EG4 Community Forum

12000xp open loop b...
 
Notifications
Clear all

12000xp open loop battery voltage bounces between float and bulk after batts are full

18 Posts
3 Users
2 Reactions
162 Views
Posts: 91
Topic starter
(@dtbaker61)
Estimable Member
Joined: 1 year ago

I have a 12000xp up and running off-grid, configured open loop as "lead acid" with:

absorb = 56.2

float = 54.0

eq = 54.0

eq period = 0

eq time = 0

...the charge control is not behaving like I want. It charges up to 56, and drops to float at 54, put then flutters between 54-56-54-56 rather than actually float at 54 for the remainder of the afternoon while the sun is up. I don't see any setting for time or voltage hysterisis to prevent the flutter so that the system returns to float rather than go all the way up to EOC voltage multiple times in the afternoon.

Is this a known bug ?

solution coming ?

 


17 Replies
Posts: 91
Topic starter
(@dtbaker61)
Estimable Member
Joined: 1 year ago

Here is a picture.... voltage is floating at 56 instead of 54 as defined. I consider this to be a BUG, and not happy about it since holding batteries at this voltage can shorten their life.

251110 NOT floating at 54

Reply
EG4 Eric
Posts: 761
Admin
(@eg4eric)
EG4 Online Support
Joined: 1 year ago

What batteries are you using with your system?


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

Estimable Member
Posts: 91

@eg4eric  6x Midnite Solar powerflo-5. I do NOT want to use closed loop comm cables, which is why I set parameters as shown in top post to enable the inverter to manage charge control by voltage setpoints.

Interesting note is that charge voltage absorb -> float looked fine when there were no Loads after I set the EQ voltage= float. However, once the system was operational with "real" loads, the voltage keeps returning to end-of-charge (absorb) voltage.

I suspect that there is faulty programming when running "open loop" which needs to consider that return to EOC is not required after small dips below float of short duration in the same day... System should just return to float voltage, not EOC voltage. They probably need to open up the allowable sag a little to create logic that the system doesn't need to return to EOC voltage unless voltage drops to "operational" voltage for a minute or more to account for temporary cloud cover or momentary sag from heavy output.


Reply
(@dtbaker61)
Joined: 1 year ago

Estimable Member
Posts: 91

I had the same problem with the first 6000xp I installed.... and it was eventually resolved with a firmware update. I would have hoped the same logic is used with all models when running off-grid, self consumption, open-loop... Lithium battery charge curve "faked" with EQ=float=0 days, 0time


Reply
(@krzysiek_kta)
Joined: 12 months ago

Trusted Member
Posts: 57

@dtbaker61 

Are you positive the issue was resolved?

 

I had some issues with my EG4 Wall Mount battery BMS comms with 6000xp (finally resolved) and was forced to run in open loop (Lead Acid) mode for months.  Below is an sample  screenshot of the battery voltage throughout the sunny day when the PV array production exceeded the demand and the battery was topped up during the day.

image

NOTE that I used AC Grid to charge it overnight and there was no issues whatsoever with maintaining the Battey in float voltage as set in the  6000xp

image

That issue was main driver for me to fix the BMS comms issues and go back to closed loop comms (Lithium Battery type 1) where the voltages are managed better - see below example of sunny day when the PV array production exceeded the demand and the battery was topped up during the day:

image

NOTE that no overnight AC Grid charging.

I run the latest firmware ccaa-191011

Just my 5 cents.

 

Brgds

 

K


Reply
(@dtbaker61)
Joined: 1 year ago

Estimable Member
Posts: 91

@krzysiek_kta I don't want closed loop comms, I want to program voltage setpoints. This is working correctly finally on a 6000xp system that is offgrid, with no AC-in available. Same settings are NOT working on an off-grid 12000xp

 


Reply
(@krzysiek_kta)
Joined: 12 months ago

Trusted Member
Posts: 57

@dtbaker61 

Got it.

Yet it seems when in Off-Grid mode with open loop comms during the day my 6000xp did not maintained 54v float voltage as programmed and voltage behaved as in OP 12k case. See example below.

image

My point is that even with latest 6000xp firmware the Off-Grid voltage management in open loop comms does not work as programmed.

