Since the newer ICC software (2.4.3 and 2.4.4) has been used,  each Pylontech battery values could be read independently.

I have noticed the last 2 days (using the batteries every night) that batteries number 1, 2 and 3 discharge almost at the same rate, but that battery number 4 is discharced at least 20% more than the other 3.  Is that normal and is there something I can do about it?

I will try and take a picture tomorrow when it is discharged to a SOC of between 35% and 40%. The current picture indicates fully charged batteries.

This is how they are set up at the moment.

 

Edited August 1, 20186 yr by Daniemj
Added picture

Try this

The other thing I have noticed is that if you mix firmware types for the Pylontech you end with some errors in reporting. My units report a collective SOH of 51% but individually they report 100%.

2 hours ago, Daniemj said:

I have noticed the last 2 days (using the batteries every night) that batteries number 1, 2 and 3 discharge almost at the same rate, but that battery number 4 is discharced at least 20% more than the other 3.  Is that normal and is there something I can do about it?

Note that battery 4's temperature is lower too. It might very well have something to do with it. It is common for the different cells to not have quite the same capacity. The battery is empty when the lowest cell drops too low. The battery is full when the highest cell is too high. The balancing circuitry will over time get them closer together. The balancer tends to create some heat. Also, the one with the higher cells might in fact run a bit cooler and might therefore be tasked with providing a bit more power.

In other words, this might be normal. Don't know for sure though.

3 hours ago, Daniemj said:

Since the newer ICC software (2.4.3 and 2.4.4) has been used,  each Pylontech battery values could be read independently.

I have noticed the last 2 days (using the batteries every night) that batteries number 1, 2 and 3 discharge almost at the same rate, but that battery number 4 is discharced at least 20% more than the other 3.  Is that normal and is there something I can do about it?

I will try and take a picture tomorrow when it is discharged to a SOC of between 35% and 40%. The current picture indicates fully charged batteries.

This is how they are set up at the moment.

 

In a DC circuit, during a discharge, electrons flow from negative to positive, hence causing what you observe here. 

When being recharged, electrons flow from positive to negative. 

2 hours ago, Chris Hobson said:

Try this

The other thing I have noticed is that if you mix firmware types for the Pylontech you end with some errors in reporting. My units report a collective SOH of 51% but individually they report 100%.

Thank you. This makes sense to me since I have added a new battery to 3 fully charged batteries that has been in use for about 2 weeks.

4 minutes ago, Daniemj said:

I have added a new battery to 3 fully charged batteries

The battery with the better capacity will provide more current during discharge. This is natural and happens for other chemistries too (including lead acid). The trouble is usually on recharge, but for Lithium batteries (where their impedance increase significantly as they become full) there is also more of a natural split of charge current.

In short... they just need a bit of time to "find each other".

Some batteries even have a specific procedure that has to be used. I know that with both LG and BYD, you can't just throw another module into the rack without doing this. I am not sure what the procedure is though.

This is what it looks like this morning.

Going to do the battery balancing as per video posted by Chris.

As you could see, the SOC of the first 3 batteries range between 36% and 37%, but battery number 4 have a SOC of only 21%.  20% SOC is the minimum what the batteries want, so this difference is huge. I have tried the battery balancing tonight, so I will see what happens tomorrow. The initial picture showed that the SOC was 100% on all 4 batteries before they have been used through the night. They also powered the geyser (2000W element ) for 35 minutes between 23H55 and 00H30.

 

I have not balanced mine and noticed that the SOC difference is much smaller after 2 weeks. Now at 6:41 this morning  SOC is 69, 69, 68, 70. The last two are the older modules.

15 minutes ago, Chris Hobson said:

difference is much smaller after 2 weeks

According to the information I have, these batteries also balance at the top end. The way to get them fully balanced will be to hold them at something above 52V for a long time (like more than a day) to give the balancers time to work. The balancer usually transfers really small amounts of current, so this takes time.

Or as Chris says, just forget about it and give them time. They will balance a little bit every day, getting closer by the day when you recharge and eventually arrive fully in sync.

My Victron batteries did something similar. For the first week or two, the 12.8V module on the right was warmer than the one on the left, and the bluetooth app showed the cell voltages were higher. It evened out in the end.

Thank you very much guys. It is good info. Will give it time.

The uneven discharge did not get better. I bought the batteries from Gary Bowes at Solar and Inverter Warehouse in Meadowdale, Johannesburg. (R13750 each, Vat incl) 

He followed it up with Pylontech but he said they mentioned something about firmware but also that it is within specs. They will take some time before they change it, so he offered to change it for me now. I  wanted a 5th battery so I bought another yesterday (at the same special price) while changing the  problem one for a new one. 

