1 hour ago, OzzyMozzy said:

Hi @Tinbum

The invertor bulk charge setting is 53.2 V. Pylontech allows for a 0.8V margin so it should have stayed under 54V.

All batteries are connected to the invertor via the CAN cables.

Would you know how to pull the data from my remaining batteries? As you say it would be interesting to see if there were spikes on my remaining batteries and to see if they correspond to my failed 2.

Most people think that 53.2v is too high for the Pylontech batteries. (Victron limit their inverters to 52.4 or 52.5v (can't remember which)). I also limit mine to that voltage except for once a week.

Providing your CAN is working OK then the inverter should be limited to 53.2v by the BMS.

The attached file gives some details on the Pylon BMS, though this is for the C type batteries.

pylontech-official-statement-for-us2000c-us3000c.pdf

23 minutes ago, OzzyMozzy said:

closer look at the quality of the lithium that goes into the battery

I've started importing Solar batteries into SA and just to give you an idea of what testing the batteries are undergoing. I'm old-fashioned, I must believe in the product and feel confident that it's a quality product. Just a few of the tests conducted below;

1) Charge and discharge repeatedly at the specified C rating and then to measure the drift of the cell Volatge before and after. (A quick way to see if the cells are inferior or poorly matched). This is done over a 2-week period, constant charge, discharge at 1C.

2) Inspect internals of the battery, cell origion, construction, wiring, BMS settings etc.

3) Actual vs specified Ah, drain to 100% DoD repeatedly at C rating (1C in this case)

We basically punish the battery (painfully so) to see what it can actually do.

We then also test compatibility of battery with different inverters, does it talk nicely to the inverters.

All this to make sure the battery does what it says on the label....

Oh yes, the BMS did shut down the battery when we overloaded the battery current draw, both charge and discharge!

 

 

 

 

That impressive @FixAMess, out of interest, at 1C discharge, what do you see the cell difference go up to? Keen to compare, I get the impression around 0.1v is expected at max discharge (for around a 100Ah/5kW battery), but I've seen higher on my batteries (though they aren't new). Assume it varies a bit over the discharge cycle as well, less during flat voltage section and higher during the lower and upper SoC?

2 hours ago, FixAMess said:

I've started importing Solar batteries into SA and just to give you an idea of what testing the batteries are undergoing. I'm old-fashioned, I must believe in the product and feel confident that it's a quality product. Just a few of the tests conducted below;

1) Charge and discharge repeatedly at the specified C rating and then to measure the drift of the cell Volatge before and after. (A quick way to see if the cells are inferior or poorly matched). This is done over a 2-week period, constant charge, discharge at 1C.

2) Inspect internals of the battery, cell origion, construction, wiring, BMS settings etc.

3) Actual vs specified Ah, drain to 100% DoD repeatedly at C rating (1C in this case)

We basically punish the battery (painfully so) to see what it can actually do.

We then also test compatibility of battery with different inverters, does it talk nicely to the inverters.

All this to make sure the battery does what it says on the label....

Oh yes, the BMS did shut down the battery when we overloaded the battery current draw, both charge and discharge!

 

 

 

 

That's impressive! Would you be able to recommend a BMS to work with my 2 X Mecer SOL-I-AX-5M invertors (circa 2015) and 2 X US2000 and 1 X US3000C Pylontech batteries? I don't trust the built-in BMS, for obvious reasons...

And where do sell your batteries btw?

I have 4 X US3000 Pylontech batteries and regularly check up on the cells with Batteryview and yes, every single battery has over voltage errors logged. Usually when one introduces a new battery to the bank it takes a few cycles to settle down, irrespective of whether or not all the batteries are charged to 100% then left overnight to equalise etc. The point is, I have taken care not to wreck the batteries but there are still issues logged.

So who knows....I don't have a problem with the pylontechs but will they last 6000 cycles, 10 yrs? I doubt it. I'm sure in 10 yrs time the battery technology will have changed so much that our current day batteries will not be supported anyway.

 

19 hours ago, Tinbum said:

. (Victron limit their inverters to 52.4 or 52.5v (can't remember which)).

