EG4 Chargeverter Battery/Inverter Communications

[wpns]

You had current and voltage reported wrong. "90 amps (VOLTAGE) at 54 volts (CURRENT)" should read "54 volts (VOLTAGE) at 90 amps (CURRENT)"

You were reporting the inverter's battery voltage readings, not the inverter's battery SOC readings, which led me to believe that you were in open-loop (voltage reading) 'communications' between the inverter and the batteries.

If your inverter is in closed-loop communications with the batteries, I wouldn't (in fact I don't) worry about what the voltages are. Use the SOC numbers from the inverter, the inverter will talk to the batteries and determine what they need for a voltage and charge them to whatever voltage makes the batteries happy (this will change as charging progresses).

If the CV is charging the batteries based on voltage it may do something different, so you'll probably want to keep the CV voltage at 55.0V or something so as not to overcharge the batteries. I keep mine at 52V, which is sufficient to prop up the SOC to about 25% even though the CVs read 0.2 volts apart, and contribute different currents to the battery bank. I'm away for hurricane season now, so I can't make any changes, but when I get back I'll probably set the CV voltages to 55.0V and use the Gen Dry Contacts to enable the relays that power the CV AC inputs. Then I can set the 'generator' to come on at 20% and 'stop' at 40% or something. I'm working on a better control algorithm that'll probably run on the Raspberry Pi that's inside the primary inverter, so I can control charging based on weather (fully charge before hurricanes arrive, don't use grid power to charge the batteries if it's only a short time till sunrise, etc. But that's an optimization for later...

For instance, my batteries were full at 11:25AM, and the voltages, as reported by the inverters (Red and Green) and batteries (Blue and Cyan), wander all over the place. But again, voltage is a _really_ bad way to tell SOC from LFP batteries, so I don't worry about it, I figure the batteries are full and balancing themselves, so fully charged is fully charged, and I'm good. Once I get the CVs under control I'll probably set them for 55V so they won't overcharge the batteries, and let the inverter take care of the 100%SOC/balancing regime.

 

[Officient]

I read in the manual that there has to be a 5 volt difference between battery start and battery stop. I have mine at 5.2 volts difference and it still maintains a 90% charge. i will be making a relay soon between grid source and grid plug.

This is valid point, I only have 2 volt difference in my setup right now. It seems to not impact the Chargeverter though in my last two charging sessions which is odd. Shouldn't it not charge if it was adhering to the config settings of 54 volts BATT START 56 volts BATT STOP? I would think my last charging session started with my batteries well below 54 volts. It seems to me that without the communication cable or a two wire start generator connected to the CV, that the VC is ignoring the BATT START / BATT STOP settings, that's unless I'm overriding it somehow and don't know it.

Here's the Battery Voltage graph from yesterday, you can see the voltage was a little over 52 volts when I started. I charged up until I went to bed so the CV didn't have a chance to stop itself at 56 volts.

 

[wpns]

I have written software that emulates an EG4 Lifepower4 battery. The communications are "EG4" which is in a HexAscii format. I do expect the CVGC it is acting as a "Master" which is why you can't have an inverter and the CVGC communicating with the batteries are the same time.

I will simply connect the CVGC to my computer where I can change SOC and battery voltages being reported to the CVGC to hopefully see how the CVGC controls charging based on the software settings. I also hope I can figure out if the CVGC settings can also be changed by software

I am also building a software "Master BMS" which does allow me to merge both EG4 LifePower4 batteries and SunGoldPower batteries. The inverter thinks it's communicating with a Master BMS when in fact it's communicating with my software. The software takes the data from all the batteries and properly reports (in response to inverter commands) data back to the inverter, for example an average of all battery SOC's and total amp hour capacity of all batteries in the bank.

Really looking forward to your software, assuming you are going to make it available? At this point, all the 18Kpv can tell me about the batteries is a single SOC number, a single voltage, a BMS current that makes no sense, and the max/min cell voltage across all 6 of my batteries, which is a 250mv span, but it's impossible to tell how the cells within a particular battery are balanced.

I keep holding out hope that using a special cable you could talk RS-485 to the batteries from the CVGC and CAN from the 18Kpv, but I can't even find anyone at SS/EG4 who knows what I'm asking, so maybe I'll have to try it when I can take my system down and try it myself...

