Upgrade to Battery Firmware 3015

Hello, have upgraded to latest battery firmware and the battery says calibrating but has been stuck on discharging at 9w and stuck on 41% for hours. How long should it take and is this normal?

When I updated to 3015 it took a little over 4 hours but it should be discharging or charging at more than 9w.

It does not sound right to me.

I had trouble with the online process and had to call Givenergy and do a manual update, in the end, using the USB sticks in the battery and a BMS update file from them.

Same with the subsequent Inverter update too. It failed and the inverter went offline and would not work. I was on the phone from 10:00 to 15:00 on and off trying to get a fix and in the end, a manual update was done on the inverter too. I am up and running now but it was a stressful day.

The Givenergy technicians on the phone are great but this update process should not be like this.

Hi folks, would appreciate some advice. After the update from portal failed as did an attempt from Helpdesk, I'm updating my two 9.5kWh batteries manually following the KB article. Looking at this I need switch the green handle switch to the down position then turn off the battery itself.

My switch seems to be on the battery, not separate - (if this is not the right switch then let me know), I have switched these to the down position AND pressed the battery switch, BUT the lights are still on the battery? Should they be off? I does seem to be off since there are no battery readings on the dashboard.

https://photos.app.goo.gl/5SSxJ3buHezP6V4r6

I also have a solar panel isolator between inverter and battery, I assume I also need to switch that off?

https://photos.app.goo.gl/EhJnRiB2L6nhB86L9

The 9.5 battery has the isolator on the LHS behind a clear plastic cover.
If you have another isolator between inverter and battery that sounds like overkill , but you turn that off as well to be sure.
The on/off button is on the RHS of the battery. Press and hold and it should beep and that will shut down the battery.
When I did the upgrade from 3012 to 3013 it did not work and I had to do a reboot but the on/off switch did not work and GivEnergy had to talk me though some sort of manual unplug the power lead inside the battery to kick things back to life.
All fine now

SteveCook Thank you Steve. The instructions just said turn it off, was not clear it needed a press and hold to do so. I'll turn off and give it an hour to sit, as per instructions.

https://photos.app.goo.gl/RRia5rnrVJ6ZoRWJ9

Looks like that's a USB stick on the blue PCB - I'll be annoyed if that is glued in.

EmmetBrown
It was harder cutting the glue off the USB stick than doing the update

SteveCook Oh, wonderful, which bright spark decided to glue it. How did you cut it? I don't want to agitate the board.

Edit... I've put the cover back on and switched things back on again. Will discuss with support in the morning - they told me that my batteries were on different versions, 3008 and 3009, could explain why the over the air update didn't work.

Edit 2: Dashboard... Battery Firmware Version was 3008... now says... 0.

EmmetBrown It is a nuisance for sure. I used a Stanley knife blade.

Why are you doing this upgrade? What is in 3015 except bugs? Can anyone tell me?

gwynnej3 I've got two batteries for some reason on different versions and I get the annoying jump when charging. Generally speaking I'd always rather be on current version, unless it has issues which is why I'd held off previous updates.

When I asked about adding another 9.5 this is what they sent me on the basis I had upgraded my existing battery to 3013 and a "new" battery turned up but happened to be 3012

"in regards to adding another battery, i recommend the installer cross checks the voltages and % of both batteries before installation, he may have to force discharge the current battery or charge it to match the capacity of the new battery,

with the master battery being on 3013 i also recommend a manual firmware update on the master battery before installation of the new battery, then do a manual firmware update on the new battery to 3012 as well. please ask the installer to call us up as soon as he is ready to do the work and we will provide him with the firmware.

just a quick rundown for simplicity:

1. ask us the delete the pre-existing commission
2. read capacity (%) and voltage on the primary battery and new battery
   -force charge/discharge to match the new battery
3. perform a manual firmware update on the master battery first to 3012
4. add the new battery to the system
5. manually update the firmware to the new version on the new battery to 3012
6. go through a new commission making sure to add the S/N of the new battery on top of the pre-existing one
7. after this the system should reflect the installation
8. When this is complete i recommend putting the batteries through a recalibration to pull the batteries back in-line with each other"

EmmetBrown Fair enough. I'm disturbed by the absence of a release note for the new version however - also the absence of anyone from GE on this forum. I'll stick with 3012 for now.

Do let us know how you get on.

Cheers.

gwynnej3

3015 release notes here: https://kb.givenergy.cloud/article.php?id=22&oid=25

pacemaker But they're pretty poor as release notes go. We don't know, for example, which of the bugs fixed by 3014/3015 were present in 3012, or introduced by 3013/3014.

gwynnej3 The release notes are on the Givenergy website in the knowledge base.

I have four 3015 a definite improvement on 3013. No more SOC step changes. Caveat emptor:- 3013 was much worse than 3012 in that regard for me.

