GivTCP died at ~00:10, corrupt inverter date

Well that's 2 hours of my life gone trying to work out why GivTCP was repeatedly failing to talk to both inverters.

Noticed predbat was giving repeated REST and API errors and found in the GivTCP logs the errors

2024-12-31 02:21:42,241 - G - read        -  [ERROR   ] - inverter Update failed so using last known good data from cache: ((<class 'ValueError'>, ValueError("time data '2024-12-31T02:21:42.233111+00:00' does not match format '%Y-%m-%dT%H:%M:%S%z'"), <traceback object at 0x7f22dbdcbf80>), 1687)
2024-12-31 02:21:42,242 - G - read        -  [ERROR   ] - runAll2 Error processing registers: ('KeyError', 'read.py', 1825)

Repeated for both inverters over and over again. I won't duplicate all the investigation detail here because its all on a GitHub issue I logged https://github.com/britkat1980/giv_tcp/issues/318 but in short it turned out that the date on the inverter was corrupt, and was appearing on the BBC app as '01--2025'

Screenshot from the BBC app after I had fixed the first inverter but not the second:

Re-syncing the date/time from the GivEnergy portal to the inverter fixed it.

Seems like I was not the only one with this issue, see https://github.com/britkat1980/giv_tcp/issues/145

2.4.9 is broke for me but GivTCP3 is ticking along fine, ill try time thingy despite nothing being wrong.

Maybe has something to do with the issues I had earlier with portal.

Jellybaby interesting. At least one other AIO user reported the problem on GitHub.
Maybe some specific AIO firmware version as well as Gen 1 and Gen 3 inverters affected?

Apparently several other people on the GivTCP facebook forum have reported the same issue, both on GivTCP 2 and 3.

If its all running fine on GivTCP3 then you probably don't need to do anything.

That sorted it, how very odd.

geoffreycoan I did it and 2.4.9 is back up and running, I am running 2 atm side by side

My setup on AIO firmware wise

I actually said elsewhere "RIP Predbat users that are effected by this overnight" lol

Jellybaby I actually said elsewhere "RIP Predbat users that are effected by this overnight" lol

Before I had figured out the fix I had manually scheduled a charge via the GivEnergy Portal from 6am in the morning tomorrow as we've got a free electricity event that Predbat would have handled. Glad its sorted and it fixed 2.4.9 for you, were you getting the 'no serial number error' which is what someone else reported on Github

geoffreycoan yup as well as a the getting erm, damnit cant remember exactly but something to do with consumption reading getting spammed as well as a couple of others which was weird as it found the inverter and gateway fine lol.

very very very odd

Edit
aha

Oh well, tumble dryer on now so time to hit the sack 😃

geoffreycoan I woke up at 4am this morning to the same issue, no data in home assistant since midnight. Fortunately for me I checked here first and was able to fix my system in about 30 seconds by re-synchronising the inverter date/time.
I can imagine the hours of investigation I could have spent looking for the solution so a big thank you to Geoffrey for posting so promptly.
5kW Gen 3 Hybrid Inverter

geoffreycoan much appreciated seeing this, I clicked set clock in bbc app and resolved the problem.

Single AIO

same happened to me.

Woke up to an exporting battery. Same errors in predbat. I set it to read only and manually set the giv app to eco.

My date was also saying 01-2025

Cheers for posting. Looks like i have the same issue. Was just trying an inverter restart but will try time sync too.

geoffreycoan Thanks for this. More epic fail from GE, they really are going down the toilet.

I'm glad there's a cheap slot shortly as if I'd woken to an empty battery and expensive electric I wouldn't have been overly happy.

Mines a gen3 hybrid, BTW, for info.

i have a Gen2 5kw inverter and have same issue. Running GIVTCP 2.4.9. Can someone explain how you resync the inverter Date/Time please?

Jon-boy go to this Web page

https://portal.givenergy.cloud/account-settings

Scroll down to you get to the sync time button and press it.

You should get a message back from the app (if you have it installed) when it is done.

