These forums have been an excellent source of information for my problem.

After consulting several topics, I am seeking closure and recommendation on using lithium batteries on an Axpert (Mecer) inverter - SOL-I-AX-5P

It seems that the inverter has a RS232 coms port and three are Mecer lithium battery coms boxes to modulate the inverter charge performance based on input from the lithium battery BMS, see here: https://wcctv.co.za/wp-content/uploads/2022/04/SOL-I-AX-BMS-BOX-Manual.pdf

I'm considering a Dyness A48100 battery but i believe the Pylontech batteries can also work.

I've also been advised that the 48vdc charging circuit of the inverter is insufficient to charge such lithium batteries - which doesn't sound right as the charging voltage (at least on ac charging) can be set as high as 58vdc, where the Dyness has a max charge voltage of 54vdc.

What is the consensus on this application and what are the latest recommendations. Should I operate simply on voltage charge or should I get the coms box or is there a way to flash new firmware onto the inverter for this application.

I  have not committed to anything yet, only the inverter which has given me 5 good years of service using lead acid batteries. I am now ready to upgrade and want to make the best decisions possible.

Looking forward to the discourse. 

@Pie No problems to charge 3x Pylontech US 3000 batteries with 5 kW Axpert. Use Solarassistant to monitor the state of batteries and monitor and do settings of Inverter remotely. Battery tipe set to User. Float voltage set at 52.5v Using this setup for more than 4 years . No communication between Inverter and batteries. Can also use multiSIB software with Solarassistant cable to check battery .There is a SOP for Pylontech batteries with Axpert inverter. Reduced floatcharge from 53.2v to 52.5v for better battery life suggestion  of @Coulomb

7 hours ago, Pie said:

I've also been advised that the 48vdc charging circuit of the inverter is insufficient to charge such lithium batteries - which doesn't sound right as the charging voltage (at least on ac charging) can be set as high as 58vdc, where the Dyness has a max charge voltage of 54vdc.

? But the maximum charge voltage isn't what determines whether there is enough grunt to charge three LFP batteries adequately. You just need enough voltage capability.

So then it comes down to charging current or power (more or less similar considerations since the battery voltage is approximately constant). And you can just charge the battery more slowly if you run out of current/power. But with frequent load shedding and intermittent solar power, you may want a lot of charging power so that you can be ready for the next load shedding.

Thank you @Chris Louw and @Coulomb for the prompt responses guys,

So are we saying that the lithium battery functionally is not lost when an Axpert units is used without the BMS to inverter interface? I just charge on voltage and make sure float and bulk charge voltage are set correctly? I was told that this approach results in the battery only being charged to 70% of capacity? 

@Chris Louw I will look into solar assistant, are you saying that solar assistant can communicate and module the axpert charging settings as required based on BMS feedback?

@Coulomb I also thought it comes down to power, so when the guy that sold me the axpert said it didnt have the legs (becasue of voltage) to charge the lithium i thought that was abit strange. I see on ace charging i can get up to 60A charging and there is 4000W on solar charginging mode. From the axpert rating plate Im abit concerned that the solar charging is at nominal voltage of 48V - again not sure if i am reading the specs wrong.

I am very budget sensitive so I am thinking of getting one Dyness unit to try and test concept and maybe another one with panels in january. I just need something urgently as my current AGM bank is "kaput". I think this should be more than what i need and should be markedly better than even the best performance I got out of this AGM bank

8 x Omnipower 60aH units, two banks of 4 in serries. I dont think I could even get 0.5kW out of it. Hoping performance with lithium will be significantly better.

Regards

 

I have spoken to Hubble today.

They have advised that their batteries are compatible with this inverter and recommend using the cloudlink device to monitor battery performance and communicate between battery bms and inverter. I do like the one stop shop solution in terms for the battery and monitoring system. I just got a RasberryPi but that has been commited to running octopi and 3d printing. 

So, I am thinking of getting one 5.12kW 1C battery this month and seeing how it goes before committing to a second one. 

Not sure if the unlimited cycles is a gimick or is actually worth the 2kZar price difference. Any comments?

I would go with pylon but the installation location requires a wall mount solution. 

