell is just a few tens of millivolts. Likewise any heating is very low, because of the low internal resistance of the cells and because power is given by I²R.

As an extreme example, say you had two of those BASEN 302Ah LiFePO4 cells with an internal resistance of 0.18mΩ (from their spec sheet) and a voltage difference of 100mV and you connected them directly in parallel. Ignoring the interconnection resistance, the initial current would be given by the sum of the internal resistances (0.36mΩ) and the voltage difference, using Ohms Law, so I = 0.1V / 0.00036Ω = 277A, which looks a bit scary. However, the power dissipated in each cell is given by I²R, so is 277A² * 0.00018Ω = 13.8W per cell. That's not enough to get the things even lukewarm.

Also worth noting that the maximum charge rate for these cells is 2C for 60s, 1C continuous, so they are never going to be the slightest bit bothered by an initial surge of only 0.92C.

I am almost understanding this. My inverter has a 100A peak discharge and that is often throttled back to 75A due i guess to the BMS data sent from the Pylontechs. At 100A we would be 0.35C on the LF280 pack alone. Add in the 6 Pylontechs and its barely a tickle and the whole thing should run even cooler.
280+150= 430Ah
100/430 = 0.232C, should make quite a difference to longevity I hope.

[nowty]

A few observations from the last few posts about balancing and fuses.

Balancing, what do you mean by balancing ?,

(A) All the cells have the same voltage all the time ?

(B) All the cells have the same voltage at the top end ?

(C) All the cells have the same voltage at the bottom end ?

(D) All the cells have the same voltage in the middle range ?

If you run without a cell balancer you have to decide whether your going to balance the cells at (B), (C) or (D), you cannot have (A). Some older cell balancers ONLY balance at the top end, so if your cells rarely charge fully in winter they can became imbalanced. I am not going to explain the technicalities of cell balancing theory because it will take several sides of A4.

You can ONLY achieve (A) by using an active cell balancer, like the one I use, so lets assume your using one of those so it gets a lot easier.

First, don't go charging LFP cells to 3.65V, that's the absolute maximum and your going to stress the cells for no benefit. LFP cells are up to high 90's % at 3.4V, probably 98% or 99% at 3.5V.

There is no point trying to get the cells bottom or top balanced, it serves no purpose because your using a balancer. Just get the cells similar to each other, then preferably connect them all in parallel with no load, only needs a few minutes, thereafter reconnect in series.

If your going to connect in parallel to another existing battery string, get the whole battery string as close as possible, preferably less than 1V difference before connecting. Remember, Pylontech batteries don't have an issue when connecting other unbalanced ones together because they have charge and discharge control. If your making up your own string and only using a simple cell balancer you don't have that protection.

Fuses, what are fuses actually for ?

(A) To protect the batteries ?

(B) To protect the wiring ?

(C) To protect the Inverter ?

(D) To prevent fires and explosions ?

If you take a Pylontech its (B) as the Pylontech's will have its own internal fuse and also its own protections.

Now, what about a DIY battery ?. Its NOT (D) as that depends on your charger settings, whack the voltage up to max and the fuses aren't going to protect them going bang.

Its not (C) as the inverter should have its own protections.

Fuses are primarily for (B), but in this case we don't want to exceed any limitations to the batteries either so its both (A) and (B).

For my system I have 2 battery banks of Growatt batteries and each Growatt battery bank is made up of 5 battery strings in parallel. Each battery string has its own fuse, then that battery bank also has another larger fuse to an intermediate DC bus bar.

For my Pylontech battery bank I also use a fuse, remember if you already have Pylontech batteries you might not have bothered fusing it to an intermediate bus bar but before you did not have any other parallel battery banks to worry about. But if adding another battery bank which might have thicker cables and higher fuse ratings like mine it becomes essential. A stack of 8 Pylontech's is still going to have more than enough current to melt its own intermediate cables if you get a short circuit somewhere in the new battery stack.