 

[wpns]

This is valid point, I only have 2 volt difference in my setup right now. It seems to not impact the Chargeverter though in my last two charging sessions which is odd. Shouldn't it not charge if it was adhering to the config settings of 54 volts BATT START 56 volts BATT STOP? I would think my last charging session started with my batteries well below 54 volts. It seems to me that without the communication cable or a two wire start generator connected to the CV, that the VC is ignoring the BATT START / BATT STOP settings, that's unless I'm overriding it somehow and don't know it.

Here's the Battery Voltage graph from yesterday, you can see the voltage was a little over 52 volts when I started. I charged up until I went to bed so the CV didn't have a chance to stop itself at 56 volts.

View attachment 228246

Where are you reading "vBat(V)" from?

 

[Officient]

You had current and voltage reported wrong. "90 amps (VOLTAGE) at 54 volts (CURRENT)" should read "54 volts (VOLTAGE) at 90 amps (CURRENT)"

Good catch, I thought you were refering to the BATT START / BATT STOP settings. I've corrected this to avoid confusion for others in my first post.

You were reporting the inverter's battery voltage readings, not the inverter's battery SOC readings, which led me to believe that you were in open-loop (voltage reading) 'communications' between the inverter and the batteries.

Gotcha! Yeah, the EG4 18k has both readings when using closed loop comms if I'm understanding correctly.

If your inverter is in closed-loop communications with the batteries, I wouldn't (in fact I don't) worry about what the voltages are. Use the SOC numbers from the inverter, the inverter will talk to the batteries and determine what they need for a voltage and charge them to whatever voltage makes the batteries happy (this will change as charging progresses).

Agreed, but in this case the inverter isn't doing any charging. The CV is connected directly to the batteries and I don't currently have grid/solar providing any other power source. Unless I'm mistaken, the inverter is in the dark about the CV and what it's doing to the batteries, so I'm looking to regulate the batteries based on voltage by reading the average battery voltage reported on the inverter.

If the CV is charging the batteries based on voltage it may do something different, so you'll probably want to keep the CV voltage at 55.0V or something so as not to overcharge the batteries. I keep mine at 52V, which is sufficient to prop up the SOC to about 25% even though the CVs read 0.2 volts apart, and contribute different currents to the battery bank. I'm away for hurricane season now, so I can't make any changes, but when I get back I'll probably set the CV voltages to 55.0V and use the Gen Dry Contacts to enable the relays that power the CV AC inputs. Then I can set the 'generator' to come on at 20% and 'stop' at 40% or something. I'm working on a better control algorithm that'll probably run on the Raspberry Pi that's inside the primary inverter, so I can control charging based on weather (fully charge before hurricanes arrive, don't use grid power to charge the batteries if it's only a short time till sunrise, etc. But that's an optimization for later...

What you're saying makes sense I think, in my scenario I'm trying to get the CV to charge the batteries from 20-30% back to 80-90% as I don't have any other charging source. Hope this makes sense.

 

[wpns]

in my scenario I'm trying to get the CV to charge the batteries from 20-30% back to 80-90% as I don't have any other charging source.

And you are charging from a generator so you can’t just leave it running to balance? I’d (also I did) use the inverter’s Gen dry contacts to enable the power to the CV but you could always put the CVGC in closed loop mode and leave the inverter in open-loop mode till you get panels.

 

[Officient]

And you are charging from a generator so you can’t just leave it running to balance? I’d (also I did) use the inverter’s Gen dry contacts to enable the power to the CV but you could always put the CVGC in closed loop mode and leave the inverter in open-loop mode till you get panels.

Yes, I can leave it running to balance assuming the CV does shutoff at the BATT STOP setting. In my testing so far, it just wasn't clear if this was happening b/c I saw the battery voltage over 56 volts when I was charging last time with my setup and the CV set with a BATT STOP of 56 volts. I charged to 100% on the batteries and the CV did ramp down, but I wasn't sure if it was the CV causing this or the battery BMS stopping any further charging.

In theory the CV should slowly ramp down as it approaches the BATT STOP setting correct?