13 hours ago, Pie said:

 

I have spoken to Hubble today.

They have advised that their batteries are compatible with this inverter and recommend using the cloudlink device to monitor battery performance and communicate between battery bms and inverter. I do like the one stop shop solution in terms for the battery and monitoring system.

 

So glad i found this thread:

Looks like the cloudlink solution is “’n goeie plan wat sleg werk".

 

On 2023/11/23 at 8:01 PM, Pie said:

So are we saying that the lithium battery functionally is not lost when an Axpert units is used without the BMS to inverter interface?

There is one scenario where you lose out without the cable. That's when one cell goes to too high a voltage, and the only way to protect it is to reduce the charge current. Without the cable, the inverter wont "get the message", so the battery BMS will have to disconnect the battery. 

So the question is, how often does that happen? It's hard to say. Maybe you'll be lucky, maybe you won't. 

Edit: I should have mentioned that there is a second, similar situation where one cell goes too low of a voltage, yet the overall battery voltage isn't under the limit to the inverter continues discharging. The BMS will likely disconnect in this situation also.

Edited December 7, 20232 yr by Coulomb

@Pie visit solar-assistant.io page for a full introduction and video off solarassitant capabilities. I only use SA to read info of the batteries because I also have external mppt .The control of charging is done by the extra mppt. Al info and settings of Axpert is done by SA.

@Coulomb Solar assistant does display the highest and lowest voltage off the cells of each battery.This bring some comfort that I will know if something goes wrong with the batteries. The external mppt I use does also give a warning and record if safe voltage is exceeded. 

On 2023/11/21 at 2:59 PM, Chris Louw said:

@Pie No problems to charge 3x Pylontech US 3000 batteries with 5 kW Axpert. Use Solarassistant to monitor the state of batteries and monitor and do settings of Inverter remotely. Battery tipe set to User. Float voltage set at 52.5v Using this setup for more than 4 years . No communication between Inverter and batteries. Can also use multiSIB software with Solarassistant cable to check battery .There is a SOP for Pylontech batteries with Axpert inverter. Reduced floatcharge from 53.2v to 52.5v for better battery life suggestion  of @Coulomb

Interesting note there Chris. Reducing float to 52.5v. I've never heard of that?

Perhaps @Coulomb and you can elaborate a  bit more?

Thanks mates!

 

On 2023/11/24 at 11:19 AM, Coulomb said:

There is one scenario where you lose out without the cable. That's when one cell goes to too high a voltage, and the only way to protect it is to reduce the charge current. Without the cable, the inverter wont "get the message", so the battery BMS will have to disconnect the battery. 

So the question is, how often does that happen? It's hard to say. Maybe you'll be lucky, maybe you won't. 

This is good information. 
In the context of my system will i get this benefit with a cable from the hubble battery to the inverter or even via the cloud link or solar assistant?

@Chris Louw checking out that solar assistant link now. 

Sorry for radio silence last week  partner was traveling so i looked after the kid alone. No time for solar research:(

Regars Guys

 

 

 

 

 

 

10 hours ago, Pie said:

In the context of my system will i get this benefit with a cable from the hubble battery to the inverter or even via the cloud link or solar assistant?

I don't know those monitoring systems, but my guess is yes, you would; that's the whole point of talking to the BMS.

On 2023/11/27 at 5:01 PM, Sass said:

Reducing float to 52.5v. I've never heard of that?

Perhaps @Coulomb and you can elaborate a  bit more?

It's very simple: even LFP batteries degrade less with a lower battery voltage. You do need the bulk voltage to get the battery charged properly, but once it's charged, you don't need to maintain such a high battery voltage. On my system, I find that I lose half a percent of SoC by using a float voltage that is even lower per cell than that, but I expect to extend my battery life by doing this.

Some people say why does an LFP battery even need a float setting? You do need a voltage set-point to aim for when balancing load power and charge power (whether that's utility or solar or other). In effect, the battery voltage falling below the float voltage is the signal to the charging system that the battery can accept more charge; when above this voltage, no more charge is desirable.

Of course, if you have BMS communication, then the inverter ignores these voltage settings, and responds to messages from the BMS to charge more, charge less, or stop charging altogether.

