Axpert King/Pylontech/Watchpower

[ShaneDS]

Hi guys, 

Quick question.....

I have a 3.5 pylontech connected to axpert king inverter using watchpower app for android to monitor for now. When the power goes I immediately get battery capacity ast 65% on app even with no load. Could this be a setting issue or an issue with the app? Batteries are fully charged at the time. 

Thank

Shane

[Coulomb]

1 hour ago, ShaneDS said:

I have a 3.5 pylontech connected to axpert king inverter using watchpower app for android to monitor for now. When the power goes I immediately get battery capacity ast 65% on app even with no load.

You probably don't have the special cable for connecting the BMS of the PylonTech to the RS-485 port of your King. It has to be a recent model (made in the last 12 months or so, preferably later than that). Then you need to change the Battery Type setting (setting 05) to PyL.

Without the above, the SOC reading is for entertainment value only.

[ShaneDS]

15 minutes ago, Coulomb said:

You probably don't have the special cable for connecting the BMS of the PylonTech to the RS-485 port of your King. It has to be a recent model (made in the last 12 months or so, preferably later than that). Then you need to change the Battery Type setting (setting 05) to PyL.

Without the above, the SOC reading is for entertainment value only.

It's brand new, installed a week ago. Cable is installed. I'll check the battery setting and confirm it's at pylon. 

Thanks

[Paul Vermeulen]

I have a similar problem. I have replaced 32 x 6V US2200 Lead Acid Batteries with 2 x 3.5kW PylonTech batteries (2 x 5kVA Axpert Inverters 6000W of Solar Panels), and had the same problem with the voltage drop as soon as you start loading the batteries. So I got the ICC Software with  a Raspberry Pi and all the correct cables. Now my ICC Software shows the battery at 94% capacity with 19 hours of use to 20% DOC but the Axpert shows 19% capacity. I thought the ICC Software would control the inverters. Am I doing something really silly?

 

[Coulomb]

7 hours ago, Paul Vermeulen said:

I thought the ICC Software would control the inverters.

There are limits to what ICC can do to control the inverters. It can't change the SOC guesstimate, for example.

If you don't have a recent Axpert King or Axpert VM III with the special cable to the PylonTech BMS, and battery type (setting 05) set to PyL, then if you want accurate SOC measuring, you'll need a Victron BMV. It will have a good SOC estimate. I'm not sure if ICC can read this estimate (I think it can), or if you have to read it from the BMV display. You then ignore the SOC reported by the inverter's LC Display.

[Paul Vermeulen]

Thank you for the information. I was looking at upgrading the firmware but I see this only works with the MKS Inverters, I was "tricked" into buying the KS ones. Can I change the settings on the inverters to compensate for the inaccurate voltage sensing by the inverters? I have noticed that by dropping the cut-off voltage from 47.5V to 45V the inverters do not send out the 04 alarm. But they still show the battery to be flat.

[Paul Vermeulen]

To continue my comments I see that the Dyness 3.5kW Lithium Ion battery has the discharge cut-off voltage at 40.5V the charging cut-off voltage at 54V and nominal voltage 48V. Will these settings damage the PylonTech batteries? I was led to believe that the built in Battery Management System will protect the batteries in most circumstances.

 

[Coulomb]

2 hours ago, Paul Vermeulen said:

Thank you for the information. I was looking at upgrading the firmware but I see this only works with the MKS Inverters, I was "tricked" into buying the KS ones.

It's certainly possible to update the firmware on KS (PWM) models. It's just that they're less common, so there are fewer firmware updates for those models.

2 hours ago, Paul Vermeulen said:

Can I change the settings on the inverters to compensate for the inaccurate voltage sensing by the inverters?

The voltage sensing is usually pretty reasonable, though at times they could do with adjustment, and there are special commands for that purpose. The problem is that voltage, even measured perfectly accurately, is a very poor estimate of state of charge, for the LFP chemistry in particular. No settings adjustments will compensate for the fact that Axperts don't have a battery current sensor, and so can't accurately estimate battery SOC.

2 hours ago, Paul Vermeulen said:

I have noticed that by dropping the cut-off voltage from 47.5V to 45V the inverters do not send out the 04 alarm. But they still show the battery to be flat.

45 V is awfully close to 44.5 V, where Pylontechs are said to disconnect via their BMS (see this post). If one or more cells is weaker than the others, that voltage could increase.

 

 

[Coulomb]

1 hour ago, Paul Vermeulen said:

To continue my comments I see that the Dyness 3.5kW Lithium Ion battery has the discharge cut-off voltage at 40.5V the charging cut-off voltage at 54V and nominal voltage 48V. Will these settings damage the PylonTech batteries?

Using 40.5 for the low DC cutoff voltage setting (setting 29 on Axpert inverters) probably won't harm the battery too much (though it encourages deep discharges, which are not good for battery life). It could damage the inverter, as it may go crazy when the battery disconnects while the inverter is powering a load. Using a bulk/absorb voltage setting (setting 26) of 54 V will again prompt the BMS to protect the cells, with possible damage to the inverter.