I'm on 3013, which, to be fair, has been and still is working OK for me. Tried twice now to upgrade to 3015 - update sticks at 90% verification, then fails, and I have to reboot inverter and batteries to restore normal working.

Nothing new for me - have never successfully auto updated, always have to get the files from support and put them on the memory sticks manually, which is no biggy. Pretty certain my installer didn't set something up correctly - already found battery 2 with the DIL switches set wrong.

I'll contact suppport when it suits me and get the latest firmware.

Updated to 3015 last night during the Intelligent Octopus off-peak period (for tre caljbration) which appeared to go fine at the second time of asking, in that the update seemed to take, and then the calibration took place, fully discharging then fully charging.

However afterwards the inverter just wasn't interested in discharging normally, and the local modbus API (as used by the Home Assistant givenergy-local integration) appears to have vanished.

Tried changing settings in the app, soft rebooting inverter, hard rebooting battery and inverter and eventually discovered a bunch of settings in the portal (e.g. zero charge and discharge) had been changed to essentially prevent the battery from being used for its intended purpose - to charge and discharge, no matter what was set in the frontend of the app. I suspect the vast majority of owners would not have discovered this without assistance and it took me a while as there was no reason to edpect such a thing might have happened. Many other settings were untouched.

Very frustrating. Why on earth would:

• an update not save and restore settings if it really has to wipe them?
• this not at least be documented if it is unavoidable?
• default settings (which I assume they are, as noone in their right mind would deliberately set these as part of an update) be designed to prevent the battery from operating as intended?

Seems crazy, but is also a total waste of owners' time, and just feels thoughtless.

The API is still gone so there's that too that requires further investigation and more time wasted. May have to revert to 3012 if I can't fix this soon but who knows if that will bring the API back or how easy that even is...

I appreciate software, batteries, inverters etc can be complex but it's not the complex under the hood bits that go wrong with GivEnergy, it seems to be the simple, easy to test for, easy to prevent stuff that causes problems. Feels like working with tech in the 90's when you daren't touch anything for fear of the whole pack of cards collapsing.

Sorry for the rant, but I'm tired of this, it really shouldn't be this hard to live an appliance like this.

andyn assuming the inverter firmware also updated. Newer inverter firmware versions include additional checks on the comms with the modbus interface, last I checked givenergy-local had not updated to include support for those check and thus would not work with new inverter firmwares. I moved to GIVtcp for this reason.

Doesn't explain the non charging etc but may explain the loss of the data in HA.

Arg0t That's interesting, thanks!

What you say makes sense - where did you find this information about the additional checks and impact on givenergy-local?

I only saw a battery update to 3015 in the prompt to update, and I went ahead specifically for that reason. I see the inverter is now at 912 though, whereas I think it was 909 before so it does seem to have updated I think. Great :-(

I checked here beforehand:

https://kb.givenergy.cloud/article.php?id=22

where there is no mention of an update to the inverter firmware too. If there had been I'd have maybe looked more deeply for potential API issues. I checked anyway just in case, and none appeared on the surface to be relevant, but see now in a comment I missed within one of the issues there is mention of something which is likely what you are referring to?

givenergy-local is dependent upon the givenergy-modbus library which seems to have abandoned its old stable version, while the pre-release of the new version is not yet incorporated into givenergy-local, so maybe I have to choose between givtcp or rolling back the updates (if that is even possible, I've no idea yet).

The idea of switching to givtcp doesn't exactly thrill me - afaik givtcp is not really a drop-in alternative to givenergy-local, so I'd have to review and likely redesign lots of automation logic and probably dashboards, and of course introduce more complexity with mqtt, and I don't have time for a project like that right now, but my hand may now have been forced as everything is broken, so we'll see. I could really do without this right now...

andyn The idea of switching to givtcp doesn't exactly thrill me - afaik givtcp is not really a drop-in alternative to givenergy-local,

GivTCP includes a copy of givenergy_modbus that has been updated - that's the version I've decided to use. (See also https://community.givenergy.cloud/d/3317-which-modbus-library ).

DD Yes, that's what I discuss above, but moving to givtcp is non-trivial.

andyn

Yes 912 is the updated firmware on the Gen2, I noticed the upgrade 6 days ago. It is a weakness in the giv portal that they do not have separation of the battery and inverter firmwares. They did announce on here some time ago the change in API checks but agree not everyone can spend time keeping up to date on breaking changes.

GIVtcp took me a few weeks to really settle down. MQTT wasn't complicated, just creating a Mosquitto container and adding the integration. The biggest challenge being the change in entity names between the two solutions and the impact on energy history. I ended up recreating all my dashboards and automations plus hacking around in the DB statistics to update entity IDs for history and running totals for the new GivTCP entities to ensure the HA energy dashboard retained full history.

andyn Ah - I thought you had meant that adopting givtcp whole would be an issue - didn't realise the givenergy-modbus inside was incompatible with the original one.

