Idiot's guide to a self build battery

[Joeboy]

I do love it when a Chinese label package arrives. Sadly not quite big enough to be the batteries.

[Joeboy]

That's the buckboard power supply professionaly cabletied to a bit of old board. Ran it up, tried the pot, all good.

Decided to check out the top voltage on the Pylontechs.

My batteries slowdown considerably after 88% soc, they can charge there for a long time then go quickly from 88% to 100%. Here's what I found.
88% SOC @ 51.3V
95% SOC @ 52.6V
100% SOC @ 52.9V
The above percentages are as indicated on the Goodwe hybrid inverter, there is lag so pinch of salt please.
Nonetheless, after it indicated 100%SOC the Victron continued to charge as did the grid and the solaredge, all of which I left doing their thing. At 53.3V the Victron shutdown and almost immediately the solaredge power was being exported. The grid charging had been off for about a minute by then as controlled by the Goodwe hybrid inverter.
Did another check after a few minutes and holding at 53.3V

Disconnected the can lead and the main feed leads to the stack and watched it self balance for about 6 minutes then all the Pylontech lights were out. A quick check of the voltage and it had fell back to 52.9V. Hooked it back up, reboot and all fine. I mention the change in voltage seen pre&post balancing to show that it does make a difference and is worth doing. I am choosing to do it on an annual basis even though its a small amount of power being redistributed (to be honest, I just like watching it do its thing).


The main thing I am taking away from this is that my particular Pylontech stack has a peak of 53.3V. Not sure what the rest of you guys top out at?

As each battery unit has 15 individual batteries within it in series I can take the 53.3V seen at 100%SOC and divide in by 15 which gives me a figure to stay below when balancing the new LF280 batteries on their Nov arrival. 3.553V being the figure.

That was fun.

[nowty]

The main thing I am taking away from this is that my particular Pylontech stack has a peak of 53.3V. Not sure what the rest of you guys top out at?

Very similar, I have typically measured about 53.2V when mine go into standby (all leds off). Even if you charge them at 54V (max charge V in datasheet) they still never reach that as they self throttle back the charge and go into standby. And they are practically full at 52.5V (min charge V in datasheet).

[Oldgreybeard]

My batteries slowdown considerably after 88% soc, they can charge there for a long time then go quickly from 88% to 100%.

Mine do the same, they sit at around 88% for ages, then next time I look they are at 100%. I'm not at all convinced the reported SOC is that accurate, TBH. I have the charge limited to 50A, mainly to limit the peak power drawn from the supply when charging overnight with other stuff on (to keep us under the 15kVA supply limit), which also means that more gets diverted to the water heating when it's sunny. At a steady 50A charge the battery packs should gain 10% SOC in about 50 minutes, but often the jump from the "sticky" 88% point to 100% is a fair bit less than this. If I had to guess then I'd say the 88% indicated SOC is close to around 95% in reality.

[Joeboy]

The main thing I am taking away from this is that my particular Pylontech stack has a peak of 53.3V. Not sure what the rest of you guys top out at?

Very similar, I have typically measured about 53.2V when mine go into standby (all leds off). Even if you charge them at 54V (max charge V in datasheet) they still never reach that as they self throttle back the charge and go into standby. And they are practically full at 52.5V (min charge V in datasheet).

Cheers Nowty, can you link in your XL sheet showing your SOC/percentage please? There are so many variables out their for these Pylontech units to be hooked up to. Would be nice to think we could help someone on the flyby?

Now on standby for a test, parallel balance, maybe a charge variance depending and we will be a go! Drums fingers, bangs head against wall as a distraction and waits for slow boat from China..

[Joeboy]

My batteries slowdown considerably after 88% soc, they can charge there for a long time then go quickly from 88% to 100%.

Mine do the same, they sit at around 88% for ages, then next time I look they are at 100%. I'm not at all convinced the reported SOC is that accurate, TBH. I have the charge limited to 50A, mainly to limit the peak power drawn from the supply when charging overnight with other stuff on (to keep us under the 15kVA supply limit), which also means that more gets diverted to the water heating when it's sunny. At a steady 50A charge the battery packs should gain 10% SOC in about 50 minutes, but often the jump from the "sticky" 88% point to 100% is a fair bit less than this. If I had to guess then I'd say the 88% indicated SOC is close to around 95% in reality.

I agree, logic and physics too I think. Voltage reading is king.
Once I am up and running with the LF280's I will be hunting down a nice big LCD screen to display battery voltage out there in the garage. TBH, I find the bank of green lights on each unit as good as anything but want to add some bells n whistles for the damn of it.

[Oldgreybeard]

When I posted the above our battery was sitting at 88% as it happens, and was charging at about 1200W. It's still charging at about 1200W now, some 40 minutes later, and the SOC is indicating 89%. Pretty good confirmation that the SOC does indeed "stick" at around the 88% level for some reason. During the past 40 minutes the battery has actually gained about 0.8kWh, so about 3.8% of true SOC change, not the 1% indicated.

[nowty]

Cheers Nowty, can you link in your XL sheet showing your SOC/percentage please? There are so many variables out their for these Pylontech units to be hooked up to. Would be nice to think we could help someone on the flyby?

No point showing you mine, as they are a blend of two battery chemistries and different cell numbers.

But below are my estimates of a 15 cell LFP battery/cell Voltage v SOC which should work for you. Its only an estimate so you might want to tweek it for your own setup once you get a feel for your own measurements.

LFP batteries have a very flat charge / discharge gradient and most of the action happens between 48.5V and 50V on a pack basis. These are my estimates at zero charge/discharge. So if your charging, the V is a bit higher and if your discharging the V is a bit lower than the values shown. Exactly how much depends on your battery size and charge/discharge amounts. If your checking it from your inverter display then you also have cable voltage drop to take into account unless your at zero charge/discharge.

[Joeboy]

Cheers Nowty, can you link in your XL sheet showing your SOC/percentage please? There are so many variables out their for these Pylontech units to be hooked up to. Would be nice to think we could help someone on the flyby?

No point showing you mine, as they are a blend of two battery chemistries and different cell numbers.

But below are my estimates of a 15 cell LFP battery/cell Voltage v SOC which should work for you. Its only an estimate so you might want to tweek it for your own setup once you get a feel for your own measurements.

LFP batteries have a very flat charge / discharge gradient and most of the action happens between 48.5V and 50V on a pack basis. These are my estimates at zero charge/discharge. So if your charging, the V is a bit higher and if your discharging the V is a bit lower than the values shown. Exactly how much depends on your battery size and charge/discharge amounts. If your checking it from your inverter display then you also have cable voltage drop to take into account unless your at zero charge/discharge.

Thanks Nowty, that is an enormously cool thing you have done. . I plan to buy a 2 screen units, one for garage and t'other for lounge . Would be nice have a built in ct/amp clamp at the sender end to have as much data as possible. WiFi would be nice too. I guess I'm half aiming towards drawing all the critical system data together in one. Hmmn, don't expect much do I?

As an aside, I very much enjoyed the afternoon's work as that's the first time I've really watched it all fall into place and knowing what to expect and when.

PMSL at the split stack/number/chemistries.

[Caesium]

Notification out of the blue that DPD will deliver my batteries today by 1:30

Didn't get any intermediate tracking info like customs or entering the country etc, just straight to delivery. Exciting though, now lets hope a box of bricks doesn't turn up