EG4 Community Forum

6000xp bug in charg...
 
Notifications
Clear all

[Solved] 6000xp bug in charge controller setpoint voltages with open loop

53 Posts
6 Users
6 Reactions
5,888 Views
Posts: 15
(@ouijifour)
Active Member
Joined: 1 year ago

yesterday I had closed loop and for some reason it went up and down 5 times from 55.6 to 53.9 v. I decided to go back to Lead Acid settings and it didn't act so good because it went down to float but then started climbing again to absorb. I shut off the pv inputs because I can divert them to direct pv loads. After both of those were off the inverter started giving me faults and shutting off and restarting the 240v inverter for no apparent good reason. The fault code was E019 Bus voltage too high which is weird since there was no pv going into it at the time. I shut everything down and restarted and it seems to be goign okay now. I reprogam via computer Maintenance screen. When I have done that I also notice that when I make changes to pv charge settings it will shut down the inverter. This has me wondering if the whole unit needs to be shut down and restarted after making any changes.  I do have a different charge controller on the same battery but I look at those and they will drop down as programmed to their float value - it is only the 6000XP that acts odd.  At first I did still have the closed loop cable still plugged in when I changed it to Lead Acid mode but that shouldn't make any difference - does seem like it since it was not displaying the bms SOC. Seems like maybe this inverter is geared more towards having AC input in it than being strictly off grid capable. The firmware really seems to act strange with just PV input and using the inverter.

 


Reply
1 Reply
(@jacobranchsystems-com)
Joined: 5 months ago

Active Member
Posts: 5

@ouijifour Yes, I agree that algorithm is sort of hit and miss. There are days where it sort of works, other not at all.  Frustrating...


Reply
Posts: 5
(@jacobranchsystems-com)
Active Member
Joined: 5 months ago

I'm a new 12000XP customer, and also shocked by the lack of a functioning Bulk/Float  feature for my SLA bank.   

I own a company selling wireless devices, and if we had that level of bug in our firmware it would be top of the list to fix!

Any updates from EG4's ?


Reply
2 Replies
(@jenzilla88)
Joined: 5 months ago

New Member
Posts: 1

@jacobranchsystems-com same, really regret buying this.  I have FLA and the inverter says it works with them and has a built in charge controller, but obviously it shouldn't be.


Reply
(@ouijifour)
Joined: 1 year ago

Active Member
Posts: 15

@jacobranchsystems-com I have the 6000xp and was amazed when I got it, and disappointed that the battery charging does not work very well. There is no absorb time in the lead acid mode.  But don't think that if you had lithium and were using closed loop that would work better. It seems to pick it's own float point even though the bms is set for a different setpoint - and the bms should dictate the charging. I thought so anyway. The inverter part seems to work fine -but the battery charging seems have been more of an afterthought than something well designed for off grid use. All the other grid or AC input setpoints - which are much more elaborate in user configuration , I don't use since I am off grid.


Reply
Posts: 5
(@jacobranchsystems-com)
Active Member
Joined: 5 months ago

I also have the problem that the Bulk/Float setting doesnt work.   Is there a new firmware for this now?  Jacob


Reply
Posts: 5
(@jacobranchsystems-com)
Active Member
Joined: 5 months ago

@dtbaker61: Did this issues ever get resolved - I have the same... thanks Jacob


Reply
3 Replies
(@dtbaker61)
Joined: 1 year ago

Estimable Member
Posts: 117

@jacobranchsystems-com there is an update available for the 6000xp that MAY fix the open-loop Li charging, but the 6000xps I have installed are very remote, and with no wifi or cell service. There is nothing available for 12000xp as far as I know.

It's getting pretty ridiculous that there is no reliable Li-friendly CC-CV charge possible open-loop. The 12000xp will hit and hold at "absorb", but never settles to float. Holding at 56v is NOT good for Li battery or BMS life, but you have to hit it to enable proper top-balancing to occur at low current.

I want to see CC-CV with end of charge at 56.4, and drop to float at 53.0 when batteries are full and sun is still out!

 


Reply
(@ouijifour)
Joined: 1 year ago

Active Member
Posts: 15

@dtbaker61 well said about the problem.  I regret getting the eg4 6000xp . I just assumed that it was advertised as offgrid use and has closed loop that it would be modern and work great. I am just using the inverter part of it for now and using a Midnite Hawkes Bay and a Victron 450/100 controller . It would be nice to be able to use the two mppt in the 6000xp again.  Maybe Eg4 pushes new products out so quickly that they don't have the resources to fix the older stuff and make it work correctly. Or maybe it is not capable hardware wise. I still can't believe that the controller lacks the basic essentials for lead acid mode that have been around for such a long time - Absorb, Absorb Time, and Float.  Not to mention ending amps but I could do without that if there was an absorb time and if Float would stick until the next day.  I was hoping that closed loop would work, but it is a bit of a mystery what the 6000xp uses from the bms to make it's determinations. It certainly doesn't seem to follow the bms parameters all the time.


Reply
(@jacobranchsystems-com)
Joined: 5 months ago

Active Member
Posts: 5

@dtbaker61 Thanks - I'll be very curious to see if the fix for 6000xp actually fixes your problem, since I assume the firmware for 12000 is very close so the fix will make it there as well...