@Chris Louw

 

this solar assistant looked great. Definitely a great way to go. Quite excited, will test this woth existing pie. But seems like it allows me to monitor whats happening woth battery and change inverter charge setting to suit. 
 

seems like the way to go. 
 

so going ahead with hubble am5 and this. 
 

 

Update

Picked up the Hubble AM5 on 18 December 23. 
 

Its huge and heavy. Took my time to do the install as i had very limited space in the installation area. Lots of measurements and a CAD sketch but it went in like a glove. 

Setup the axpert as per hubble settings for this inverter. Bulk and float charge voltages of 55.2V, so a good bit higher than your recomendation @Coulomb

So far its been good, probably have not run more than 200W for longer than an hour yet, but it seems to be performing. Charges are really quick at the recommended 30A charging.

So now to monitor and maintain. 
I guess ill need to reduce the float voltage at some point and drain the battery down so i can recharge from SOC 20% back up to 80%. 
 

Hubble recomended cutoff voltage is 48V. This actually suits the inverter better than the AGM batteries I had. On this inverter you cannot set cutoff voltage higher than 48V, which is way too late for a lead acid battery in my view. 
 

Anyway thanks for the help and support guys. Will continue posting as i learn more.

17 hours ago, Pie said:

Update

Picked up the Hubble AM5 on 18 December 23. 
 

Its huge and heavy. Took my time to do the install as i had very limited space in the installation area. Lots of measurements and a CAD sketch but it went in like a glove. 

Setup the axpert as per hubble settings for this inverter. Bulk and float charge voltages of 55.2V, so a good bit higher than your recomendation @Coulomb

So far its been good, probably have not run more than 200W for longer than an hour yet, but it seems to be performing. Charges are really quick at the recommended 30A charging.

So now to monitor and maintain. 
I guess ill need to reduce the float voltage at some point and drain the battery down so i can recharge from SOC 20% back up to 80%. 
 

Hubble recomended cutoff voltage is 48V. This actually suits the inverter better than the AGM batteries I had. On this inverter you cannot set cutoff voltage higher than 48V, which is way too late for a lead acid battery in my view. 
 

Anyway thanks for the help and support guys. Will continue posting as i learn more.

48V cut off for a lead acid battery is quite a bit above 50% SOC as it would be measured under load and SOC is normally measured under no load. The no load level could even be around 49V.

Correct but i was trying to nurse my batteries and esnure they never went below 65% soc as there is a big drop off in battery when you get to 48v.  
 

Because I ran very small loads (again in an attempt to nurse the batteries) when the inverter did trip, the no load voltage was never more than 1v above 48V. 

It would have been nice to have the ability to set cut-out voltage to 49v or 50v and i think a maximum setting value for me was a little too low. 

1 hour ago, Pie said:

Correct but i was trying to nurse my batteries and esnure they never went below 65% soc as there is a big drop off in battery when you get to 48v.  
 

Because I ran very small loads (again in an attempt to nurse the batteries) when the inverter did trip, the no load voltage was never more than 1v above 48V. 

It would have been nice to have the ability to set cut-out voltage to 49v or 50v and i think a maximum setting value for me was a little too low. 

If one is at home one could also act on the low level alarm and switch off before the cut out level. 

I am using a light that comes on wired via the normally open contact to warn of a battery getting lower. It does also come on when using a high power item in the kitchen. This is also a good sign that one has LS and to limit the use of high power items. 

I hope some one can give me some advice.  I have an old Axpert 5kw inverter.  I have attached new lifepo4 batteries.  These are DC solar energy DC-L5000/LV.  I have given new parameters to the inverter.  Type USE, absorb 56.6V, float 55.4V,cut off 47V,  max charge current 40 amps.  These settings were provided by the supplier who is now not responding to questions.

This seems to work ok but I'm uncertain about the following issues.

1. When the batteries show soc of 100% they still take 30 amps with virtually no load.  2. The charge voltage never reaches 56.6V, only about 54V.

I know that since there is no data link between batteries and inverter some off the readings may be inaccurate.  I am using the ones displayed on the battery led display.  Any ideas ?