Today was the first morning, and I am happy to see the SOC of all 5 batteries are close to each other! Problem solved, thanks to excellent service from Gary!

2 hours ago, Daniemj said:

The uneven discharge did not get better. I bought the batteries from Gary Bowes at Solar and Inverter Warehouse in Meadowdale, Johannesburg. (R13750 each, Vat incl) 

Good price!

I’ve got 4 batteries but not stacked on top of each other.

I bought the Full Circle Solar cabinet where the batteries stand on 2 levels.

Initiall I connected them the same way as when stacking it but then split the connections to the inverter.

A pleasing result of this was that the charging amps (and off course the same when discharging) halved. This can just be beneficial to the system.  Running 24a vs 49a per cable should make a difference somewhere.

I’ve got 4 batteries but not stacked on top of each other.

I bought the Full Circle Solar cabinet where the batteries stand on 2 levels.

Initiall I connected them the same way as when stacking it but then split the connections to the inverter.

A pleasing result of this was that the charging amps (and off course the same when discharging) halved. This can just be beneficial to the system.  Running 24a vs 49a per cable should make a difference somewhere.

Hi

I have 2 x US2000B batteries about 1 month old, can I add a new  US3000 to the pack or should I stay with US2000(if I can find any)

On 2019/05/15 at 5:01 PM, Kite said:

Hi

I have 2 x US2000B batteries about 1 month old, can I add a new  US3000 to the pack or should I stay with US2000(if I can find any)

You can add the US3000, but it has to be the new master battery, i.e. its "link 0" socket is empty, and its link 1 is connected to the link 0 of your second battery, and so on.

Hi

I use 2 US3000 batteries with a Mecer MKSII inverter.

The installer disabled the battery equalization function (program 33), is this right?

On the ICC I see that they charge/discharge absolutely the same.

If you people talk about "balance" is this the same as equalization?

1 hour ago, Bert sa said:

The installer disabled the battery equalization function (program 33), is this right?

This is correct, a lithium battery does not get an equalization charge.

1 hour ago, Bert sa said:

If you people talk about "balance" is this the same as equalization?

No, this is not the same thing. A balance charge is where the BMS will make sure that all series connected cells in the battery are at an equal (or very close to equal) charge. Thankfully it does this on its own. This normally happens when the battery is fully (or close to fully) charged.  

On 2019/06/19 at 9:18 PM, Bert sa said:

"balance" is this the same as equalization?

Not the same thing, but achieves the same purpose (nearly equal state of charge for a set of series connected cells). Equalisation is a crude overcharge of all cells; this is acceptable with flooded lead acid batteries where you can replace the water; with any lithium based chemistry including LFP, it has to be done another way That's usually resistive discharge of the fuller cell(s), though there are other schemes.

Hi All,

So for those that has been running Pylontech's for a while now, how fast does your Total AH drop?

I have had my 2X 3.5Kw Pylontech batteries for just over a week now and noticed that my "Total AH" Capacity has already dropped from 148 AH to 147.93 AH?

Is this normal? At this rate i will never see 10 years life?

Hi @Valken

I have 7x US3000, 74Ah per brick nominal. Being in production from 3/2019. As of today, these are the reported capacities for the each brick:

73.904 Ah
74.004 Ah
73.864 Ah
73.092 Ah
74.008 Ah
74.004 Ah
73.204 Ah

Nominal capacity = 518.00 Ah
Reported capacity = 516.08 Ah

Ping me after couple of weeks and I will post an update here.
 

BTW, you can check your bricks using the diagnostic software: https://powerforum.co.za/topic/2322-youdas-off-grid-lab/page/2/?tab=comments#comment-53641

1 hour ago, Youda said:

BTW, you can check your bricks using the diagnostic software: https://powerforum.co.za/topic/2322-youdas-off-grid-lab/page/2/?tab=comments#comment-53641

@Youda You don't have (or know where to get) a Linux version of Watchpower?

2 minutes ago, JacquesVDM said:

@Youda You don't have (or know where to get) a Linux version of Watchpower?

Here you go @JacquesVDM

http://www.mppsolar.com/manual/Watchpower (off-grid)/Watchpower 1.13/

Edited July 17, 20195 yr by Valken

@ValkenThanks!!

Oeps, made a mistake: wanted to ask for a Linux version of Batteryview (as per Youda's link).

(But will download Watchpower as I need that one as well.)