I wonder how bad it could be for batteries if one has over-paneled PV setup and it stays fully charged most of the time ....

Brgds

 

K

 

 

 


Reply
(@dtbaker61)
Joined: 1 year ago

Estimable Member
Posts: 91

@krzysiek_kta it is not great for LFP to hold at 56v, and its hard on the BMS and charge controller to cycle on and off. There needs to be a reasonable hysterisis (like float -1.0 v) to be lithium-friendly. I'd be fine with seeing "float" bounce between 53.0-54.0... knowing that as soon as the sun sets, 53.3 or so is still "full".

More importantly, holding at 100%, or bouncing up to EOC many times in the afternoon increases the chances of overvoltage, killing the internal BMS, and fires.

some brands of batteries have this worked out internally, but you have to have closed loop for that to work. My point to EG4 is that the inverters all need to be able to have internal charge control managing by voltage setpoints as well to accommodate batteries that do not have closed com.

I don't like to used closed comm in many off-grid situations because I don't trust the durability of cat5 cable or port in a typical off-grid shed or even an RV/camper with vibration from the road long term. I want the inverter MPPT charge control to work predictably as programmed with a LFP friendly charge curve and float.

 


Reply
Posts: 91
Topic starter
(@dtbaker61)
Estimable Member
Joined: 1 year ago

@eg4eric , the charge curve and behavior we really want to see for open loop charge control of Lithium batteries is a simple CC-CV curve where:
- daily (PV voltage comes up to minimum) charging happens at max current until "Vfloat", and then
- switches to CV= Vend-of-charge once a day, and then charging stops, and allowed to drop to Vfloat, and is maintained at Vfloat for the remainder of the day as detected by the presence of PV voltage available.

typically there is some bouncing +/- float as loads and pv available vary thru the day, but a good charge controller algorithm controls that pretty tightly, always seeking to return to Vfloat for the day as long as PV power is available. When PV voltage drops < operational, the controller considers that as "night", and resets charge curve to shoot for Veoc the next "day".

Unfortunately, what I see with EG4 inverter charge control (open loop), is the Battery voltage bouncing between float and EOC and sometimes holding at EOC despite settings for 0 time and 0 days for EQ voltage.

What we want to see is 56.2 once a day, and float targetting 54.0 for the remainder of the day with minimal bounce to be expected depending on PV available versus loads


Reply
7 Replies
(@krzysiek_kta)
Joined: 12 months ago

Trusted Member
Posts: 57

@dtbaker61 

Maybe new firmware will offer an alternative - look at the latest for 6000xp - new battery type 30 for lithium batteries in open loop added  (if I read the changelog correctly).

I am reluctant to try it yet, but maybe similar version for your 12000xp would eventually do...?

 

Brgds

 

K


Reply
(@dtbaker61)
Joined: 1 year ago

Estimable Member
Posts: 91

@krzysiek_kta most of the xp series I have installed have been way off-grid with no wi-fi available, which makes firmware updates after install difficult without easy delivery from a local laptop, and expensive for the clients if I have to make a physical trip and charge them for it.

I do have one 6000xp, and one 12000xp, with wifi access for firmware updates.... but it can be difficult to do remotely if the system has to be put on standby and the residential modem is on a circuit that gets de-powered. Most clients can't get thru the process on their own to power the modem from a secondary source during an update; so it becomes a time consuming (expensive) process.

I'd want to be very sure that the update is "worth it". I'd want to see a written report, video, or something to show the results.


Reply
(@krzysiek_kta)
Joined: 12 months ago

Trusted Member
Posts: 57

@dtbaker61 

I have two 600VA UPS units -  one powering  the wi-fi and  the other internet gateway.

All firmware updates I have done were in off-grid mode thorough EG4 monitoring website.

When all went well the power loss when inverter restarted was just few seconds and easy handled by UPS.

The longest it always took for the Wi-Fi dongle to come back online (couple of minutes).

So in theory such setup can be upgraded remotely.