That's only because of their topology and not being able to regulate fast enough, so they just lowered it until they stopped getting high voltage alarms. If you have an inverter with a decent control loop, you won't have that issue. 53.2V is 3.5466V per cell, which is not at all a problem for LiFePO4.

4 hours ago, FixAMess said:

 yes, every single battery has over voltage errors logged. Usually when one introduces a new battery to the bank it takes a few cycles to settle down, irrespective of whether or not all the batteries are charged to 100% then left overnight to equalise etc. The point is, I have taken care not to wreck the batteries but there are still issues logged.

 

Yes, but I guess they are cell overvoltage errors and not battery pack overvoltage errors. Their is a big difference.

Edited June 14, 20233 yr by Tinbum

3 hours ago, P1000 said:

That's only because of their topology and not being able to regulate fast enough, so they just lowered it until they stopped getting high voltage alarms. If you have an inverter with a decent control loop, you won't have that issue. 53.2V is 3.5466V per cell, which is not at all a problem for LiFePO4.

Really, Vicron say that.

Many would disagree with taking the voltage that high, their is little point and it's always best to leave a margin.

39 minutes ago, Tinbum said:

Really, Vicron say that.

Many would disagree with taking the voltage that high, their is little point and it's always best to leave a margin.

Who are the many that disagree? If you make a list of batteries on the market and what they charge their cells to, you'll see that is on the lower end.

24 minutes ago, P1000 said:

Who are the many that disagree? If you make a list of batteries on the market and what they charge their cells to, you'll see that is on the lower end.

Its been mentioned on this forum and another that i am a member of many times. You only have to look at the charge curve to see.

Edited June 14, 20233 yr by Tinbum

1 hour ago, Tinbum said:

Yes, but I guess they are cell overvoltage errors and not battery pack overvoltage errors. Their is a big difference.

Yes, unfortunately I cannot see over voltage errors by cell, only a number of times that a "battery" has had over voltage errors logged, but yes, the batteries have never shut down or displayed red error lights.

6 hours ago, FixAMess said:

So who knows....I don't have a problem with the pylontechs but will they last 6000 cycles, 10 yrs? I doubt it. I'm sure in 10 yrs time the battery technology will have changed so much that our current day batteries will not be supported anyway.

 

This may interest you- it is for the none C batteries but I would imagine they should be very similar.

Very true- their are some amazing ones already out already there but still very expensive at the moment.

48V Pylontech Battery Performance 0320.pdf

Guys I hate to say that one is better than the other. But that is one big reason to buy a FreedomWon battery. Local support & they entertain your battery issue should it arise. Why buy a product that has no local OEM presence. Another issue: A BMS is designed to protect whatever is thrown at it. Complete rubbish that the BMS cannot do this job, then they say you have cooked the Battery. 

I'm sorry I just can't accept that statement, otherwise it's not up to the task to begin with. 

I'm not saying one is better than the other but having dealt with FreedomWon, they have provided investigation, service as well as facilitated return of the units to the labs to service cells & BMSs. 

Must admit it's shameful to see ppl spend money then request support the get it refused. 

 

19 hours ago, Tinbum said:

Its been mentioned on this forum and another that i am a member of many times. You only have to look at the charge curve to see.

It seems to me that it's been mentioned a bunch of times, but only by you. The charge curve does not really provide much insight into why you would choose a lower voltage. It all depends on what the chemistry does when you do go over this voltage, and in the case of LiFePO, the answer is not much (up to around 3.65V).

4 hours ago, P1000 said:

It seems to me that it's been mentioned a bunch of times, but only by you. The charge curve does not really provide much insight into why you would choose a lower voltage. It all depends on what the chemistry does when you do go over this voltage, and in the case of LiFePO, the answer is not much (up to around 3.65V).

The charge curve shows it all. Their is very little point taking the voltage that high as you gain very very little extra capacity.

Some examples,
https://powerforum.co.za/topic/19327-best-practice-for-charging-litium-battery/#comment-168095

https://powerforum.co.za/topic/6807-guidance-requested-on-shoto-sda10-48100l5-battery-settings/#comment-154152

From the Victron web site.

Seplos BMS settings for their packs.

Edited June 15, 20233 yr by Tinbum

1 hour ago, Tinbum said:

The charge curve shows it all. Their is very little point taking the voltage that high as you gain very very little extra capacity.

Some examples,
https://powerforum.co.za/topic/19327-best-practice-for-charging-litium-battery/#comment-168095

https://powerforum.co.za/topic/6807-guidance-requested-on-shoto-sda10-48100l5-battery-settings/#comment-154152

From the Victron web site.

Seplos BMS settings for their packs.

Your victron link says exactly what I did - they get overvoltage alarms on victron equipment, but they won't on equipment with a fast enough control loop. The snippet from seplos does not really say anything at all - just their standard settings.

 

While it is true that there is very little extra energy stored by going higher voltages, it has no detrimental effects to the cells - and might be needed to balance the cells, depending on the BMS config.

Victron equipment has a reputation for being good and of a high quality as does the SMA equipment I use. They have been around for years and are specialists in their field unlike a lot of the manufactures that are around now. I would take their advise on settings.

Yes the cells are designed for a max voltage but it doesn't mean that you have to take them to that voltage. A BMS that requires them to go to that voltage to balance should be put straight in a dustbin.

Their is no benefit of taking them to max voltage and allowing a margin is good practice and advisable. Any system, be it quality or rubbish, is going to find it hard to control voltage spikes when large loads are suddenly turned off. I dont believe for one minute that Victron have a problem with their control loop- your making assumptions.

Watching from the sideline, amused...

Pylontech seems known to run their batteries hard. 15 cells vs 16 and such. One of my systems have 2 x US3000 on a Victron inverter and I occasionally see slight over voltage warnings, duly ignored. My standard installation 14s 100p 18650 Li-Ion cell batteries (don't laugh, they work very well) run a very conservative 44-55V profile regulated by the inverters, sans BMS. When I occasionally don my BMS hat, I sometimes replace a few cells, but then, they are virtually free... Ran a 53s 34p backup bank the other day and thought to check health: 34 changeover switches to the 100W test load makes that easy, and not a single cell needed changing. NMC chemistry still rocks! My 12yr old Samsung laptop still has 3 hour battery life. (Used Samsung battery management, 80% max.)

Wow Wow - Pylontech commenting that the BMS is "passive"? really.

I rest my case / hat / cape  what ever you prefer - as someone noted we're into the first 5yr window of battery/inverter warranties and the alarming overtone is that warranties seem to not be worth the paper they're printed on and of course the importers and dealer are dare I say it as unscrupulous as always - any excuse to not honor a warranty.

I personally just on the comments thus far wont look at Pylontech - there is an alarming read between the lines situation brewing - the battery swelling issue is not new and has been happening for years with various 2000 and 3000 series models probably more than most people know/admit/report

The bottom line is this - international standards which lets be honest seem to mean little in SA dictate that LiFePO4 and other/simlar Li-chemistry are required to have a  BMS, Battery Management System, a mandatory component for LiFePO4 batteries. LiFePO4 or all lithium battery cells are sensitive to over-voltage, under-voltage, and over-current. If a LiFePO4 battery is kept under one of the above conditions for a long time, it can easily cause capacity degradation, battery damage, or even the risk of fire. There was an EU/EE statement I read once that was much like the above and added ..."should invoke (I love that word) a hard disconnect from the inverter to circumvent the damage and or fire and or exhausting of chemical gases or other unsatisfactory expulsions at all costs" I've butchered it a bit most likely, but I know some manufactures actually employ a solenoid/engergised disconnect that physically breaks the incoming charge or out going load if an inverter doesn't respond in xx millisecond of a fault being registered.

If Pylontech want to assert that they have a Passive BMS then I think they should be "passive aggressively" banned from selling their product here unless they can demonstrate that their BMS does what its actually supposed to do - protect the internal cells and pack as a whole from what they are claiming is the cause of the issue(s) experienced by users.