Mine said it had failed, but it did automatically retry and confirmed successfully sorted.

Cancel that request, i found the Set Date/Time <SEND> in Portal settings

TX200 In same situation with gen 3 hybrid. Fortunately a few cheap slots to fill battery. Easy to fix once you know how. Thanks everyone

TX200 Thanks, its worked

Leeshore Easy to fix once you know how.

Just like many things in life.

This isn't a GivTCP bug it's a Giv FW bug. I don't even run GivTCP and both my Gen 1 old FW inverters reported impossible dates this morning.
Im guessing everyone will wake up to this

hoggy a very belated millennium bug 🤪

Just adding my name to the list of people this affected. Fixed via BBC app. 🙄

hoggy With no support available as they've all left/been fired.....

The trustpilot reviews over the last three months (excluding fake reviews, probably by GE) are astoundingly bad.

If this company isn't going bust then it certainly deserves to!

hoggy yes, I was thinking the same. I wonder how many customers would actually notice!

Only those that are paying close attention to their inverter and/or IHD (e.g. why am I using the grid now!) assuming they have an IHD that works and is on the right screen!

Same here. Woke to two empty batteries and Predbat full of nonsense. I did have a Predbat update which I initially thought had borked everything (not the first time). Many thanks to @geoffreycoan for spotting the fix so quickly. Snarky looks to GivEnergy for ballsing it up again.

I have managed to get some IOG slots by requesting a charge from 80% to 99%>. That should get the batteries charged to hopefully get me through the day.

We'll see if this is a one-off "bug" tonight I guess.

Going to be interesting to say the least if it recurs daily, especially for G1 owners who won't be getting any fixes....

I hope it's not daily. I don't want to wake up at 0015 each day to resync the time!

Giv do that for clock change i think. So hopefully they could push out a sync daily if they had to.

BBC app seems to be able to reset it so there must be a way to do it from home assistant possibly ? (If we have to take matters into our own hands…)

TX200
I use home assistant to monitor my system because I like the way you can customize the extra data that you can't see with the app or easily see with the portal.

Whilst the impossible date completely messed up home assistant data it didn't affect the simple battery charging between 2am and 5am and the energy graph in the app appears normal.
So it's quite possible that many people may not notice the issue if they only use the time slots in the Givenergy app.

Interesting no problems, running on software DA0.007-DD0.007-A0.092

Same problem here. GivTCP basically flatlining on all graphs from around midnight.

Super happy that the solution was published here and on github by @geoffreycoan Thanks very much!

pacemaker Chicken and egg situation - if GivTCP can't communicate with the inverter due to an invalid date on the inverter, then it can't reset the date....

I do have to note that it’s worryingly bad that the inverter software thought today was 1st of the 13th month in 2025. What kind of date/time code / framework is in there? 😱 - GMT/BST hour shift… ok - cross region time conversion… ok - but 13th month ?!?!

Vestas good point! Think the bbc app is not using givtcp so handled it more gracefully

pacemaker It should be a fairly simple change to GivTCP. I'm assuming it doesn't have an error handling trap for invalid dates (like month=13) which would be trivial to fix.

As things stand however there's no way to automate a time sync using GivTCP....

geoffreycoan I had exactly the same issue at the same time. Sadly I was awake at that time of the morning and reset the date/time of the inverter which fixed it. I assumed it was due to time drift of the inverter.

Assuming its only the month being incremented to an invalid value then at least its only going to be a monthly issue.

Still appallingly bad coding by "GE" however, but that's normal.

I wonder if there should be a routine included in predbat which sends a reset time signal every day at say 00:05?

It was a leap year this year wasn't it?
13th month thing aside, it did jump to 2025 a day early so I guess it may be a 4 yearly event.

I wondered if you can sync the time via the portal API route?

Vestas Chicken and egg situation - if GivTCP can't communicate with the inverter due to an invalid date on the inverter, then it can't reset the date....

Despite HA being unable to record anything this morning, Predbat still managed to run multiple import/export cycles correctly. So I think the issue isn't preventing all comms to the inverter.

