For reasons I have discussed elsewhere on this forum, I need to make some configuration changes to my Goodwe ES. Please note

• The physical batteries will remain the same
• The BMS will remain the same
• The battery type will change (from BYD to self-define)
• Once I have changed to self-define I need to change some settings around charge and discharge current to bring them in line with the battery maker's specs
• I will be disconnecting the comms cable between the inverter and the BMS. This is a recommendation from the battery manufacturer. I will disconnect this at both ends because I suspect that the inverter (at least) senses the physical presence of the cable and starts wanting certain inputs from it.

I am going to do this during the day, when there is plenty of PV available.

OK... this seems not too scary.

As I see it, I have to key in all the settings and then hit save, the inverter will ask if I want it to restart, and I will respond with a "yes". Then I disconnect the comms cable.

Question: Am I doing this in the right sequence? I know that right now, running in BYD mode, if I disconnect the comms cable, the system will shut down the batteries pretty quickly.

Edited February 9, 20223 yr by Bobster

That is pretty poor show from Revov. Running lithium in self-define is not optimal. Have you tried setting the inverter to Default lithium 100Ah? Or even pylontech?

The way you want to do it should work, but if you are worried, why not do it when it is dark and the batteries are empty?

52 minutes ago, P1000 said:

That is pretty poor show from Revov. Running lithium in self-define is not optimal. Have you tried setting the inverter to Default lithium 100Ah? Or even pylontech?

The way you want to do it should work, but if you are worried, why not do it when it is dark and the batteries are empty?

The Goodwe comes pre-configured for lots of brands of battery, so you can select BYD or Pylontech or LG... but for Revov there is nothing. I don't think this is Revov's fault.

This recommendation came about because I asked them 

1) If the Goodwe is on their list of approved inverters?
2) If so, then they do they have another site reliably running the same batteries and same inverter?
3) If so, then can they determine what settings are on that Goodwe inverter so that they can be replicated at my site?

I want to do it during the day because experience tells me that the Goodwe needs PV to get the batteries working again.

 

28 minutes ago, Bobster said:

I want to do it during the day because experience tells me that the Goodwe needs PV to get the batteries working again.

 

That is definitely not true.

2 minutes ago, P1000 said:

That is definitely not true.

Much earlier on in the life of this system, the batteries used to turn off for no reason that anybody could ever figure out. This was intermittent. There is a feature on the Goodwe that sends a pulse of DC to the batteries to wake them up. But this only happens when there is PV. Or I could only get it to happen when there was PV.

I would wake up in the morning. Batteries not working. I'd reset everything and make sure this "kick start" setting was on. Nothing would happen, but this was early in the morning. By the time I got to work the sun would have come up, PV was working, and the batteries came back on line.

10 hours ago, P1000 said:

That is pretty poor show from Revov. Running lithium in self-define is not optimal. Have you tried setting the inverter to Default lithium 100Ah? Or even pylontech?

The way you want to do it should work, but if you are worried, why not do it when it is dark and the batteries are empty?

@P1000Thanks for your feedback thus far.

A partial history of this system is
1) Currently it is running as BYD B-BOX 10 with comms between BMU and inverter. This is a modified cable - only the middle 2 conductors are left intact and the cable is not crossed.
2) Revov say this results in too high a charge current and voltage. This could damage the cells eventually.
3) The system has been tried in default lithium 100a/h mode. This resulted in an even higher charge current and voltage which could not be changed.
4) Then the system was put into self-define mode, using Revov's settings and with the comms cable still in place. The inverter complained of "battery communication failure".
5) It was reset back to BYD B-BOX 10.
6) Revov found another site with same batteries and inverter running reliably. That site has no comms cable, which is why I am proposing to disconnect the comms cable when I change the setting. They say the inverter is running in self-define mode.

(6) is how my system was originally set up, but the batteries would switch off from time to time, even if they had charge. However, the inverter firmware has been updated since the change to (1).

Why do you say that self-define is not optimal? Is that intended for lead-acid batteries?

Edited February 10, 20223 yr by Bobster

13 hours ago, Bobster said:

@P1000Thanks for your feedback thus far.

A partial history of this system is
1) Currently it is running as BYD B-BOX 10 with comms between BMU and inverter. This is a modified cable - only the middle 2 conductors are left intact and the cable is not crossed.
2) Revov say this results in too high a charge current and voltage. This could damage the cells eventually.
 

The resultant current and voltage are those requested by the BMS in those modes - that should be configured in the BMS commissioning tool. So I find it strange that the manufacturer should suggest those are wrong?

 

13 hours ago, Bobster said:

Why do you say that self-define is not optimal? Is that intended for lead-acid batteries?