Reply
Posts: 15
(@ouijifour)
Active Member
Joined: 1 year ago

I have local control for monitoring my EG4 6000xp and could probably do some control too in the future. There are two ways to approach this. One is to use a Home Assistant integration.

This one works well https://github.com/ant0nkr/luxpower-ha-integration

Go in your router and shut off internet access to the wifi dongle . It will still work locally on the home network. This way you won't have the data going out to remote servers, which can slow things up locally since it prioritizes to servers first. 

The second method is to unplug the wifi dongle and make a connection to the rs485 jack. I just got mine working that way but only using it for monitoring right now. Here is info on that 

https://github.com/larduino/EG4-6000XP-Home-Assistant-Local-Control/blob/main/README.md


Reply
8 Replies
(@dtbaker61)
Joined: 1 year ago

Estimable Member
Posts: 117

@ouijifour your reply is completely off-topic for this thread. The issue is faulty charge control by the 12000xp when running "open loop" with user defined voltage setpoints for generic LFP batteries. This is not a monitoring issue, it is a charge control BUG.... which EG4 says they are working on.


Reply
(@ouijifour)
Joined: 1 year ago

Active Member
Posts: 15

@dtbaker61

I would differ that this is off topic. With direct modbus control to home assistant , it is possible to write automation that would get close to achieving the goal of having a better charge cycle that both of us desire . For example it could be probably be forced into float at the appropriate time. I have not gotten to that yet, because the pv charging was rather unreliable in doing what I wanted so I have other controllers now. I may eventually use the mppt in the 6000xp again though. I have a very reliable accurate Victron Smart Shunt in my system which could provide point of truth setpoints for when to make voltage transitions. 

Regarding the comment that EG4 says they are working on this - hopefully, but it has been two years  or more. When companies are releasing new products over the years, with a limited staff the priority to fix older products is most likely much lower than supporting or fixing bugs in the newest generation stuff. There may also be hardware limitations on fixing the mppt charge behavior too. 

So we are left to do what we can on our own sometimes to make improvements. I would love to fix  firmware code myself and assume the risk of breaking something, but that is not possible. So we work with what we have to work with. 

Thanks for your comment. 

 


Reply
(@dtbaker61)
Joined: 1 year ago

Estimable Member
Posts: 117

@ouijifour As I noted.... this bug does appear to be fixed for the 6000xp.... but not for the 12000xp yet.

There is no need for the average user to hack into the charge controller software, or use external charge controllers if EG4 fixes the bugs the way they should...


Reply
(@ouijifour)
Joined: 1 year ago

Active Member
Posts: 15

@dtbaker61 There is no Absorb Time for the Lead Acid mode so no way to give a bms time to balance cells .


Reply
(@dtbaker61)
Joined: 1 year ago

Estimable Member
Posts: 117

@ouijifour the time it takes for voltage to rise from 54 to 58 (during the CV portion of charging) is typically plenty for internal top balancing for lithium. There is no absorb time required for correct LFP charging which is simply CC-CV to end-of-charge voltage. no absorb, and no equalize time is required.

Once Veoc is hit, the system *should* relax to a lithium-friendly float target of 53.0v or thereabouts.

260116 6000xp correct Veoc to Vfloat

 

The current 6000xp charge curve is pretty much correct


Reply
(@ouijifour)
Joined: 1 year ago

Active Member
Posts: 15

@dtbaker61  That is assuming someone wants to charge to 58v . I only charge to 55.2v which is 99% full. My balancing does not turn on until 55.2v so it needs some time at that voltage to let the balancing happen. Your bms is obviously different so I guess what you have suits your batteries and bms needs. 

 


Reply
(@dtbaker61)
Joined: 1 year ago

Estimable Member
Posts: 117

@ouijifour most mfg BMS don't start balancing until 56.0 or thereabouts as the current drops during the CV phase of charging. An end of charge voltage of 58v for a typical 16 cell LFP batteries is not just personal preference, it is what is suggested by mfg.

No "absorb" time is required for LFP per any mfg I am aware of with a CC-CV charge curve.


Reply
(@ouijifour)
Joined: 1 year ago

Active Member
Posts: 15

@dtbaker61 I am glad you have something that works for your system and how you want to manage it. Most other controllers are very flexible to work with any different kind of bms and different charging strategies.  I follow guidance from Andy at Off Grid Garage on youtube. He has done extensive testing and explains his results. Charging to a higher voltage can charge the batteries faster, but going that high can be at the expense of battery cell longevity. There are studies on this . To me it is not worth the risk of this degradation by trying to get the last 1 or 2 % of battery capacity charged. LifePo4 cells are quite happy and efficient being charged lower . My batteries all have bms that the setpoints can be programmed so I am not stuck with setpoints from years ago when the understanding and science of LiFePo4 was different.  

I am happy that you have what works well for your system, but there is nothing wrong with different approaches to cell charging and maintenance.  I compare what is available in the Eg4 6000xp and it is not flexible to even what I have available in my older Midnite Classic charge controllers.  And some people still use real lead acid batteries and the lack of absorb time setting  in the 6000xp is a deal breaker for charging those. 


Reply
Page 4 / 5
Share: