Slow start in the morning, is this a (known) MPPT issue ?

I had my hybrid 5.0 Gen2 installed D0.906-A0.906 about a month ago, and so I am just getting used to what is "normal".

I have 2 strings, south and west.

What I am seeing looks like the inverter doesnt track properly, at least first thing. As the sun comes up it starts to produce, then drops back to "noise" and eventually starts to produce significantly. I thought at first it was cloud/shadow ( shouldn't be any), but early days.

However: I have now had a couple of blue sky days, and today is uniform grey.

An example of the pattern I am seeing: this is where the system " self recovered". Its not always the same time, and not at the same light levels as far as I can see. This was 3rd April, blue skys. Homeassistant graph at 10 sec samples.

Having noticed the pattern I tried restarting the inverter on the following day, Blue skys, 4th april Instant recovery.

And today, grey skys, so much lower production, again reset point is obvious.

This is obviously not supposed to happen, but is it one of the issues being worked on in the beta firmwares ?

Thanks.

I believe this is a known issue and being worked on but maybe @TheDragon (GivEnergy) can confirm please. I am due a gen 2 retrofit next week (2 strings) and this is an issue I'm also keen to know about.

Known and fixed for Gen2. Current test release D0996-A0.908
Gen1 98% good.

TheDragon (GivEnergy) Excellent, thanks for the update

Thanks for the reply, and good news.

TheDragon (GivEnergy) How do you get access to the beta firmware ?

TheDragon (GivEnergy) I too see this characteristic of a slow start on sunny days but seems ok on shady days ?? system is Gen 1 3.6 you mention a firmware update I also would like to know how to get access to this firmware to resolve this problem, can you help please

Just a heads up for anyone the mppt behaviour affected.
909 did indeed get rid of this effect, but it seems to have reappeared in 911.

Its all pretty academic, mostly for interest.

Seems to recover earlier than 906 which were the pictures above, but that was April, so cant really say if it is exactly the same as 906 or not. The pickup at 6:20 was when my automation restarted the inverter based on an automation monitoring higher string voltages, That kicks the mppt.

Characteristic seems to be that the string voltage gets higher than normal before the current and therefore power is increased. You can see the differnce between 909 (left) and 911 (right)

911 does fix the battery charge discharge pattern when PV hits inverter limit, AND battery is full.
Hopefully 9xx will fix both :-)

Interesting that it was fixed for you on 909 because I'm on that currently and still have a slow startup before the string jumps to _140W and it settles into what I would expect. Voltage has climbed above normal during the slow startup as well. My graphs look very similar to your 911 ones (Don't have access to my HA instance on this PC to post them).

Phil_H its so unique to each person, 99% see no issues
but some see the stuck startup.
cant even find a trend based on panels brand, or optimizers

TheDragon (GivEnergy) Happy to run any tests that will help. From previous threads it did seem to be linked to multiple aspect strings, but might be coincidence. For the record, 11 south, 6 west 420w JA Panels, no optimisers ( sadly ...)

Phil_H
For me its really obvious.

906 was bad, ( 9am start in april ! )
909 instantly made it smooth as silk,, power from 5:30 at start of june giving 150-200W to cover base load.
911 is 7 o'clock start in July.

909 showing lovely clean start in the morning, but battery charge cycling when Inverter hits maximum.
and zoomed in to same scale as above:

I need to look back at my data as I’ve noticed the last couple of days that there is some solar quite early in the morning - if the inverter was already on from having some charge left I could see an earlier startup - but if it had shutdown if the battery was flat it would take a bit of getting going - and I have 2 strings on different orientations.
(following a recent firmware update)

TheDragon (GivEnergy) The 911 firmware has noticeably different behaviour on MPPT startup than 909 but with lots of clouds could not be definitive until today when we've had clear blue skies. 909 power oscillated a little when below 100w per string on startup but fairly consistently showed an upwards trend until it settled down. 911 oscillations are more extreme more and spends a lot more time at 0 output.
This morning I restarted the inverter at 7.20ish which seemed to cause it to settle into more normal behaviour.

Looking back at 909, one noticeable change is that the magnitude of the voltage oscillation seems to have increased significantly in 911 (from c.50V to >100V) so wonder if that is a symptom or the cause of the changed behaviour. 909 example graphs below

Sting 1 is 7 panels south facing without optimisers. String 2 is 9 panels 3 south, 3 east and 3 west facing - all with Tigo optimisers. The behaviour is pretty similar between the two strings, so it seems unlikely orientation or optimisers are the cause.

Happy to run any tests that would help.