Quote

I was led to believe that the built in Battery Management System will protect the batteries in most circumstances.

In an ideal world, the inverter and BMS would co-operate to charge the battery. This is what happens with battery type PyL and the appropriate cable and firmware, with inverters (Kings and VM IIIs) that have the PyL option. Absent that, BMS intervention is really an emergency brake; good to have it for safety, but not to be used daily.

[Coulomb]

10 minutes ago, Centurionsolar said:

So, what we do in ICC is we use the cable we make to rather speak directly to the BMS, and ask it what the SOC is.

Are you saying that ICC can talk directly to the BMS, even in systems without a battery type of PyL? But this communication is used only for SOC display?

That would seem to mean that you'd never need a Victron BMV if you're using a system with ICC and a PylonTech battery.

Is PylonTech the only battery so supported?

Do other battery brands emulate the PylonTech protocol sufficiently to achieve the same end?

[Paul Vermeulen]

Thank you for the information, I have a lot to learn about Lithium Ion, I believe that I have made the correct decision going that route. I also assume I have all the cables, one from the Battery Supplier, it looks like a data cable and is marked Inverter on one side (plugs into the RS232 port) and Battery (plugged into the CAN port) on the other side. I have one that came with the Raspberry Pi that looks like a phone jack on the one side and plugs into the Console port on the Battery, and the USB3.0 port on the Raspberry Pi, and the Raspberry Pi is connected directly to the USB port on one of the Inverters. I can see that I need to be able to store more electricity, my batteries are 100% full most of the day, with the house running directly off solar. I will double up next month if the batteries are available.

[0 |>\/\/3|<|<3Я]

5 hours ago, Centurionsolar said:

Hi Paul, you are correct yes, the Pylontechs (as well as most other batteries with BMS's built in) will auto-protect themselves when they go past a certain SOC/Voltage. Pylontech for instance cuts out at 9%, and then requires you to manually press the red button again to get them going.  Then regarding the inverter that reads the battery:  It's a bit dof of Voltronics to make a port for the inverter with software running on it to read to the voltage of the battery, as voltage is too easily manipulated.  Let me explain:  Let's say you have a 50W light-bulb connected to your battery without it begin charged, and you monitor the pylontech through the inverter (Axpert king) that reads the Voltage directly from the battery.  That light will shine but because it's only 50W, the voltage will barely drop, as it will recover almost instantly.  So it will read 53.1V almost constantly, even though we are pulling 50W from it, so it will eventually just go flat and turn off after a long while, but the inverter will not know about it, as it will constantly measure only the 53.1 odd volts.

In contrast though, let's say you run a boiler (or some other huge load) for only 5 minutes on the same battery.  Because of the huge load, the voltage will drop suddenly to let's say 48V, but you are only running that huge load for a few minutes, so the battery isn't really flat, BUT because the inverter can only read Voltage to make it's decision between flat and not flat, it will think the battery is flat and then engage eskom...

So, what we do in ICC is we use the cable we make to rather speak directly to the BMS, and ask it what the SOC is. This essentially means that the BMS keeps track of the amount of amp-hours that have flow regardless of voltage, and then works out what the percentage SOC is that is left in the battery and then reports on that. So doing SOC based switching then allows you to run the battery correctl to the right SOC, which is a true reflection of the amount of energy used from the battery, rather than using voltage only as the inverter does.

I hope this makes some sense

 

Kind Regards

 

 

Have you tested the Pylontech BMS integration via the new Axpert King and VM III inverters to determine how ICC would compare? Apart from extended monitoring capabilities, would there be any further optimisation by using ICC, such as extended battery backup time without reducing battery lifetime?

[Paul Vermeulen]

I dis-connected the cable from the batteries to the inverter during load shedding a few hours ago. Immediately the battery capacity on the inverter was within 10% of the battery capacity on the ICC Software, we did not hear the inverters screaming with their 04 alarm. I believe it works very well. Thank you for the advice!

 

[0 |>\/\/3|<|<3Я]

3 hours ago, Centurionsolar said:

Yup yup,

 

The thing with the VM III and King is that it still only reads Voltage on the Pylontech directly, which is not a true indication of SOC at all, so by adding ICC to the mi you can accurately measure the SOC of the battery, giving you a much clearer picture of what's going on with them and ultimately extending their lifetime exponentially as you will be able to turn off devices pulling the big loads, etc before the battery gets too flat.

Hope it makes a bit of sense?

 

Thanks, I'm now getting seriously interested in this ICC solution. Will it work with a mixture of Pylontech 2000 and 3000 batteries in the same parallel stack? I started off with 2x US2000 batteries and will now be adding 1x US3000 battery, with more 3000's as required in the future. Pylontech supports mixing these up, per the attachment. Will ICC monitor and manage a mixed stack as effectively?