I'm away from home again and got all sorts of error messages. I'd remotely restarted the router, HA, the inverter and then the NAS drive HA is running on before finding this thread and the fix. Running a 5kw Gen 2 inverter. Thanks @geoffreycoan!

pacemaker 13th month in 2025

Well it could be US format so Jan 13 2025

I had the same problem with the daylight saving, it just wasn’t set 😕

CW if so, jumping from 30th Dec 2024 to 13th Jan 2025 is an even bigger worry !

hoggy You've had a system since 2020 haven't you? Presumably you don't remember this happening then?

If it didn't then its a bug introduced with D448-A448 f/w onwards as that was the first production update after 2020.

Edit - I note D450-A450 had a change made regarding time which is also present in AC coupled f/w D533-A533 - Removal of the 2 am automatic restart function. Both dated middle of 2021. Could be that restart masked the bug - not that it matters what the cause is, just speculating.

Same thing here but GE App & Portal show todays date and charge window ran as normal overnight. No longer run HA, just use the App or portal as through winter have the same charge strategy daily.
Inverter App shows the error so I did a resync anyway and now shows correctly.

Well if nothing else, this got me switching GivTCP from 2.4.9 to 3.0 as I initially thought it was HA (or my RPi4 SD card again)

Thanks @geoffreycoan for getting this post up - after updating HA etc and re-checking logs (issue appeared around 00:05) I'd decided to just revisit the issue on the morning of the 1st and see how things were, with my initial thought being some predictive value that couldn't calculate across year changes.

Last minute thought was to dip into the forum and check if anyone else had post anything about it - and now, knowing it is GE and not HA/MQQT etc, this is rather more annoying (alongside the poor EV integration) and indicative of insufficient testing

rjp predbat wasn't able to run for me.

It charged to 100% and that was it.

I'm looking at the giv app to confirm that, battery stayed at 100% since 0250.

Vestas mine went in on Dec 4th 2020 and I don't recall it being a thing BUT -
We are now on the 3rd (or 4th, I forget) Givenergy App.
GivTCP wasn't even started until well into 2021.
We are on the second Givenergy Portal.
They had no API then.

So I don't think I had any concrete way of seeing the inverter date time and given it just charged up based on time (that isn't currupted) it would probably have gone unnoticed.

I do recall there being a quirk in the OLD old API where you could ask it say... 45/01/2021 and it would return the corresponding day in Feb fine but that ended when the API did. I used to use that trick to increment a counter from 0 to 365 each night to pull down daily totals without having to handle dates specifically, but that's an API thing not direct from the inverter.

My HA dashboard and Predbat has all gone AWOL. I found my inverter had gone offline for some reason. I have restarted that and inverter is back online.
Now resent the time and date.
Dashboard data starting to populate now.

Weird

hoggy Bug has probably been there since day 1 then and that code has been reused across product ranges and generations.

SteveCook is that the bug that clears the export and import values to zero on inverter restart?

Mine does that. Reported over 12 months ago and accepted as a known bug to be fixed at some point.

We have a Gen3 Hybrid inverter and the date seems fine today so far, although not using GivTCP or HA. Is it only Gen1, Gen2 and AOI? That would be odd because wasn't AOI after Gen3? Or is it perhaps because I didn't update the firmware since April, so still on D0.307-A0.307?

Thanks all for advice.
Wonder what caused the time to go awol now unless GivEnergy doing something with getting ready for New Year switchover

I wonder if it has anything to do with this SSO malarkey?
https://givenergy.co.uk/givenergy-transition-to-new-sso-service-in-further-commitment-to-data-security/

Lots of useless info on what they are forcing users to do and why.
No information on actually how to set up SSO.....

Interestingly, when I re-synced the date/time at 06:48:59 this morning, the portal figures in the meter tab reset:-

The Black Cat
Same as what mine did at 10:15 when all my HA was messed up and I re-synced the time.
Dont know why I had to re-sync now though, everything has been fine for months

geoffreycoan A big hank you for posting!
I was checking everything ( HA, GivTCP, MQTT, GivEnergy portal, WiFi settings, etc) for 2 hours but couldn`t find out the issue. 😃

After a very late night last night, I’ve just got up to see all the messages on here and on GitHub (and I assume on the GivTCP Facebook group) about the bug and fix I discovered.

Good suggestion about doing the time set in the BBC app as an alternative to to the GivEnergy portal time sync, I didn’t think of that. I did discover that it was only the BBC app (and probably presumably @Rubikcube Android) where I would see the corrupted date.

The error affected GivTCP v2 and v3 and also GivLocal as there’s a GivLocal bug linked to my GivTCP report. GivTCP did try to recover the corrupted date but there was a failure in that process which was the ‘can’t convert date/time, bad format’ message we were seeing in the GivTCP log.

Predbat carried on trying to run throughout the GivTCP failure, repeatedly complaining about REST issues. I suspect that as GivTCP wasn’t working it was running entirely on cached HA data and those for whom their battery charged as normal was due to the commands having already been sent to the inverter prior to the failure (e.g. IOG/Go/Eon Drive).
For those of us on Agile we weren’t so lucky. I had both of my inverters with PauseDischarge set so they weren’t going to be doing much.

The Black Cat I saw the same at 2:54, despite the meter data being correctly reported in the right date/time, resending the date/time causes the today portal figures to get reset to zero.

I’m hoping it doesn’t happen again tonight, at least I’ll be up then to check.

I count about 50 people on the different groups that have reported the error. Since the GivEnergy app/portal just work on time-based charging I would imagine that there are a lot of people affected but don’t know about it.

Did suggest on the GivTCP bug that people send me beer tokens and GivEnergy sends a free battery as a thank-you. I can but hope 😉

I've been working on this for 2 hours now but even resetting time isn't working for me (but then I have uninstalled GivTCP so I might now have a config issue.)
I've reset the inverter time via the accounts setting page, https://portal.givenergy.cloud/account-settings but what is this BBC app being discussed as I can't find a way to pull the current time?

Thanks for this folks, same issue but syncing time sorted immediately.. would have spent hours troubleshooting otherwise

nophead I have a vague recollection of @hoggy saying something along the lines of "the AIO is basically two G1/G2 inverters wired together with some additional control".

I may well be misremembering but the G1/G2 inverters aren't that dissimilar so code reuse is likely.

Lawrence_P I've reset the inverter time via the accounts setting page, https://portal.givenergy.cloud/account-settings but what is this BBC app being discussed as I can't find a way to pull the current time?

There is a BBC basic app that is available for our inverters, on lots of different platforms, see https://community.givenergy.cloud/d/5051-bbc-basic-inverter-app-significant-update

Gives a direct connection to the inverter and in some respects is better than the GivEnergy app. Worth having it as a backup because the GivEnergy app requires the GivEnergy portal to be working, even in Local mode

nophead no it affected gen 3 inverters as well. I have a friend who I asked this morning and he doesn’t use HA and he had the 2025 date as well.

Hook So perhaps it is a new bug introduced by a firmware update that I haven't done.

For those with my app, just press the home line to see the date/time on the inverter.

To fix the date with my app, you would need the premium version.

As far as I can tell so far, my app is not adversely affected by this bug. Schedules are based on time only and continue to run correctly.

I am quite baffled though, usually HA fills in missed hours if you shut HA done in the hour it can register stuff and just overfills the first it starts back up in again nada

Usually it would have put the first 3 hours in that 4th one

geoffreycoan is it on android, can't find it in the store?

Scratch that. Found the link to the package.

Just reading some posts on facebook, pretty eventful night some had hehehe

The Black Cat Yep this happened to my AIO too as soon as I updated the time on the portal to fix my HA monitoring

All affected clocks are re-set

Now go enjoy the rest of NYE

So I advanced the time on my test inverter ...

.. and waited ...

... and it seems everything will be fine for tomorrow.

Happy new year!!