Dw Thanks for the confirmation its not just me! Exactly same behaviour as my system ( all above) The MPPT starts to apply load, then gives up and you get high voltage, no current load until something triggers it, and we are back to normal. The restart trick works, (906 and 911). To me, 911 start up behaviour is EXACTLY like 906, swapping battery when full issues of 909 for the slow start.

From a firmware perspective it should be really obvious what was changed, and what was reverted. Any chance of going opensource :-)

Just to add another data point, I have observed the MPPT startup (and shutdown) oscillations with GEN2 906, it has not been resolved with 909, this is an example for this morning

System outline:
GEN2 Hybrid FW909
Array 1: 9x 380W JA Solar panels 170 dge orientation (180 dge = south), 35 deg pitch, Tigo optimisers fitted 4x panels, min shading
Array 2: 6x 320W JA Solar panels 170 dge orientation (180 dge = south), 5 deg pitch, Tigo optimisers fitted 4x panels, high shading early morning.

Startup oscillations observed every day, but not clear why the inverter suddenly start to behave.

The GEN1 hybrid with FW 451 at the same location with the same array did not show the issue.

athresuk Just checked back on this thread as I had a particularly poor result today, was waiting for the full days print before I posted, as it looks like a blue skys all the way day.
Very interesting you say 909 didn't fix for you, as my behaviour changed dramatically on the day 909 was installed, and bad again when 911 was. I think I will ask to be revertedd in a couple of days, just to check the theory. Be interesting if it is register based issue that is coincident with an upgrade, but my first thought it is the firmware version.
My setup is 11 south, 6 west, so big discrepancy in the two strings at startup.

Isn't 911 the newest version on the portal ? Have you tried that ?

Hi

I believe I experienced something like this today, had system installed late July and all working well. This morning at 7 ish, sun usually starts to charge batts but when I checked at 9ish my panels were showing like 90w generation (not even a slow ramp up) and it was only when I rebooted my inverter did it then jump up to the usual 1.5KW.

the inverter PV stats at the time show this:

2023-08-10 09:42:19 138.5 151.6 10.7 0.6 1518 90 1608
2023-08-10 09:42:09 141.5 143 10.7 0.6 1517 89 1606
2023-08-10 09:39:07 172.9 169.5 0 0 4 2 6
2023-08-10 09:38:44 172.9 169.5 0 0 2 33 35
2023-08-10 09:38:33 172.9 169.1 0.5 0.4 75 1 76
2023-08-10 09:38:22 173 169.5 0 0 1 62 63
2023-08-10 09:38:08 172.6 162.3 0.4 0 114 1 115
2023-08-10 09:37:39 172.6 163.3 0.6 0.5 86 1 87
2023-08-10 09:37:26 172.8 169.5 0 0 2 70 72
2023-08-10 09:35:29 172.9 169.5 0 0 1 19 20
2023-08-10 09:34:02 173 169.5 0 0 1 85 86
2023-08-10 09:33:49 172.7 166.1 0.3 0.2 48 1 49

Inverter Data for the time shows:
2023-08-10 09:43:07 368 40.6 1.6 247.4 50.14 0
2023-08-10 09:42:59 359 40.9 1.6 247.4 50.04 0
2023-08-10 09:42:56 364 40.8 1.7 247.4 50.06 0
2023-08-10 09:42:49 366 40.8 1.7 247.5 50.05 0
2023-08-10 09:42:45 359 40.8 1.6 247.5 50.12 0
2023-08-10 09:42:39 394 40.7 1.8 247.6 50.05 0
2023-08-10 09:42:34 357 40.7 1.6 247.4 50.16 0
2023-08-10 09:42:29 365 40.7 1.7 247.3 50.14 0
2023-08-10 09:42:23 354 40.6 1.6 247.5 50.05 0
2023-08-10 09:42:19 356 40.5 1.7 247.5 50.03 0
2023-08-10 09:42:09 358 40.3 1.6 247.3 50.03 0
2023-08-10 09:39:07 -54 40.4 0.6 246.9 50.1 0
2023-08-10 09:38:44 -52 40.2 0.6 246.9 50.12 0
2023-08-10 09:38:33 -44 40.2 0.7 247.1 50.09 0
2023-08-10 09:38:22 -18 40.2 0.7 247.1 50.12 0
2023-08-10 09:38:08 64 40.4 0.8 246.8 50.1 0
2023-08-10 09:37:39 -24 40.2 0.9 246.9 50.09 0
2023-08-10 09:37:26 -15 40.3 0.7 246.8 50.11 0
2023-08-10 09:35:29 -67 40.3 0.6 246.8 50.05 0
2023-08-10 09:34:02 -17 40.2 0.7 246.5 49.96 0
2023-08-10 09:33:49 -37 40.4 1.2 246.9 50 0

My firmware for inverter was
D0.303-A0.202
My Batt Firmware was
3012

I checked and there was an update so i applied this so now im on

D0.305-A0.305 & 3013

I have the following setup (supplied from ocotopus)
2 strings (5 panels on each)
Inverter Model GIV-HY3.6
Batt GIv 102 Ah | 5.22 kWh | (1 Pack)

I will see what tomorrow is like, maybe just a one off 🙂

justpassing The plot I posted above was with 909 .........