 

But.... errors can happen (my first hand experience) and I am always on site when performing upgrade. Just in case.

Good luck

 

Brgds

K


Reply
(@dtbaker61)
Joined: 1 year ago

Estimable Member
Posts: 91

@krzysiek_kta aha.... see.... what if the customer doesn't have a separate UPS for their wifi ?

don't you have to turn off EPS to inverter in standby to initiate a remote update of the inverter via app or portal ?

This means if the customer doesn't have a separate UPS for wifi, I HAVE to go on site with one, and execute the update while I am on-site. That is the time consuming (expensive) part for the client.

The only other way around it would be a "bypass" circuit from the generator that skips the xp completely to keep all loads powered up while the update is in progress. This would be a great thing to build into the xp since they are intended for off-grid use, but currently you'd have to design/build a bypass in externally.


Reply
(@krzysiek_kta)
Joined: 12 months ago

Trusted Member
Posts: 57

@dtbaker61 

Well,

As hobbyist  BETA tester I always build backup plan. I do have externals bypass transfer switch separating Inverter from the house power and allow grid or other more conventional power source to be available to power the house independently - that was the first think I did starting my Inverter/battery project journey....

Yet the power blips issues with EG4 inverter/battery (read my other posts) made me invest in multiple UPS and even it home Surge Protector in my main load panel....

I was just sharing my experiences with 'remote firmware updates'. I think the UPS  for WiFi and Internet gateway is a MUST for any Off-grid setup. Just my 5 cents.

 

Brgds

 

K


Reply
(@dtbaker61)
Joined: 1 year ago

Estimable Member
Posts: 91

@krzysiek_kta The issue with planning for UPS for the wifi is problematic whether the install is grid-tied, or off-grid, since normal functionality desired by clients is to have the wifi modem on a normally protected Load circuit.

The two best (least expensive for client) solutions would be:

1. adding a "bypass" switch from either/both the AC-in and Gen-in terminal blocks to power the Load breaker directly for updates, or if the inverter fails completely and you are waiting for a replacement.

2. providing firmware updates (and instructions) as files that can be uploaded from a laptop thru the Dongle  or rs-232 port or BT local connect WITHOUT internet. i.e. connecting a laptop to the dongle wifi and doing updates from a laptop locally rather than worry about www internet connection.


Reply
(@krzysiek_kta)
Joined: 12 months ago

Trusted Member
Posts: 57

@dtbaker61 

I can't agree more with you.

Whenever the system is marketed as Off-grid it should have to be manageable (upgrades, serviced and so on) without Internet. Old school USB stick upgrade should be the way to go.

My point of UPS for any off-grid setup was purely practical - whenever one needs to work on the system (read: 'interrupt it') -  uninterrupted access to internet to Wi-Fi/internet becomes essential.

Re 1. -  I think I made my point earlier - any NEW technology implementation needs backup (read bypass) solution as a starting point.

Re 2. - These days I doubt any company will go for that path as they're aiming for online profits. I highly doubt any viable offline solution will be available. PPL (including vendors) tend to get Internet  access for granted. Unless one decides to rely on phone data access as backup.

Concluding

I am nowhere close to work on commercially viable solutions to provide service to others ...

I am just voluntarily early adopter of the EG4 offerings (inverters, batteries hybrid mini split ACs...)  had some fun but not without the 'early adopter pains"

 

Brgds

 

K


Reply
Posts: 91
Topic starter
(@dtbaker61)
Estimable Member
Joined: 1 year ago

ok, getting back to the original concern....

As far as I can tell, there is NO WAY to obtain a "proper" open loop charge curve, (charge controlled by Inverter using voltage setpoints for end-of-charge and float) for a LFP battery with the 12000xp.

To me, this is a serious bug and deal breaker for using the EG4 with future clients. The manual and documentation indicates that it should be possible, but it is not.

I really need to hear back from EG4 that they are aware of this, acknowledge it is a problem, and are actively working on it.

How about it @eg4eric ? Is there any work being done to resolve the problem? I'd be glad to Beta test at my closest customer site if Engineering can give me a way to install a beta update.


Reply
Share: