The 6000xp firmware ccaa-191011 seems to have fixed the bug with open-loop charge control for LFP... mostly. The resulting charge curve is "correct" most of the time as far as I can tell.
But, the 12000xp is NOT correct
Notifications
Clear all
January 23, 2026 4:50 pm
February 28, 2026 2:45 am
@dtbaker61 you seem to have some insight regarding the lack of Absorb time the EG4 inverters are lacking. With that in mind, I have an EG4 18Kpv that I just found out from tech support doesn't allow a setting for Absorb time. Over the past couple of weeks, I returned home to see the inverter had my bank of 10 SimpliPHI 3.8 51.2 batteries (BMS but no comms) at Absorb 56.4v for 3-4 hours. At the time, my batteries were set to Lithium using the Lithium 31 setting I believe. I was told to change them to Lead acid which I have now done, but there is still no Absorb Time. So, I've read through this post and it seems you are saying I should set the Absorb to 56.2v (with Float at 54v) and the system will take care of the rest. Is that correct? Do I need to change the batteries back to Lithium or leave them at FLA? Or, do you know of a 3rd party arrangement that could communicate between the batteries and the inverter so an Absorb time could be implemented?
February 28, 2026 3:28 am
@mtjag Take a look at this project - it seems to do what you are are thinking. There is also another on github called yam bms that might be similar. I haven't had time to get them going. I did get the smart shunt working with this one one but then didn't get the canbus part of it working . But now I understand the canbus stuff and the different can adapters so might give it another try sometime. Actually I just ordered an espboard with integral canbus port , that is what the developer moved to using . https://github.com/sijones/DiyBatteryBMS
February 28, 2026 3:34 am
@mtjag this looks like it has support to work with eg4 https://github.com/Sleeper85/esphome-yambms/blob/main/documents/README/Supported_devices.md
March 1, 2026 3:09 am
@ouijifour thank you for sharing. I'll start reading to try to figure it out. Strangely, today after switching my settings back to FLA lead acid, my system hit 56.4v stay there at least 4 minutes but less than 8 minutes and then started the decent back to Float at 54v.
February 28, 2026 10:53 pm
March 1, 2026 12:44 am
@dtbaker61 you may want absorb time to give the bms time to balance the cells . They only will do that at the top of the charge cycle. It really depends on what the bms is programmed for, and what voltages you charge to. For example I charge to 55.2v which is 3.45 per cell. That is the voltage the balancing starts at . So depending on how matched the cells are it can take between 15 minutes to an hour to balance. Also depends if a battery has passive 150 ma bypass balancing or if like in my case it has active balancing which does the balancing if needed much faster. So I don't agree with your statement that lithium batteries don't need absorb time.
March 1, 2026 3:20 am
@dtbaker61 I definitely don't want 3 or 4 hours at that voltage to be sure. I've spoken with the SimpliPHI tech over the past 4 years with these batteries and they seem to think it's best to hit 56.4v for about 6 minutes then drop back to lower voltages for the remainder of the day. I still have 6 years of warranty on my bank (guaranteed 80% of capacity at 10 years) so I hope to exceed that by a lot. I see in your last post EG4 came through on a firmware update for charging on the 6000xp open loop charge control. How long did you have to lean on them to get that done? I hope they are more receptive to doing these kinds of requests. It just seems like going backwards when my old Outback inverter/CCs you were able to program to set points and time without issue.
March 2, 2026 4:34 pm
@mtjag exactly.... by setting vBulk to 56.4, or 57, or even 58.0 give the BMS 5-10 minutes to balance at low current, and then allows a drop to vFloat. For the 6000xp, the behavior now with firmware ccaa-1c1212 is actually cycling between vBulk and (vFloat-1v) for the afternoon while PV is available. This is "acceptable", although I would prefer vFloat +/- 0.5v
February 28, 2026 11:25 pm
looks like update to ccaa-1c1212 fixes this for the 6000xp. see details and notes: https://forum.eg4electronics.com/community/inverters/6000xp-bug-in-charge-controller-setpoint-voltages-with-open-loop/paged/5/#post-20821
February 28, 2026 11:20 pm
I'm following up on the best available fix for the 6000xp open loop charge control.... which is to do a firmware update to version ccaa-1c1212 which is the latest available as of 2/27/2026.