I haven't tried 911 yet as the portal is trying to force a BMS 3013 update which I don't want to do.
This is an issue with the current portal FW update implementation, you can't selectively pick which FW update to carry out when there is more than one pending.

Morning All

Sadly I had the issue again this morning, making it the 2nd time since system installed. Sunny day and panels were showing only generating like 100w.....rebooted inverter and now its back to 1.6kw.... a call to support maybe?

athresuk I reverted back to 909 , from 911, and back to smooth as silk startups. So for me at least with a south string and west string, there is clear improvement on 909. My strings have very different early morning incident light, particularly on clear days, with reduced scatter. Have you looked at the current (A) graph for your two strings ? On 911 I was seeing the start of the south string being held back until the west could maintain 0.6 A. 909 I don't see that at all. I dont have optimisers, though I wish I had on the west string for evening shade :-)

Hi

Getting my head around this, so are we thinking that on 2 string set ups, there might be an issue with one strings current (A) being low?

DGinUK logged with support who are downgrading my inverter FW , Im currently on D0.305-A0.305, will keep you posted

justpassing

For my install 909 is as bad as 906.
My troublesome install also has 2x strings with wildly different irradiance levels first thing in the morning,
MPPT#1 array 1 is on an attached garage south facing at 35 dge pitch with minimal shading
MPPT#2 array 2 is on a workshop flat roof at 5 dge pitch, heavy shading from a neighboring tree.

Due to the oscillating nature of the PV curves it is very difficult to get definitive numbers on when things improve, however the MPPT#2 generally needs > 0.4A before the system start to behave and the allows the correct operation on MPPT#1
This is an example this morning:

This obviously isn't a fundamental HW issue as an inverter reset or change of FW appears to reduce the effects of the issue.
I guess I could take a trip to site and reverse the MPPT channels over to see if this has any effect.

DGinUK To be clear, what I see is that with very different strings, both strings are suppressed until the lowest producing string hits a particular value. Once the lower current string hits a magic number, then they both produce normally.

athresuk Just FYI, my west string is MPPT#1, so its that that needed to hit 0.6 before it started. So swapping the strings is not the obvious answer.

downgraded to firmware now D0.303-A0.303 via support..... 😃

Morning all - Just to add my experiances,
I have 12 panels - string 1 is 5 panels on sw side with Tigo's on each. String 2 is 7 panels as 3 on NE and 5 on NW, again tigo's on each panel. All on to a Gen2 5kw inverter.

I was put onto the 909? Beta to resolve this issue, which worked for me over the main summer sunny days.
I recently updated to D0.911-A0.911, which had been working fine until the sun actually came out (lots of cloudy days). Over the last two days I have been seeing this error, on Monday I tried the software restart via givtcp, when it came back online, still had the same issue, did via the web, same outcome, and in the end I had to physically power down and back up the inverter, which has then resolved. On Tuesday, just a restart from givtcp has resolved the issue. Today, at 7:35 it kicked itself into life.

What I can see with the GivTCP stats is that PV production on string 1 and 2 jump between each other - almost as if they are feeding each other and there isnt enough power (current specifically) to engage and start generation on either string?

For any Giv Staff - happy to be a Beta tester and also provide additional diagnostics as required.

Had same issue again this morning after downgrade yesterday, Support have passed the case over to head of firmware development ..... hopefully cause can be found, is there a way to schedule inverter reboot ? will keep you posted

DGinUK is there a way to schedule inverter reboot

Set a schedule on HA on your RPi?

Tim yeah, Good idea, I was looking at HA today after my post, was reading about GIVTCP I thought about that until a fix is found....

unclebob This explains the behaviour of mine this morning. I rebooted the inverter to fix it. Quite glad others have the same issue.

unclebob I have reverted 909, which doesn't have the slow start of 906 or 911. For me at least. 909 has the oscillating battery charge when full issue. I have set up HA to set battery pause charge when it's full. That stops the oscillations, while still allowing dynamic supply to the house load. I prefer using ha to stop the battery charge issue, rather than restarting the inverter to kick the mppt. YMMV 😊

+1 for the regression to slow start post 911 upgrade. 909 worked perfectly, 911 I've had to restart twice already today - particularly poor PV.