DD No that is what I was saying: to migrate from givenergy-local to givtcp (because it works with 3015/912) would be a lot of work that I want to avoid.

I'm assuming you can't just take the givenergy-modbus from givtcp and plonk it in givenergy-local (in fact this never even occurred to me), as there is an open issue for givenergy-local for adding support for the newer version in givtcp, but if you know differently that might be a solution.

Arg0t Yes indeed, I rarely come to these forums unless I need to find something out so wouldn't see new posts in the meantime.

And yes, that is the sort of work I am worried about - that simply isn't gonig to happen for me, so it seems I will either need givenery-local to be updated (which doesn't look likely imminently) or to revert the firmware. From searching the forum, it seems that will require a call to GE support...

andyn if you know differently that might be a solution.

No, sorry, I don't know for sure. I had just assumed that the GivTCP people forked a copy of givenergy-modbus reluctantly, and weren't making gratuitous incompatible changes to it.

I took a look at the GivTCP copy recently, and cli.py had rotted slightly - that's a piece that GivTCP wasn't using, and hadn't updated - but with a few minor updates (now merged back), it seems to work fine.

DD Sorry I missed this post earlier! I'll take a quick look at that when I get some time, could be a shirt-term solution! Also, in another post I've mentioned someone else in the HA forums has managed to do something similar and I've asked if they can share what they did 🤞

Well i bit the bullet and clicked on update from 3012
Got a message straight away saying update failed.
Refreshed the portal and tried again and it is now updating.

SteveCook Goodluck!

Went perfectly, and all calibrated and working.

So this is interesting I was finally able to get my firmware updated to 3015 but did not have to do it manually. A whole back I had the same issue as others, the update getting to 90% (verification of the file) then failing. Spoke to support and they said they had decided to pull remote updates for some time. Come to last week, the update button was back... but when I tried to update there was the same issue again. I called support, they tried from their end, to no avil. Remote updates just don't seem to work for some people. Support chap thought that perhaps it was because my two were on different versions (3008 and 3009), and for some reason the new version wasn't getting to both batteries. Anyhoo... next step is manual upgrade which I hadn't got around to due yet due to the USBs being glued in.

So last night I was on the dashboard and thought what the heck I'll try a remote update... and it worked (both batteries confirmed on 3015). Maybe support trying to push the software did something, or they recently changed something about how the update is delivered.

I am one of those for whom auto updates never work. Manual works just fine though, and 3015 seems to be behaving itself so far.

I've just successfully upgraded to 3015 from 3013. Inverter is stating version 3015 but no calibration occurred. Any ideas?

mbrennen my recent upgrade also had no calibration. The advance notification indicated that would be the case.

Thanks, looking forward to seeing if there are any improvements to my peak discharge rate.

I think if its just a battery firmware upgrade that no calibration is required. Its if the inverter firmware is changed that it needs to recalibrate

geoffreycoan Other way around IME.

3012 to 3015 did not seem to need a calibration, although one could be useful.

Maybe they've changed the calibration rules ... calibration definitely ran when I upgraded to 3015 back in November.

I think the new style firmware calculates the SoC in a different way, or has extra checks to make sure it is hitting the low and high parts of the battery more accurately.

So to move from old style firmware to new style firmware calibration was strongly recommended. To move from new style firmware to new style firmware isn't as necessary now - still sometimes useful depending on usage patterns of the system, but not needed for everyone.

mrand31 Same Inverter to DO.20-A0.205 and Battery 3012 to 3015 on AC3.0 2x9.5 went ok (tried several times before but it always failed).
It did not do a calibration afterwards.

So far - no issues on 3015 + latest Inverter firmware

I would have liked to have a calibration run as as far as I know it has never been done sinve installation.

TX200 I think the new style firmware calculates the SoC in a different way, or has extra checks to make sure it is hitting the low and high parts of the battery more accurately.

I vaguely recall someone saying that it would figure out the endpoints over time, rather than requiring an explicit calibration step ?

Since upgrading to 3015 a month ago despite full calibration manually initiated by GE support I had multiple SoCs between 10 and 40%. Anyone else is having this as well?

15 to 20% SOC difference between my two 9.5 batteries when discharged. Oddly the primary has a lower SOC when discharged than the secondary but higher on the primary than the secondary when charged.

mbrennen isn't 3015 to 3013 a downgrade? I'm still on 3012 considering moving to 3015 so wonder why you've gone backward. Thanks!

gwynnej3 The majority of us who have stuck with 3012 are reasonably comfortable with our decision.
IIRC while 3012 does suffer some rapid SOC corrections (rapid, not instant), 3013 caused many BMS over/Under voltage issues. 3015 provides more capacity and is better but is still not right for some. 3017 is out in the field as a beta, but some users still have issues. 3018 is being talked about.
I intend to stick with 3012 until there's a groundswell of opinion that a newer BMS version has properly fixed stuff.