I suspect we'll see the same problem at the end of 2028 (can't cope with an extra day in the year). I'm just glad Predbat still managed to charge the battery during the night (I'm on Agile). My first hint something was wrong was that HA thought the battery was at 4% SoC all night. Luckily before I did more than the obvious troubleshooting (restart HA) I thought of looking here and found @geoffreycoan post 🙂.

Rubikcube so your also a Time Lord too? "Who" knew... (I'll see my self out)

Side note, it could be worse, there's a bunch of Irish owners of K-Star AIO (the Libbi ones) and theirs just shutdown at midnight and won't come back on!

SJB I suspect we'll see the same problem at the end of 2028

Correct ...

... the date advances to 01/13/29.
I have checked for 2025 and 30/12 advances to 31/12 correctly, so a strange leap year bug.

hoggy Man, leap year causing problems for everyone eh lol

Jellybaby
Pretty poor when we had all the Millennium bug stuff 24 years ago. Does software ever get checked, or "just wait till it crashes and then issue a patch"

SteveCook More like "China doesn't have a leap year every 4 years" and the code got reused outside the region it was intended for.

Rubikcube

It worked ok last night and mine are in sync without any inverter restart required.

My master battery did error this morning but I think that was down to me joining the Agile thrashing routine again after several sedate weeks on Octopus Go.

So not a problem on my old inverter.

lol just checked the inverter page on portal to write down how much charge went into battery overnight and saw only 10.3 went in according to the meter data and I was like WTH

Then I scrolled down and saw this lol

M_J hmmm saw that it might have been sorted by the GivEnergy team before I checked it just now?

Thanks so much. Wish I had checked here first, spent ages updating GivTCP on docker, restarting routers etc

I thought I'd take a look at my AIO (no predbat on it - only use Inverter Monitor to switch on/off discharge) and I had a "funny" although at 9:30 this morning...

Vestas learn something new everyday.

The Chinese calendar has leap years featuring leap months - rather than leap days as in the Gregorian.

https://www.dw.com/en/no-perfect-calendar-why-we-have-leap-years-and-why-the-earths-always-out-of-sync/a-19077064

Robgy Indeed, making me wonder if 1/13/xx is some sort of "correction date placeholder" when the product is used in China and its never been noticed until now in its various "western" incarnations - GE being one of them.

It'd be interesting to see what happens if the inverter was left to rollover from 1/13/25 - does it then show the "correct" date of 1/1/25?

Vestas It'd be interesting to see what happens if the inverter was left to rollover from 1/13/25 - does it then show the "correct" date of 1/1/25?

Almost certainly yes. The date&time is stored internally in binary. Just a conversion bug really. I haven't been able to test because I cannot set the time when the date is invalid. The conversion to binary has some validation!

For the next century, the validation will not allow the year to be set to zero. It will allow the year to be set to 100, but the date gets set to crazy values!

Rubikcube 2100 is not a leap year so that's a common bug. Or it could be the RTC in the dsp/cpu only storing year data as two digits.

I can't reset the inverter time either automatically or manually, it says there are items missing. Does anyone know a fix for this?

Phil If you are setting the time via the portal, check you have your time zone in account settings set to GMT 🤷‍♂️

Maybe change it to something else then change it back again. There is no option to set the time in the portal, only to sync the portal time (and timezone) to the inverter

hoggy

Oh that’s brilliant.. month 13 😂😂

kram unlucky for some 😁

Considered lucky in some places including China.

Again this morning my HA was unable to talk to the inverter! Looks like it went unreachable at 8:30am.
I tried restarting HA, no luck. Then resetting the time in the portal but that failed to talk to the inverter too. Resetting to defaults from the portal also failed.

Restarting the inverter from the portal, then resetting the time from the portal, then rebooting HA seemed the only sequence of events to bring everything back. worth a try if others are stuck

Hopeful No problems with mine at all today, it went through midnight with no rollover issues, then at 1am we had a power cut. When I restarted the HA server this morning it all connected up fine.

TX200

I’m just amazed month 13 made it vast design verification testing.

