as continuation of an other tread i start here a new thread for the Axpert Max 8000 which has the newer DSP Processor.

People which have low PV production issues with this Inverter and search for solutions will initially not know which DSP Processor they even have.
So, if your Main Firmware is 72.xx or 92.xx ex works its likely a model with the newer 28066 DSP Processor.

My Inverters are factory date 05/2022 and seems to be the early ones with new DSP Processor.

Some Inverters have issies with low PV production as MPPT sticks at around 90 Volt and i have additional AC Out losses and F80 Errors in as 3Phase parallel configuration

Already ongoing modifications and optimizations of the 92.04 Version did not solve my issues yet.

I got now original Firmware 72.62 HV
Affected users could try this Version and the technicians here could compare changes of this newer 72.62 HV with 92.04

Here is the download. Use at your own risk! VERIFY that you have the correct Inverter Model before trying this Firmware.

MAX_HV8k 72.62.7z

I have it running now for a day, panels still covered with snow.
 

Up to now system was stable, the MPPTs did not stick at 90v
the MPPT kept voltage on 120v and increased amps

PV Production and peak watt is only a bit less than yesterday where i had installed patched Firmware Version 92.06
i had (yet) no AC Out losses and no resulting F80 Errors

Next days will become warmer and more sunny. so the snow will melt down from the panels and i will be able to see the behavior with more watts

but hey, already 30 hours without Can Bus error F80

 

9 hours ago, mjmx8k said:

Up to now system was stable, the MPPTs did not stick at 90v
the MPPT kept voltage on 120v and increased amps

That sure looks like the minimum MPPT voltage is set to 120 V; the MPPT voltage graph is ruler flat apart from a very few very brief excursions. So I had another look in the new firmware (version 72.62). I didn't realise that in the function that checks various EEPROM values for sanity (the minimum MPPT voltage is stored in EEPROM), it checks the minimum MPPT voltage, and sets it to 120.0 V, if it was anything other than 120.0 V! So the effective minimum MPPT voltage is indeed 120.0 V for this firmware. My apologies for not noticing this. I've edited my earlier post to this effect.

But I note that it's very possible that the maximum power point for these panels, even under a low of snow, is more than 120 V, so this isn't a complete solution, it's just giving back 30 of the "missing" volts, which could be a hundred or more volts.

It would be interesting to see this firmware with the "solar" plug on the main power supply unplugged, if possible. But certainly, it's worth establishing a base of a few days of "factory" ("as intended"?) behaviour.

Quote

but hey, already 30 hours without Can Bus error F80

That's possibly because there were no sudden jerks of PV voltage, being ruler flat. In other words, it could be that the "stuck at 90/120 V" bug is actually masking the F80 bug or hardware fault. That agrees with the observation that patched firmware seems to exacerbate the F80 problem. This is highly speculative, of course.

Edited December 22, 20222 yr by Coulomb

2 hours ago, Coulomb said:

It would be interesting to see this firmware with the "solar" plug on the main power supply unplugged

Helper PCB Solar cable is unplugged on all 3 inverters

 

Yes, panels could do much more volts under snow (220 and 250 depending on the string) but i guess this firmware tries first to go up to 0,5amps and then change voltage to next level. (just a guess as we know that in original firmwares is set the 0,5a value)

thats the question - will it stick at 120v or not 🙂

production in total watts with 92.06 (20.12) and 72.62 HV (21.12) was very close in very similar weather,
start time and shutdown time was only 2 minutes difference.

now snow is melting, so i will have a chance to see what will happen with the 120v ruler

 

2 hours ago, Coulomb said:

apart from a very few very brief excursions

this breakout from the ruler at 13.00 is a result from grid off/on switching

 

interesting phenomen:
3P3 MPPT2 (other make of panels) do 60v close to midnight

i was not onsite to measure with multimeter, so i dont know if its real volts or an inverter measuring error

 

Startup nice today and survived without F80

Another positive news:
- the EEPROM fault messages seems to have disappeared.
- Dessmonitor now show all 3 inverters as green (Normal) instead of orange (Warning)

Lets hope this Firmware now does not make the same MPPT nonsene @120v as the MKS4 model

 

yes! breakout from 120v after reaches 0,5a without interaction

In the user manual is written that in 3Phase Parallel configuration a longer transfer time will occur.

With the 72.62 HV now Inverter power switching (grid off, Inverter go to battery mode) is now in my 3Phase Parallel Configuration much faster.
So fast, that my Bypass Automatic Transfer Switch (50ms) do not recognize it any more

this is a huge improvement for me.

