Idiot's guide to a self build battery

[Joeboy]

Well that was the blind casual test!.Went to bed at half 6 and slept like a log for 12 hours. During that sleep the stack did its thing. It went from 11% soc as shown on the hybrid inverter all the way up to 88% and held there.

That doesn't make sense unless there is a venturi between pylontech and LF280 pack.

This seems to be the case as the pylontech pack has shown a 10% decline over 3.5 hours while two nights ago the same unaugmented pack showed a 5% decline over the same window.

Ergo, power is continuing to transfer from pylontech stack to LF280 packs after Go faster window has closed and the whole twin pack settles out.

Hopefully we'll have a greater starting soc for the charging cycle tonight and we'll see how it goes at the top end. There's a bit of space for the PV to fill while I open up the shading forest canopy a little today.

Note- the system charged at 3.35kWh to 3.45kWh across the whole window so we'll call that 17kWh in(hooray). Background load is about 240W averaged per hour in the home (3 freezers & 2 fridges). Kettle has been on once at 06.30hrs and twin stack is sitting at 76% soc & 3.32v showing at 06.50 hrs. This is great!

Added Nowty's excellent cheatsheet for reference.

[Caesium]

Hi Joeboy

Yes I expect this is due to imbalance between SOC between the packs. I saw similar oddities for a good few days when hooking things up. I fiddled, probably unnecessarily. I'd just leave it alone for at least a week and see if it resolves itself.

I've also noticed that the charge/discharge profile seems ever so slightly different between the Pylontechs and the EVEs. That is to say, while discharging from say 100% to 70% of SOC, both supply current fairly evenly. But then there comes a time where the Pylontechs actually stop supplying current to the load and basically hand over to the EVEs; the Pylontechs actually sit idle for a while, staying at 70% SOC, but the stack is still clearly supplying the house as my import remains at zero

After an hour or so (will depend on a lot of factors obviously like size of load) the EVEs discharge sufficiently that the Pylontechs come back into play and start supplying current again. You might notice it if you keep an careful eye on the Pylontech lights (or if you obsessively graph everything like I do)

[Joeboy]

Hi Joeboy

Yes I expect this is due to imbalance between SOC between the packs. I saw similar oddities for a good few days when hooking things up. I fiddled, probably unnecessarily. I'd just leave it alone for at least a week and see if it resolves itself.

I've also noticed that the charge/discharge profile seems ever so slightly different between the Pylontechs and the EVEs. That is to say, while discharging from say 100% to 70% of SOC, both supply current fairly evenly. But then there comes a time where the Pylontechs actually stop supplying current to the load and basically hand over to the EVEs; the Pylontechs actually sit idle for a while, staying at 70% SOC, but the stack is still clearly supplying the house as my import remains at zero

After an hour or so (will depend on a lot of factors obviously like size of load) the EVEs discharge sufficiently that the Pylontechs come back into play and start supplying current again. You might notice it if you keep an careful eye on the Pylontech lights (or if you obsessively graph everything like I do)

Thanks Caesium, my expectations are high for these units going forward. We missed an hour of charge on the way home from Turkey as the Chinese Goodwe servers are still running on old time. I've adjusted at my end. This missing hour pushed us down to 11% SOC last night as did SWMBO incredible post holiday washing machine frenzy.

Another handy little wrinkle in the system is that the off grid Victron system will have access straight to the LF280's so there might be a 1/10th of a Volt or so extra to play with there on the Victron shutdown settings come Summer time. Allowing the LF280's to act as a slight further store beyond the capacity of the Pylontechs? I have no clue as to the internal wiring for the Pylontechs and what they actually shut down internally. Just a thought.

I am going to hunt down a wee unit to display cell voltage on the LF280's, would like to go to 3 digits beyond the point but have not found anything yet.

[Stinsy]

Hi Joeboy

Yes I expect this is due to imbalance between SOC between the packs. I saw similar oddities for a good few days when hooking things up. I fiddled, probably unnecessarily. I'd just leave it alone for at least a week and see if it resolves itself.

I've also noticed that the charge/discharge profile seems ever so slightly different between the Pylontechs and the EVEs. That is to say, while discharging from say 100% to 70% of SOC, both supply current fairly evenly. But then there comes a time where the Pylontechs actually stop supplying current to the load and basically hand over to the EVEs; the Pylontechs actually sit idle for a while, staying at 70% SOC, but the stack is still clearly supplying the house as my import remains at zero

After an hour or so (will depend on a lot of factors obviously like size of load) the EVEs discharge sufficiently that the Pylontechs come back into play and start supplying current again. You might notice it if you keep an careful eye on the Pylontech lights (or if you obsessively graph everything like I do)

+1 to all of this.

You can expect the new batteries to require a dozen or so cycles before they sync up with the Pylontechs.

[nowty]

At high charge rates even small differences in cable lengths will result in slightly differing charge voltages on the actual packs because of the voltage drops in the cables. Then if the charge rate drops or finishes, the voltage drop in the cables reduces and you have two parallel packs with slightly different voltages so the higher one discharges into the lower one whilst they equalise.

You will have a similar effect after a long heavy discharge.

Its your new normal.

[Joeboy]

Got a few minutes to start the tidying up. We have 3A flowing into the LF280's from Victron PV. Just going to enjoy this quiet wee moment before the whirlwind that is SWMBO arrives in the garage.

bp amoco gas station locations

[Joeboy]

At high charge rates even small differences in cable lengths will result in slightly differing charge voltages on the actual packs because of the voltage drops in the cables. Then if the charge rate drops or finishes, the voltage drop in the cables reduces and you have two parallel packs with slightly different voltages so the higher one discharges into the lower one whilst they equalise.

You will have a similar effect after a long heavy discharge.

Its your new normal.

Cheers Nowty, and a beautiful place it is to be too.

[Joeboy]

Got it tidied. Nowtys cheatsheet in place for easy reference. Absolutely excellent watching it work as I did my bit. I'll add an ampclamp with screen and individual cell monitor and that should be enough.

Thank you to all of you for your assistance and advice during this upgrade.

[Caesium]

Looking smart! Very nice.

[Stinsy]

Really tidy! Nice work!

Any plans to swap the Pylontechs for another string of LF280ks?

You've got £5400 worth of Pylontechs there!