The update takes about 30 minutes, and requires a solid wifi connection. The update CAN be initiated remotely without any issues; it downloads, preps, uploads several pieces to itself, and then automatically reboots so everything takes effect. The reboot will turn off AC output to load for a minute, so be prepared for a power outage at the very end unless you have Grid AC input.
to initiate the update, all you have to do is login to the web portal at https://monitor.eg4electronics.com , and navigate to the screen Maintenance>RemoteUpdate screen, and click on "standard update" button . It takes about 30 minutes with several pauses in between pieces. DO NOT end the process until you see the final "success" in the history window.
THEN, you will have to review/change a couple settings:
- Battery Type has to be changed from lead-acid to Lithium
with sub-type = 31 for generic open loop 16-cell LFP
- Battery control should be changed back from SOC to Volt since SOC doesn't work if you are open loop with no comm cables, right ?!
- default vBulk ok at 57.4 , default vFloat ok at 54.0
- I changed system v limit from 59.5 to 58.0
- I set LV disconnect to 46v, it's a bad idea to go lower than that, and I noticed "disconnect resume" after LV disconnect is hard-coded to 3V, which is ok since by the time batteries get back to 49, they can support loads again
.... system appears to be functioning as desired and has settled to Float plus/minus while PV coming in is greater than
loads, BUT the system does not really float at vFloat as you might expect. I reviewed data history the next day, and it appears that vBatt will cycle between vFloat-1v, and vBulk. This is not a true float, but is acceptable as far as I am concerned since it is not holding the batteries at vBulk for any length of time. It is NOT desirable to hold LFP at an 'absorb' voltage, but is acceptable to cycle between 57.4 and 53.0 (which is effectively 100%-99%)
March 1, 2026 12:57 am
@dtbaker61 That cycling from float to bulk back to float is exactly why I quit using the eg4 xp6000 for charging lithium batteries. There are a lot better mppt chargers out there that don't do that kind of thing. I have no idea why the xp6000 would do that since it is unlike most any other charger which goes bulk to absorb , then back down to float and stays there until the next day. 55.2v is around 99% full. I understand that there are lots of different theories and thoughts about the best LiFePo4 charging practices. I do the approach to stay away from the top voltage to stress the cells less and because it isn't worth it to try and go way to the recommended limit to gain just 1% of capacity. Off Grid Garage on youtube has done extensive testing of charing voltages and load capacities with the results all graphed out and explained so that is where I get my numbers from .
March 2, 2026 8:43 pm
Since setting my batteries back to Lead Acid, the last three days show my system reaches 56.4v for 4-8minutes then drops back lower eventually getting back to Float at 54v (1st screenshot) as opposed to (2nd screenshot)
March 7, 2026 11:58 pm
@mtjag is this with the new firmware ? In your lithium plot on the right did you have closed loop with bms talking to batteries ? The first one on left ( lead acid ) does look like it is holding float pretty good. Did you have loads on it through out the day ?
March 2, 2026 9:22 pm
very interesting.... can you export all your settings: Maintenance>read>export PDF and post so I can see what's different?
you might take a look in Data> DataHistory to see if vBatt is actually hanging at float, or just taking a long time to sink to vFloat-1.0 v if you have very small loads. I don't see the cycling until/unless loads > PV pull down the vBat lower than vFloat-1.
March 7, 2026 12:05 am
@dtbaker61 The screenshot on the left is with Lead Acid selected as battery type and the one on the right was with Lithium selected as battery type. Vbatt is holding a float for most of the day based on Date>DataHistory with the settings Lead Acid
March 7, 2026 3:45 pm
I am ok with the updated behavior provided by ccaa=1c1212
Page 5 / 5
Prev