Out of interest and this isn't a "plug" for DEYE inverters, but one of the options is a STOP on BMS error which I've seen many an installer disable because its considered a "nuisance" but what it does is when there is a BMS error it can  disconnect from the battery and isolate the inverter and loads and battery(s) from each other - as far as I'm concerned it should be a password protected option that cannot be unchecked and only in very special circumstances with an installers ID. I have mine on and with my "badly misbehaving" CFE batteries it does exactly that in mere milliseconds.

Then again I think half the batteries out there in SA would never work if that was a requirement because BMS issues seem to be a large problem from scouring this forum and we need to ask ourselves - what scarp is being dumped on our shores that wouldn't pass as safe in other countries?

The reference to 'Passive' is as opposed to an 'Active' balancer.

Edited June 16, 20232 yr by Tinbum

Hi All - re the active vs passive issue of the PylonTech BMS:

I have shared the following in another thread. As it throws a bit of light on whether the BMS is active or passive I thought to share it here again. 

I have a MultiPlus II 48/5000/70 with a Cerbo GX linked to two Pylontech UP5000 batteries. I ran the following measurements:

1)    2x UP5000 batteries with a CAN interface cable from the master battery to the Cerbo, no ESS.

In this case the BMS managed the charging. The VE Config settings were essentially ignored.

2)    2x UP5000 batteries without the CAN interface cable, no ESS

In this case the Victron managed the charging and followed the VE Config settings

3)    1x UP5000 battery with the CAN interface cable but with ESS activated. This was done at a stage when I had only one UP5000 battery. The inverter seems to be more ‘active’ in managing the batteries. However the interaction between the inverter and BMS is not that clear.

The following graphs shows the Voltage/Current and SoC for the three scenarios. .

1) With CAN cable: PylonTech BMS Managed Charging (Date of graphs: 3-5 June 23)

2X UP5000 (Simple implementation, no Assistants)

The BMS seems to manage the charging with only the maximum charging current provided in VE Config/Remote Console is enforced by the MP2. The BMS sometimes shows that it will use 100A to recover from a low SoC. However I specified 60A in VE Config and 50A in the Remote Console as maximum charge current for the two UP5000 batteries – the maximum observed was 50A. Other parameters like Absorption and Float voltages are BMS prescribed. The charger sequence (e.g. per status lights) follows a combination of Bulk, Absorption and back to Bulk sequence. It never goes to float. This confirms what @JKKM indicated in his reply in another thread:

"When Victron released version 2.90 of the VenusOS for the Cerbox GX one of the changes as per the release notes was for the following:

·          Inverter/chargers: show state as "Ext. control"; instead of Bulk or Absorption, when controlled by a BMS/DVCC. When controlled by a BMS the inverter/charger would never show float, which caused questions. Now it just shows externally controlled, simpler, and more accurate: there is no bulk -> absorption -> float charge algorithm when controlled by BMS."

All that is abundantly clear is that the BMS controls the charging. After a full charge, the voltage stays roughly constant at 52.3V for some period and then slowly drops to just below 50V. At some stage at about 49.8-49-9V the SoC will show a small drop and that seems to be the signal to initiate a small kick back charge to 52.3V.

Interesting is that the period that it takes for the voltage to drop and the period how long it can stay constant at about 49.85-49.9V is quite variable.  These variations are a bit weird – I am sure there is an explanation…

The following diagrams are screen dumps of the voltage/ampere and SoC curves. It includes one loadshedding period.

 

 

2) No CAN Cable: Inverter Managed Charging (Date of graphs: 20 May 2023)

X2 UP5000; Absorption @ 52.3V, Float @ 51V, No assistants

(Window includes a loadshedding period that ended at about 01:50)

This shows that the Victron is managing the charging in that the charge curves adheres to the VE Config settings. It goes through the expected Bulk, Absorption and Float stages. It is “nice and clean”

 

3) ESS with CAN cable: BMS + Victron Managed Charging(?) (Date of graphs: 3-5 April 2023 – including a number of loadsheddings)

X1 UP5000; Absorption @ 52V; Float @ 51V.