Can see the voltage goes passed a threshold and the power drops to nothing, restart and it works. @TheDragon (GivEnergy) is this already captured on the open topics for the next firmware release?

Arg0t ok, i've reverted you to 909

While we resolve the PV starting issues in 911

Issue of a beta test for this will be via the formal Beta test group now setup, adn invited to only the official beta testers admitted.
This prevents the chaff from swamping the issues,

TheDragon (GivEnergy) roll back did not fix the issue, to make things worse I now see weird behaviour on string 2 also.

String one hits 440v and as always this results in power outage dropping to nothing, but interestingly it corresponds to string two having a higher working range and 220v ceiling and larger amount of voltage oscillation.

String two power is erratic, string one is flat lined until the startup routine somehow causes it to take control and normalise both strings.

Does the BMS have any influence on inverter decisions? Wondering if I need to get the batteries downgraded also.
Not sure what the firmware upgrade does, if that only changes certain blocks on the device or writes to the entire range. I wonder if there are blocks written by 911 that 909 doesn’t touch and the downgrade leaves some remnants of 911. Is there some kind of default reset you can perform from the admin side?

Any other suggestions?

Arg0t for reference 3013 on battery and 909 works for me. I also did 909 -911-909 and I am back to the original 909 improv.ment

Arg0t Is there any possibility the inverter thinks you've exceeded the maximum string voltage on string 1?

Looks to me like that, even though you're well below the limit. Edit - as to what "resets" the "tripped" threshold who knows? Reboot works as you said...

TheDragon (GivEnergy) any update with how this Beta is going, im having to reboot my inverter daily with all this sun.

Many Thanks

DGinUK we have a new verion in test, D910.

A blend of 909 & 911, zero issues found. However we are waiting for the 500v+ issue to be authorised to be coded in, then D912 will be created

TheDragon (GivEnergy) Paul, does that mean that you think you have a fix for the issue of production stopping if the string voltage exceeds 500v on Gen 3? As I'm planning to add panels onto our extension in around a month (once it's up), this will directly affect me if we simply add the panels to one string rather than redistributing both (which would require access to the original roof which the scaffolding won't extend to).

still waiting for a solution, any update?

Any update with this? still getting this issue

Any update with this? still getting this issue even after GiveEnergy Swapped out my inverter .....

Still waiting....ever get the impression givenergy don't actually giveafek..... I've had calls with support about batt firmware failing, inverters swapped out by givenergy and its still not any better
.....like I said to last tech, I'm happy to remove the inverter for a refund and it's clearly not working.......unless you can get the firmware?

I Have this problem. Givenergy don’t seem to have anyone working for them who actually understand their own design. Otherwise this problem would have been fixed instead of trying to patch around it with firmware.
It is upsetting to see a beautiful blue sky and not notice till mid morning that there is virtually no power being harnessed.

12 x 430W panels East, 6 x 430 panels West, 5.0 Hybrid Gen 3 + 9.5KWh battery.

Oldgit yep....im pushing support but its feels like a lost cause, 1st line response when I chased about my existing call....i liked the "......could not locate it haha"

as no response as of yet......oh like your good self, its sunny and missing that sun as per normal!

Mine was doing the same, if you restarted the inverter it would go from 80w to 800w now on firmware 912 and not had the issue. 5kw inverter gen 2 27.2kw batteries and 20 400w panels 10 SE, 10 NW all with Tigos and CCA.

not sure that firmware is for my inverter??

Inverter Model
GIV-HY3.6
Software Version
D0.393-A0.393

paulwindy FW912 doesn't always fix this issue with GEN2 inverters,
I have a site with a GEN2 and split arrays running Tigos which 912 made no difference to slow MPPT start-up, exactly the same as 909.

There is FW915 which specifically address this issue for GEN2, not had a chance to update the GEN2 to this yet to verify it is now working.

athresuk Is this the make of panel that you have as there was an issue with certain make of panel. Mine are Eurener, think it could have been Jinko or a certain model of Eurener's ? cant remember now

DGinUK This is mine from today ! maybe its still doing it ?

I have this exact same issue. Gen 3 inverter installed this week. Restart sorts the issue but can’t be doing this every day

As with anything GE there's lots of iffs, buts & maybe across Gens and Firmwares...so this odd behaviour is fixed, so long as you don't have the newest Gen 3 gear (which is a bit weird)

So yes, you are stuck with the restart until some mystery future date when I guess it gets fixed?

Hi

My PV panels installed by Octopus, 2 strings East and West and no bypass diodes or conditioners on 🙂

JAM54S31 AB Half-Cell
4kW total of Solar Power
10 x JAM54S31-400/MR 400 Watt panels
12 Year Product Warranty & 25 Year Linear Performance Warranty