At the power switch at 09:56 MPPT1 voltage drops, MPPT2 not, so they seem to behave differently

not nice, look at Inverter2 MPP

think at 10:34 snow was slipping from upper to lower panels.

but INV2 MPPT2 went down to 120v and high 2,2amps.
MPPT should increase voltage, but does not, loosing out watts here...

grid on/off settled the volt situation... but watts stay the same.

Edited December 22, 20222 yr by mjmx8k

inf fact, up to now 72.62 HV has solved a lot if issues for my setup:

- EEPROM fault messages gone
- AC Out losses gone
- F80 Faults gone
- Faster power switching
- Sudden spikes are gone or treated differently

so i can consider the system basically stable on the AC output side.

about the MPPT:
the strings with the 2 Parallel work not that bad out of the box. it sticks at 120v until 0,5amps and then go up
the strings with the 1 Parallel work bad it sticks to 120v and bring about 25% less watts as should (cloudy)

so there is a lot room for working better.

@Coulomb

would it be possible to apply the known patches with 200v to the Version 72.62 HV
in first step i would install it only in my 3P3 only with the 1 parallel strings
in a second step on the 3P2 to have a direct comparison to 3P1 as they have absolute identical panels

 

 

4 hours ago, mjmx8k said:

would it be possible to apply the known patches with 200v to the Version 72.62 HV

Yes. But I think I'll sort out the issues with @Volker first. He has ten PV strings to your three 😉 Hopefully we'll have a reasonable consensus as to what the best patches are soon.

Had 1 F80 yesterday in the morning because MPPT of 3P3 stuck and than spiked.

@Coulomb
i think the only MPPT related thing on Version 72.62 HV would be to decrease the 0,5amps threshold to 0,25amps
i am sure then it works much more nicely and it does not need to change the startup voltage at all in this version
 

18 hours ago, mjmx8k said:

i think the only MPPT related thing on Version 72.62 HV would be to decrease the 0,5amps threshold to 0,25amps

You're easily pleased!

This is patched firmware version 172.62 for the Axpert Max 1 8kW with '28066 processor and Arena reflash tool. Based on factory firmware version 72.62.

It has just one of Georg's changes to prevent getting "stuck" near 90V PV voltage (threshold current 0.50A -> 0.25A)
Same for MPPT2.
Also fixes the premature float bug, because it's easy and I hate that bug.

Do NOT use with other models, even the same model with the older '2809 chip.

Use at your own risk.

Edit: Attachment deleted due to blank lines left in the hex file. See below for corrected version.

 

Edited December 28, 20222 yr by Coulomb

2 hours ago, mjmx8k said:

Had 1 F80 yesterday in the morning because MPPT of 3P3 stuck and than spiked.

I still find it hard to believe that these things are related, but let's see if the MPPT voltage changes more smoothly, and if that prevents the F80 faults.

Edited December 27, 20222 yr by Coulomb

Thank you a lot. will keep you updated

@Coulomb
is possibly gone something wrong with version 172.62?
i get only error 32 with this version.

12 hours ago, mjmx8k said:

is possibly gone something wrong with version 172.62?

I did have a lot of stuffups this time around, but I thought I recovered them all.

I'll do a full compare to make sure.

16 minutes ago, Coulomb said:

I'll do a full compare to make sure.

Sigh. I had left some blank lines in the hex file after one of the stuffups. IDA Pro didn't mind, but evidently the reflash tool does. Here is the corrected version. I'll leave the same version number, as yours was the only download, and there can be no question whether you have the right version or not 😮

Sorry for the hassle; I hope it wasn't hard replacing with a good firmware.

This is patched firmware version 172.62 for the Axpert Max 1 8kW with '28066 processor and Arena reflash tool. Based on factory firmware version 72.62.

It has just one of Georg's changes to prevent getting "stuck" near 90V PV voltage (threshold current 0.50A -> 0.25A)
Same for MPPT2.
Also fixes the premature float bug, because it's easy and I hate that bug.

Do NOT use with other models, even the same model with the older '2809 chip.

Use at your own risk.

 

Main DSP 172.62 - Patched.zip

Edited December 28, 20222 yr by Coulomb

Thank you. will install it later today.

i still do not understand the correct behavior of "Startup Voltage".

When i look on the MKS4 which has a Startup Voltage of 120 it do NOT pruduce ANY watt before panels get to 120.x volt
When i look on the MAX8k which has a Startup Voltage of 90 /120 or patched value it DOES start to make watts after it reaches 75v

2 different behaviors for the same word / trigger. but which one is doing correct? the MKS4 or the MAX8K

 

Edited December 28, 20222 yr by mjmx8k

@Coulomb

yes, installed well,
no, does not like it.
worked 1,5 hours nicely until exactly the moment MPPTs came down to 0,1 amps -> have F80

10 hours ago, mjmx8k said:

worked 1,5 hours nicely until exactly the moment MPPTs came down to 0,1 amps

So the MPPT is behaving well, high voltage until 0.1 A?