It is not optimal, because the inverter then ignores the BMS, so it estimates all the values itself. Taking the above in consideration, that might not be the worst... The problem is it will then estimate SoC based on it's own coulomb counting - that is not so bad. It will ignore voltage and current requested from the BMS - that apparently solves the problem, because the BMS requests values that are too high, according to Revov.  The biggest issue is that the inverter does not know the balance state of the cells, so it cannot lower current to balance the cells, and also cannot lower the discharge rate from the battery. So it will go full steam until the BMS decides one cell is too low and shut down the battery.

None of the modes support lead acid with firmware from the last 2? or so years.

OK... this is interesting. So the inverter doesn't say to the batteries "I'm going to charge you at this rate", but rather the BMS asks the inverter "please charge the batteries at this rate"? Even if there is a discrepancy between the two sets of values?

But then why have all the different options?

On 2022/02/11 at 1:24 PM, Bobster said:

OK... this is interesting. So the inverter doesn't say to the batteries "I'm going to charge you at this rate", but rather the BMS asks the inverter "please charge the batteries at this rate"? Even if there is a discrepancy between the two sets of values?

But then why have all the different options?

That is exactly how it works. The BMS tells the inverter what rate they want to be charged at. Generally to lower the rate the inverter can do. I see this all the time on mine.

The different options generally are that batteries have different protocols or sets of commands they work with.

So Revov's concerns about the charge current being set too high on the inverter is a red herring. Their BMS will draw what it needs.

Edited February 15, 20223 yr by Bobster

11 minutes ago, Bobster said:

So Revov's concerns about the charge current being set too high on the inverter is a red herring. Their BMS will draw what it needs.

Exactly.

And this is is what I see now. Revov want 55.4 charge voltage. System says 56.5. But the Goodwe app shows that the batteries are charging at 55.4. So the BMS's needs trump the inverter settings.

Sounds like it is working?

If the BMS comms work, it will override the settings set on the inverter.

In home assistant for example, I can see the battery limits set by the battery. You _should_ be able to see this in PV Master as well.

In home assistant I made it look like this. 

Thanks all for your help. I have been able to establish that the BMS is controlling the inverter. In fact the inverter settings don't come into play and I am guessing the installer set the system to BYD because the BMS is made by BYD. The inverter says it's going to charge at 56.5 volts. Revov want 55.5, and that is what they get because the BMS calls the shots.
It would probably be a good idea to install that software that somebody (@Cef? ) wrote and generously put into the public domain. The installer wanted voltages at specific times. The first was fairly easy - just before midday when the battery usually hits 100% SOC. The other was early morning when SOC is lowest. So I had to sit up from 5 this morning and watch PV master and log voltages. Not the end of the world, but there are other, more accurate ways of doing this. I am accustomed to contemplating the meaning of life at that hour of the morning.

14 hours ago, Bobster said:

It would probably be a good idea to install that software that somebody ( ) wrote and generously put into the public domain.

I would recommend rather going with homeassistant and the goodwe plugin (on something like a raspberry pi). It opens up a lot more possibilities.

I agree, with Home Assistant you can log everything and have it graphed in Grafana. Then you can pretty much give them all the info they want.

@Bobster Hi !

If you find another software solution that can provide you with 1-second resolution data and keep record for 72 continuous hours (approximately 259,000 records) of all electrical parameters provided by GoodWe ES series sensors, without relying on internet, please let me to know...

14 hours ago, Cef said:

@Bobster Hi !

If you find another software solution that can provide you with 1-second resolution data and keep record for 72 continuous hours (approximately 259,000 records) of all electrical parameters provided by GoodWe ES series sensors, without relying on internet, please let me to know...

The Home Assistant plugin can do 1 second if that's what you want (default is 5s). And for as many hours as you have storage for...

1 hour ago, P1000 said:

The Home Assistant plugin can do 1 second if that's what you want (default is 5s). And for as many hours as you have storage for...

There are guys using the home assistant plugin with sub second intervals. Don't ask me how big their databases are but they crazy in my opinion.

Hi @P1000 !

I would like to know first of all if you are a developer or participated in the open source project you use.
Are we talking about hardware next to your inverter which is a Raspberry PI? Maybe it's a bit outdated on this platform, but at that refresh rate, does it write locally to an SD memory card to a database? I don't think an SD will have much life with a 1 second read/write polling rate. Another thing is that your data travel to some remote server on the Web... This implies hosting costs, which is unlimited? Of course, it depends on what you want to pay.
On a LOCAL hard drive you store everything you need. The database is yours and it is on your PC.
Several installers have told me that they have the software saved on a laptop and all they have to do is connect to the home LAN where they finished installing the Solar System. They immediately know a lot of data about how your newly installed system is working, not to mention that in the case of isolated houses or without Internet service, the solution you are using is inapplicable or it is also inapplicable for people who have no idea how to configure a Raspberry PI. . . and/or installation scripts.

@Tinuva: At least with the ES Series, up to 0.5 second intervals are possible, but the % CRC Errors increase to high values, becoming very dependent on the stability of WIFI connectivity.