Just been looking at the octopus app and energy cost per day from 25th to end of 31st December 2024 when the issue happened.

It says it cost me in import only (give or take a few pennies)
25th - £3
26th - £3
27th - £3
28th - £3
29th - £3
30th - £2
31st - £5

So because of something or someone messing with the time on my inverters between 12 & 1am, until I fixed it at Midday it’s cost me £2. Can we all ask GE for a refund? Hahahaha

Do we know why this happened, I think we know what.

I was also wondering if GE fixed the issue remotely? What about those not in these forums or away?

Daveb01 Do we know why this happened, I think we know what.

I was also wondering if GE fixed the issue remotely?

https://community.givenergy.cloud/d/5448-a-number-of-inverter-clocks-are-reporting-strange-daymonth

Daveb01 My rough import costs are very like yours with about £3 from 25th–29th with £2 on 30th.

My 31st import cost was about £2 and that was the day my inverter was out of action for over 20 hours from midnight till 8:30 pm when I was on the grid on some very low overnight rates but also the later 4-7pm rates!
Just before midnight on 31st Dec, my batteries were both empty in anticipation of getting filled from midnight 31st Dec.

I don't think I suffered unduly by what happened and was sheltered by low grid rates early on and getting my inverter back on in time to collect the really low and plunge rates 31st Dec-1st Jan. My batteries were then replenished on negative rates.

Unfortunately, GE do not bill us, Octopus does on the evidence of our external meters.
I think givTCP API was released in 2021, in which case 2024 would have been the first leap year to contend with. This looks like a software glitch, waiting to be unleashed on the last day of the first leap year.
I hope that GE do come back to us to explain exactly what happened and what they are doing to prevent a repletion in 2028.

I think I was very lucky to have escaped what could have been very costly had it happened when Agile rates were through the roof.

Rob

@TheDragon (GivEnergy) Will we see a patched firmware version for ALL the GivEnergy inverter product range that is affected by the leap-year date bug?

geoffreycoan We could do with replacement dongles which installers come out an fit.

I have a load of devices on my wifi network and the GE inverter is the only one that is on the 2.4 gHz only 'IoT' network. We shouldn't have to turn off roaming on mesh networks in an attempt to generate a reliable wifi connection to the dongle.

When I changed my old BT router (which only provided shared 2 gHz and 5 gHz) to a Mesh network, the GE AC3.0 inverter was the only device that didn't pick up the new IP address assigned to it. And I was using my old BT SSID and password. To get the dongle connected requires doing the dongle dance.

Rob

Rbor Agree that there should be a newer dongle, especially one that supports newer security protocols. But with your AC3 and my Gen 1 hybrid, both now end of life despite their age (mine are less than 2 years old), I fear that’s unlikely.

A lot of IOT devices are 2.4GHz only. All of my Shelly, Tasmota, ESP32, Tapo and the Octopus mini are on the 2.4 mesh, 25 of the 28 devices currently connected to my WiFi.

Looking on my Deco app, I have 30 clients currently online and a further 20 offline!
Many could be on the 2.4 gHz IoT network but only the GE inverter is there at the moment.

GE have sold a lot of their low voltage batteries of various sizes, 9.5, 8.2, 5.2, 2.6.
The AC coupled inverters are a real issue with their maximum throughput of 3 kW or less. Hybrid inverters serving these batteries are available up to 5 KW. And all the inverters seem to need a dongle of some type, whether external or internal. These are still being sold and there are a lot of mesh systems around.

My AC3.0 doesn't have a LAN option.
My Solaredge PV inverter installed in 2022 has LAN or wifi options. My installer used the 'easy' wifi route.

Overall the GE kit does its job especially with predbat and HA.

But are we being hung out to dry, left with old technology worth a fraction of what we paid for it?

Rob

With a Google nest mesh system the dongle works fine. It seems to be faulty routers that try to send 2.4GHz only devices to a 5GHz connection that fail.

nophead With a Google nest mesh system the dongle works fine. It seems to be faulty routers that try to send 2.4GHz only devices to a 5GHz connection that fail.