This is my big question. Seems like most are discussing the CV as if it does, but in practice and in the manual, I don't see this as happening for certain.


I can do some specific testing I suppose to find out for certain by setting the BATT STOP to something like 55 volts, then let the generator charge to that via the CV and see if the CV ramps down as most seem to predict.

I wish my generator had dry contacts, can't find a portable generator that has that feature.

 

[guru_florida]

Will you let us know if the CVGC does ignore the user start stop configs when plugged into grid power?

I've had it working for 2 days now. Confirmed the BATT START|STOP settings are ignored on shore power. However, it is working as desired as the VOLTAGE setting is obeyed instead. So I set that to the float voltage of 54.2v. Now it charges to 54.2 (99%) and stays there and the CV current goes to near 0. What is interesting is the demand current by the UPS is actually being supplied mostly by the CV and the CV current will go up with demand even though the batteries are charged. My guess is as amps are being pulled, for ex. by the HVAC, the battery voltage may drop so the CV starts to kick in again. Since the CV has a slow ramp-up it's not immediate, the power will come from the batteries first then the CV will compensate and recharge that.

I have to say it works like a charm. I can set a max current on the shore power (amps setting in the CV) so it draws for example 25A max on a 30A shore hookup. I won't have to pay the up-charge for the 50A service. The batteries mean I can run pretty much every appliance, 2 HVAC, coffee maker and convection oven and not trip a breaker. Also many campgrounds have crappy power that can kill appliances so the CV works like a power filter. Sweet!

Only downside...the CV is loud when supplying power! It's currently in my bedroom, I definitely have to move that...plus the 6000XP though it's not as bad.

 

[guru_florida]

You had current and voltage reported wrong. "90 amps (VOLTAGE) at 54 volts (CURRENT)" should read "54 volts (VOLTAGE) at 90 amps (CURRENT)"

You were reporting the inverter's battery voltage readings, not the inverter's battery SOC readings, which led me to believe that you were in open-loop (voltage reading) 'communications' between the inverter and the batteries.

If your inverter is in closed-loop communications with the batteries, I wouldn't (in fact I don't) worry about what the voltages are. Use the SOC numbers from the inverter, the inverter will talk to the batteries and determine what they need for a voltage and charge them to whatever voltage makes the batteries happy (this will change as charging progresses).

If the CV is charging the batteries based on voltage it may do something different, so you'll probably want to keep the CV voltage at 55.0V or something so as not to overcharge the batteries. I keep mine at 52V, which is sufficient to prop up the SOC to about 25% even though the CVs read 0.2 volts apart, and contribute different currents to the battery bank. I'm away for hurricane season now, so I can't make any changes, but when I get back I'll probably set the CV voltages to 55.0V and use the Gen Dry Contacts to enable the relays that power the CV AC inputs. Then I can set the 'generator' to come on at 20% and 'stop' at 40% or something. I'm working on a better control algorithm that'll probably run on the Raspberry Pi that's inside the primary inverter, so I can control charging based on weather (fully charge before hurricanes arrive, don't use grid power to charge the batteries if it's only a short time till sunrise, etc. But that's an optimization for later...

I would suggest putting the Raspberry Pi outside the primary inverter case, kind of like where the wifi adapter is. You can double-sided tape it to the side for example and bring in only the cable. Keep low voltage devices like RPis away from high-voltage...or in other words...keep your Pi out of the oven.

 

[wpns]

I would suggest putting the Raspberry Pi outside the primary inverter case, kind of like where the wifi adapter is. You can double-sided tape it to the side for example and bring in only the cable. Keep low voltage devices like RPis away from high-voltage...or in other words...keep your Pi out of the oven.

The environment inside the case is actually better for the Pi, it's hot, humid, and dusty outside, and it's far enough from the AC and battery connections that there's no concerns about it getting zapped. I mean, there's low voltage and RS485 and stuff in the same space...

 

[wpns]

I've just published https://www.thingiverse.com/thing:6696579 which is a case for a (dual) relay from Amazon that'll be used to switch the relays that power the ChargeVerters on/off/inverter-control in case anyone's looking to do a similar thing. Python code a work in progress, but functional, DM for details.