US2000B US2000 PLUS US3000 Difference and Mixture Installation.pdf

[0 |>\/\/3|<|<3Я]

On 2020/02/24 at 5:23 AM, Centurionsolar said:

Indeed it will yes, the only requirement really is that the US3000 be the master battery, other than that the protocol is virtually unaware of the battery it's talking to, as the internal comms on the different batteries will know when to switch of which one but keep the rest of the stack going, so it's all good from that perspective too

 

Having added the 1x Pylontech US3000 to my 2x Pylontech US2000 stack, I now receive an error from the inverter, fault #69. Did not receive this error with only the 2x US2000 batteries.

According to the inverter display, the batteries now charge to 53.2V, then the error message is displayed whilst the voltage drops to 52.2V, then the error disappears whilst the batteries charge back up to 53.2V and this process keeps repeating.

Will the ICC resolve this problem for me?

[Youda]

Hi @0 |>\/\/3|<|<3Я

E69 means that the Pylontech raised "stop charging" flag and therefore Axpert stopped the charger. The flag is being raised basically on the overvoltage of pack or a cell in the pack.

Reason behind this might be that the two older bricks reach their top voltage too early. Shame that the implementation of BMS protocol in the Axpert is just basic, without the possibility to manualy adjust (lower) the C.C. voltage. Victron users were observing a very same error in the past, but since Victron gear is much more configurable, they were able to solve it easily.

With the ICC, you will disconnect the Pylon BMS from the Axpert and plug it to the ICC. Then, you will be able to set battery type to "USE" and lower the C.C. to 52.5V which will work much better. ICC will take care of reporting correct SOC and controlling the inverter.

I would wait for couple of days, whether the bricks will become balanced, which would remove the error. If not, I would go for ICC (not to mention that ICC has a LOT of other benefits too).

[0 |>\/\/3|<|<3Я]

10 minutes ago, Youda said:

I would wait for couple of days, whether the bricks will become balanced, which would remove the error. If not, I would go for ICC (not to mention that ICC has a LOT of other benefits too).

Thank you @Youda. I am leaning towards the ICC solution also for the ability to monitor remotely. Plus once funds allow I wish to add solar and from what I understand, the ICC solution will be able to help me with tracking power consumption for actionable insights and to accurately determine savings and eventually ROI.

Since the ICC solution is not cheap, are there any alternatives you would also recommend for comparison before committing to the purchase?

[0 |>\/\/3|<|<3Я]

2 minutes ago, Centurionsolar said:

HI @0 |>\/\/3|<|<3Я, there are  few alternatives yes, some free, but they lack things like historic information (making sure your system works over time), pylontech integration, and a host of other things, and albeit a bit of a biased view, I personally think that the extra stuff added makes it worth the money in my mind ;).

 

 

I think I'm convinced now Plus it's also exciting to support a local business for software - doesn't happen often. Am I able to order with custom cable length? I would require 1.5m between the Pi and battery as well as 1.5m between the Pi and inverter. Longer would be fine of course, just not shorter since my OCD requires me to run these cables in the trunking already in place  

[0 |>\/\/3|<|<3Я]

5 minutes ago, Centurionsolar said:

mail to [email protected]

Sent, we continue the chat there

[Oros]

On 2020/02/22 at 7:06 AM, Centurionsolar said:

Hi Coulomb,

Correct yes, if you use the cable we make, you can essentially remove the cable between the battery and the inverter, change the inverter settings to USE rather than PYL (it doesn't REALLY matter but I prefer it that way so I can set the cutoff lower to prevent error 04), and then simply use ICC to interrogate the SOC directly from the BMS inside the battery and do switching to grid based on that.  I have found that doing it this way is even more accurate than using a Victron BMV ;).

At the moment we only have Pylontech and Narada lithiums BMS protocol adopted, with others being added as we get sample batteries to test together with their respective protocol documentation.

 

Hi, I have Narada lithium batteries and have rasberry pi with ICC software installed, but I do not have the cable connecting the naradas to the pi. I don't even have the data cable connecting the narada's to the inverter. Can I buy just the cable for the pi connection to the batteries?

Is there any software I can use to see more info from the Narada's bms, or is it basically just the soc? 

[hpb23]

Hi folks,

I have an MPP Solar PIP-5048MK = Axpert King 5K with 3 Pylontech US3000 running (in Germany).

Your inverter must support Pylontech Batteries (which is only true for these manufactored after feb 2019), which can be checked through prog 05. It must have an entry "pyl".

Set prog 05 to pyl and connect inverter and batteries with custom RJ45 cable.

Cable PINOUT:

1 - 3, 2 -5, 3 -1, 5 -2

Note: This cable is symmetrical - there is no dedicated end for batteries or inverter.

Pylon Batt Setup: Set communication speed to 9600 Baud (Dip-1 ="On") and connect cable to RS 232 port.

Set Inverter prog 05 to "pyl" , connect to RS232 port and restart.

When communication inverter - batt is ok, the batt symbol on inverter display will blink.

This setup works with Inverter firmware in my setup - cable pinout was provided by support of MPP Solar:

U1=71.80, U2=02.00, Bluetooth=00.21, SCC Version U4=1.10

Actually in australia - greeting from Brisbane.