My understanding is that the newer routers have a '5GHz preferred' protocol/standard so if a device tries to connect on the 2.4 network the router directs them to the 5 network. On a modern device this is fine but older and 2.4-only devices such as IOT and our inverter dongles, it all gets stuck.

It is what it is though. We can comment here but I doubt it will change

geoffreycoan I agree. Many provided routers have a joint 2.4 and 5 GHz option and it is not possible to turn off the 5 GHz frequency. The dongle isn't capable of connecting to 5 GHz.
The firmware on some older devices seems to baulk when confronted by 5 GHz.
Having a router with a 2.4 GHz only option is a huge advantage for making contact with 'older' devices.

Rob

geoffreycoan

I have a UniFi network system & that is the default direct devices to 5GHZ, however there is a button you can click to disable that.

I have found that most devices that can connect to 5GHZ do anyway. Since unticking the setting my network has been much better.

There’s a sound reason why 2.4ghz is preferred over 5ghz for IoT stuff, typically lower bandwidth devices in places typically furthest away from a router. 2.4ghz gives a more reliable signal and typically travels further. This cheatsheet is a useful guide:

https://www.oxitsolutions.co.uk/blog/ieee-wifi-standards-cheat-sheet

Just wondering if anyone heard anything from GE regarding the root cause and how are they planning to fix it and prevent it from happening again in the future?
The https://community.givenergy.cloud/d/5448-a-number-of-inverter-clocks-are-reporting-strange-daymonth is rather vague and not accurate (I'm not do not use PredBat but my inverter was affected).

alfwro
Wouldn't hold your breath and just take it on the chin and accept it as the way this company is going downhill rapidly ......

Northwarks Yes 5GHz is no good in our house because it doesn't go through internal walls well at all. The only devices out of about 50 that I own that support 5GHz are phones, tablets, an RPi5 and our newest TV. Before I had the Google mesh system I simply switched it off as nothing I do needs more bandwidth but I can't turn it off on the Google mesh and I can't give it a different SSID. Despite that everything works except the TV as the signal wasn't strong enough but it insisted on connecting at 5GHz despite the 2.4GHz signal being good enough. I had to install another Google point about 2m away just for that TV.

2.4GHz devices never see 5GHz signals, so if they don't connect it the fault of the router that assumes they can, not the devices themselves.

Hi all, if you have HA etc, I have found the sensor and put it in my home card.

I can now see the date and time it last sync’ed, so if it’s out I will know.

For me with 2 x AIO’s, go to devices and in MQTT find your inverter serial number.

Then click Inverter Time

Then click the Cog wheel

Copy the Entity ID (the two pages icon on right), then paste it into your chosen card/graph.
My Home Card now looks like this 👍

Daveb01 I check the inverter time (entity name though is spelt “invertor_time”) as a quick check that GivTCP is polling the inverter.

This particular bug though the date was corrupt on the inverter and GivTCP recognised it was corrupt and tried to replace it with the current time, but failed. Not sure what will happen when the ‘replace it with the current time’ bug is fixed, it might mean we don’t see these such issues any more as givtcp will mask them

https://spacenews.com/eutelsat-resolves-oneweb-leap-year-software-glitch-after-two-day-outage/

Wonder if they also use Chinese date formats 🤣

I had the same issue and the time resync solved it in part. The issue now (maybe unrelated?) is that Home Assistant cannot access the API. When I look in the portal 'Remote Control History' this also times out so I guess it is a GivEnergy issue. I manage a second system for a family member and this is working fine. Any ideas?

LukeG when you say “home assistant cannot access the API”, what do you mean?

It might be this issue that someone else has reported on givtcp https://github.com/britkat1980/giv_tcp/issues/323

Timeouts on remote control in the givenergy portal, I get them sometimes. I put it down to the portal not being able to communicate to the inverter in a timely manner. Usually a retry or two and it works. If it happens all the time then maybe look at the inverter wifi signal strength (is it good enough) and whether there is anything on the router firewall that might be blocking the request.