Charging seems to be done under some combination of BMS and Charger control. Voltage and current curves look messy with continually changing voltage and small current ‘kicks’, even the odd small push back into the main supply. The charger did not follow the  charging sequence of Bulk, Absorption and Float. (I need to understand this better.)

 

 

A big thank you to @P1000 and @BritishRacingGreen for taking the time to actually look at the data logs of my batteries.

I contacted Pylontech China directly paraphrasing what the both of you mentioned and this was their response:

"A few days ago, in the communication with Segensolar, there were two cases where this situation occurred. After re-checking the log, I think that the cells have been unbalanced before the overcharge data occurs. So I changed the warranty judgment and replaced the new batteries under warranty for the customer.

 

We will continue to pay attention to similar cases and provide customers with warranty services."

 

I was then offered 2 X US2000C batteries as replacements but decided to take a credit note instead.

 

I think many people are correct in that to get any company in South Africa to honour a warranty is next to impossible BUT not impossible!

 

They rely on customers not really understanding the technology and baffling them with multipage data logs in the hope that you just give up.

 

Thank god for this forum and the people on it who willingly help out strangers, you literally saved me a small fortune!

 

It is however a long, long process (nearly 6 months!) but worth exploring every avenue before throwing in the towel.

 

I hope this might help some other members in their warranty claims as there does seem to be gremlins in those US2000 series batteries...

 

Thanks again!

1 hour ago, OzzyMozzy said:

A big thank you to @P1000 and @BritishRacingGreen for taking the time to actually look at the data logs of my batteries.

I contacted Pylontech China directly paraphrasing what the both of you mentioned and this was their response:

"A few days ago, in the communication with Segensolar, there were two cases where this situation occurred. After re-checking the log, I think that the cells have been unbalanced before the overcharge data occurs. So I changed the warranty judgment and replaced the new batteries under warranty for the customer.

 

We will continue to pay attention to similar cases and provide customers with warranty services."

 

I was then offered 2 X US2000C batteries as replacements but decided to take a credit note instead.

 

I think many people are correct in that to get any company in South Africa to honour a warranty is next to impossible BUT not impossible!

 

They rely on customers not really understanding the technology and baffling them with multipage data logs in the hope that you just give up.

 

Thank god for this forum and the people on it who willingly help out strangers, you literally saved me a small fortune!

 

It is however a long, long process (nearly 6 months!) but worth exploring every avenue before throwing in the towel.

 

I hope this might help some other members in their warranty claims as there does seem to be gremlins in those US2000 series batteries...

 

Thanks again!

I also looked in great detail at your data logs and I really think you need to look at how your system is set up for what ever batteries you next use and the ones you still have.. I think you have been very very lucky. (I'm sure I asked for the logs for the time before the ones you supplied).

Edited July 6, 20232 yr by Tinbum

Hi @Tinbum

Any advice would be greatly appreciated on the settings. Unfortunately I don't have the knowhow on extracting the data from my other US2000 units.

Currently I'm using 3 x US3000C and the older 2 x US2000B.

Dip switches set as per Solar Invertor Warehouse recommendations are:

US3000C Master. Dip switch 1. ON. All others off.

Rest of the batteries Dip switches are off.

Invertor: 2 x Mecer SOL-I-AX-5M

Program 02: 50 Amp

Program 05: User defined

Program 12: 48V

Program 13: 51V

Program 26: 53.2V

Program 29: 46.5V

The invertor is connected to the batteries via this cable: WIOSCAN35GL1

Regards

 

Are you using the batteries as a single pack?

I presume you have a master inverter and a slave inverter connected with an interconnecting communication cable.

I don't know your inverter but looking at the manual it seems to use RS485 communication to the battery.

If your using the US3000C as a master (which you should be) then your communication cable is wrong judging from a picture I found online for that cable. See attached. Have a look in the manual for the US3000C battery for the battery cable end. (pins 1 to 3 should not have any connections on them!).

 

Edited July 8, 20232 yr by Tinbum