As to what fuse ratings to use, well there are no 100% correct answers, there will be a range of values which will be safer or less safe. But choosing too low a value makes them operate at higher temperatures. There also needs to be a hierarchy of values so a string fuse blows first, a battery bank fuse blows second and the final fuse to your battery inverter blows last.

[Joeboy]

Huge amount of info, cheers Nowty! I want to.protect the Pylontechs from the LF280'S if they develop a fault. I'll have a look at what I have for fuses in the current set up when we get back home and go from there. 42 degs here this afternoon, brain is shutting down.

[Oldgreybeard]

From a purely electrical installation perspective fuses or circuit breakers are only to protect the wiring. They do nothing other than prevent wires from fusing and possibly causing a fire in the event of something malfunctioning. They aren't likely to operate fast enough to protect any electronic equipment, as they need a heck of an overload for a relatively long time to blow.

That does present an unusual case for the mains side wiring for a battery inverter/charger, as the mains cable needs over-current protection at both the inverter end and the supply end, unless the cable length is under 3m, when it's deemed OK to use back protection. I guess the inverter may well have internal over-current protection on its output but I have wondered if that meets any regulatory requirements.

[marshman]

__

[nowty]

This guy, I'm sure he must be a relative of Nowty , tests a 5000A fuse, when I say tests he actually destroys it using a capacitor bank and a few other bits and pieces. Great video and worth a watch:

I'm happier with 56V DC, its low enough to grab hold of both bus bars with dry bare hands.
My main fuse from my DC bus bars back to my Sunny Island is 200A on a 70m㎡ cable.

Back in the 80's as an Apprentice we used to charge small capacitors up to 400V then say to unsuspecting victims "catch this", they always did, well at least the first time.

[Oldgreybeard]

Back in the 80's as an Apprentice we used to charge small capacitors up to 400V then say to unsuspecting victims "catch this", they always did, well at least the first time.

First place I worked there was a mad Welshman that would charge up any high voltage electrolytics he could find to high voltages and then put them in people's bench drawers. It became a habit to put on rubber gloves before rummaging around anywhere he might have been. The H&S police would have heart failure at someone doing that today. I'm sure.

[Caesium]

Thanks to all who've posted so far, some great knowledge being shared here.

Realising that I probably need more substantial fuse protection, I dug around a bit and realised that the 100A DC breakers that I previously linked on ebay for £4.99 can also be bought from various retailers for more like £30, for example at
https://www.12voltplanet.co.uk/waterpro ... akers.html

This makes me worry that the ones on ebay might be fakes and to be honest this seems like a part of the build I shouldn't risk it on. So rather than resettable breakers, which just seem like adding more things to go wrong, I'll simplify to using link fuses: https://www.12voltplanet.co.uk/P00523.html along with a holder https://www.12voltplanet.co.uk/victron- ... older.html. Haven't bought these yet but do have in basket so if anyone has better recommendation please do shout

I'll be using 20mm cable for battery interconnects that I can't do with the supplied busbars, which should be good for 135A. And a set of long pylontech cables to branch off the Pylon stack. So to protect the cables 100A fuses should be good I think. I reckon I could go lower as during normal operation once balanced I'd expect no more than about 10 - 20A as the LF280s will share the work with all 6 Pylons.

[Oldgreybeard]

The fuse switches I used are definitely safe and fine for up to 160A IIRC. Bit more expensive at £45 for a two pole one with both poles fused, but they are well made and you can physically see that the thing is disconnected. Not sure I trust eBay, TBH, as there is a lot of crap around on there.

According to BS7671 (the UK wiring regulations) 25mm² cable is rated to carry 101A in conduit or trunking, 114A if clipped to a wall. If you want to carry more than this then you need 35mm², which is rated at 125A in conduit or trunking, 141A if clipped to a wall. Both these cable sizes have higher ratings when used in cable tray or just dangling in free air, but the UK wiring regs require that cables be secured and not left dangling. Not ever seen 20mm² cable, unless it's an American gauge converted to metric, perhaps.

My batteries are wired up with the cable supplied by Pylontech, which is 25mm².