While hourly average historical data is unlimited in the tables I use in separate databases for continuous storage, I store live data separately for 72 hours per second (I could it do 30 days for example, which would equate to 2,592,000 records), as it rarely takes longer to diagnose your system in the event of an inverter trip, and in case you want to download that data into a spreadsheet, it doesn't take forever to convert it to .csv format.
I invite you both to try my software and after trying it, compare it. I accept all kinds of constructive, respectful criticism and suggestions. All that is needed is a Windows PC. I'd appreciate your feedback and it would be great if you could tell me what the software you're using does compared to the one I'm offering for free, so I can add what might be missing and eventually top it off.
Regards.

14 hours ago, Cef said:

Hi @P1000 !

I would like to know first of all if you are a developer or participated in the open source project you use.
Are we talking about hardware next to your inverter which is a Raspberry PI? Maybe it's a bit outdated on this platform, but at that refresh rate, does it write locally to an SD memory card to a database? I don't think an SD will have much life with a 1 second read/write polling rate. Another thing is that your data travel to some remote server on the Web... This implies hosting costs, which is unlimited? Of course, it depends on what you want to pay.
On a LOCAL hard drive you store everything you need. The database is yours and it is on your PC.
Several installers have told me that they have the software saved on a laptop and all they have to do is connect to the home LAN where they finished installing the Solar System. They immediately know a lot of data about how your newly installed system is working, not to mention that in the case of isolated houses or without Internet service, the solution you are using is inapplicable or it is also inapplicable for people who have no idea how to configure a Raspberry PI. . . and/or installation scripts.

Hi Cef, this is not a fight. I agree with most of what you have said, I just wanted to point out your erroneous assumption in the previous comment. Home Assistant and your software serve different purposes. I merely suggested that Home Assistant might provide more options - like controlling air conditioning based on solar production and SoC of the batteries.

I am thankful for your contribution and it definitely provides a lot of value to many people. Keep up the good work!

On 2022/02/19 at 4:47 AM, P1000 said:

Hi Cef, this is not a fight. I agree with most of what you have said, I just wanted to point out your erroneous assumption in the previous comment. Home Assistant and your software serve different purposes. I merely suggested that Home Assistant might provide more options - like controlling air conditioning based on solar production and SoC of the batteries.

I am thankful for your contribution and it definitely provides a lot of value to many people. Keep up the good work!

Hi P1000 ! Of course it isn't a fighth ! 😊👌
I think the point is that when you tell another Forum User that one solution is better than another for that other member's problem or situation, it's good to "give an objective opinion" and objectivity is achieved when I make a comparison, in summary: "I know because I've tried both solutions..."
I am not referring to the control of Relays by events, since it is the subject of a future version of my software (v5.0). And precisely the point is that I designed it to find and discover all kinds of possible problems related to the solar installation and not remain attentive all the time to know how a particular electrical variable behaves, waiting for the eternal, and I assure you in many cases erroneous, data delivered by the Sems Portal of Goodwe.
Thank you for your feedback and encouragement to continue sharing my development !

Hi @Cef

I have to admit, I haven't tried your software. Mostly because I do not own a windows pc. Only macs and then mini/tiny linux servers. Unfortunately I can't justify keeping a windows pc on hand for the app.

 

On 2022/02/14 at 12:46 PM, Tinuva said:

That is exactly how it works. The BMS tells the inverter what rate they want to be charged at. Generally to lower the rate the inverter can do. I see this all the time on mine.

The different options generally are that batteries have different protocols or sets of commands they work with.

So to return to this after a long pause: How does the BMS "know" what the batteries want? I presume there is a battery specific graph in the BMS's firmware; the BMS then can use this to correlate voltage to SOC and then make a decision as to what is required. Am I barking up the right tree?

This is curiosity only. I am not about to start running cables between devices. Any reports of explosions in Jhb will be nothing to do with me.

On 2022/04/19 at 10:47 AM, Bobster said:

So to return to this after a long pause: How does the BMS "know" what the batteries want? I presume there is a battery specific graph in the BMS's firmware; the BMS then can use this to correlate voltage to SOC and then make a decision as to what is required. Am I barking up the right tree?

This is curiosity only. I am not about to start running cables between devices. Any reports of explosions in Jhb will be nothing to do with me.

BMS monitor the voltage over the whole pack and then also have a tap wire to each cell individually. So it can monitor the cells for specific scenerios.

As for knowing what the batteries want, I guess the correct wording is, the bms is programmed for what is best for the batteries and then it tell the inverter what the limits are at any given stage.

For a deep dive, you can read a "How a BMS works" page.

I like this reddit wiki: https://old.reddit.com/r/AskElectronics/wiki/batteries#wiki_li-ion_bms

https://www.orionbms.com/general/how-it-works/

https://www.synopsys.com/glossary/what-is-a-battery-management-system.html

 

Edited April 20, 20222 yr by Tinuva