On 2022/12/29 at 3:20 AM, Coulomb said:

So the MPPT is behaving well, high voltage until 0.1 A?

0,25 as set in the FW.
it increases the voltage to >120 at 0,25a and it drops voltage again to 120 when comes down to 0,25a

at the moment it works stable, keeps the voltage up the whole day nicely

Full day:

Startup (today):

Shutdown (yesterday):

@Coulomb
@Volker

my MAX1 8K inverters in 3Phase Parallel are now stable since 14 days without any F80 and without any AC Out losses.
all EEPROM error messages are gone, all Inverters show up as green in DessMonitor

FW 172.62, Solar Cable disconnected from Helper PSU, Inverters set to SUB permanently.
Grid connection is controlled with a 63a Zigbee MCB based on PV Power and Pylontech SOC (HA) to avoid any possible grid feedback or consumption

I think 0,2A instead of 0,25 could gain some watts in the morning and the evening but i think this very small loss can be neglected and i prefer to thave no fault codes, never touch a running system.

btw, as you are watching out for MPPT1 and MPPT2 panel similarities in another thread:
Inverter 1 and 2 have all identical panels 6+6=12 300w on each MPPT (48 Panels)
Inverter 3 has 7x300w on MPPT1 and 8x195w on MPPT2

so i could do comparisons if you need.

thank you again for the FW 172.62
 

@Coulomb

mppt wise the 172.62 for the Max1 8k is still doing very well!

setting the parameter 03 from UPS to APL made the F80 issue again a bit more stable.
my finding is that the F80 appears when grid power and pv power are very nearby.
our grid has up to 12v difference over the 3 phases, so maybe this is a reason.

however, still did not get rid of the F80 issue completely, as a workaround i use now 3 inverters (3Phase) permanently on SBU (grid off) and the 4th inverter acts in powered off mode as a Battery charger only (no panels on it).  HA switch on INV4 Grid when SOC is low. with this setup there are no F80 anymore as the 3 inverters never swtich back to grid.

maybe not very efficient due to double conversion losses, but stable and no more AC Out flickering on changeover. a bit like the King models with 0 transfer time 🙂

still i have every day several errors 61 (Pylontech comm lost) beeping for a minute and then recovering automatically.

dont know if this originating from the Voltronic firmware or the pylontech firmware, but before touching the BMS firmware...

is it worth to try the 72.66 version?
if yes,

please could you do me a 72.66
- with threshold current 0.50A -> 0.2A for both MPPTs
- the premature float bug
- no changes in the startup voltage

2 hours ago, mjmx8k said:

is it worth to try the 72.66 version?

My notes say that 72.66 fixes a false intermittent EEPROM warning. If you're not seeing this, it's hard to say. I should try and compare them.

In my case the EEPROM fault message appears only on parameter change.
lets say you change AC Charge current from 70 to 50 amp.
then it does not save the new parameter value and makes EEPROM warning line in log
inverter reboot proves that new value not saved.
changing parameter again and it saves the parameter well without EEPROM warning.

just my findings.
 

On 2023/04/12 at 5:57 PM, mjmx8k said:

is it worth to try the 72.66 version?

 

On 2023/04/12 at 8:17 PM, Coulomb said:

I should try and compare them.

Ok, I've done some comparisons, and I find that 72.66 has significantly more logic in the MPPT control functions. It has a new pair of variables (one for each MPPT) which can be set to 4 different values depending on the solar current. 3 of those values are much higher than the default value. What the new variable does is very complex, it could take me weeks to sort it out.

In short, yes, this could be significant, although it doesn't look to me at this stage that it's Voltronic attempting to fix the "stuck at 90 V" problem.

So I'll look at patching it, for @mjmx8k and probably others.

17 hours ago, Coulomb said:

 

Ok, am făcut câteva comparații și constat că 72.66 are mult mai multă logică în funcțiile de control MPPT. Are o nouă pereche de variabile (una pentru fiecare MPPT) care pot fi setate la 4 valori diferite în funcție de curentul solar. 3 dintre aceste valori sunt mult mai mari decât valoarea implicită. Ceea ce face noua variabilă este foarte complex, mi-ar putea lua săptămâni să o rezolv.

Pe scurt, da, acest lucru ar putea fi semnificativ, deși nu mi se pare în acest stadiu că Voltronic încearcă să remedieze problema „blocat la 90 V”.

Așa că mă voi uita la patch-ul, pentru@mjmx8k si probabil si altele.

Hi Coulomb, i want to buy an Axpert Max I 8kw Twin, under ANENJI brand, i searched and is made by Voltronic like mpp, easun, do u have firmware for this inverter ?