Good day

I had a problem today with the MAX 7.2kW where it didn't want to charge the batteries at all.

It was a very cloudy day, so I changed it to SUB mode with the hope that it will charge the batteries as "first priority" (like my old VMiii did).

After no success with that, I changed the "solar only" charging to both other options but still it would only charge when there were excess sunlight (load was first priority)

So I did the right thing and attempted to load new firmware, by far the best decision this year.... (only 6 days at least)....

Original firmware taken down before disaster strike: U1: 45.07,  U2: 12.13,  U3: 00.00 (don't know what this U3 is for).

After turning the inverter into a brick, I eventually got 12.03 flashed and then the screen came on at least. But trying to get 45.04 or 90.19 loaded doesn't work at all.

I've tried through the coms cable and via USB stick, no luck.

When I power it up, it just goes through the normal countdown 6,5,4,4,3,2,1 and the 32 flashes. Tried with batteries only and AC only, makes no difference...

It "writes" 2 blocks and stops as per attached photo.

Also, with the old (working) firmware, I did notice that it backfed into the grid on SUB setting... Another reason why I though it would be brilliant to load newer firmware.

Any help would be much appreciated 😀

> Axpert MAX 7.2kW (Synapse 7.2M+)

I don't know what to think of Synapses. According to my notes, I now believe that most of the recently manufactured models are genuine. I suspect that earlier versions were clones.

On 2022/01/07 at 2:26 AM, Jopie said:

Original firmware taken down before disaster strike: U1: 45.07,  U2: 12.13,  U3: 00.00 (don't know what this U3 is for).

Ironically, 45.07 is the latest firmware that  I know of for Axpert MAX 7200 W models.

Removable display firmware version 12.13 is the latest I have (as of this post).

U3 is for the Bluetooth firmware, I believe.

On 2022/01/07 at 2:26 AM, Jopie said:

When I power it up, it just goes through the normal countdown 6,5,4,4,3,2,1 and the 32 flashes.

Fault code 32 is a lack of communications between the DSP (main inverter processor) and the removable display. In other words, that's what I expect to see if the DSP is dead or bricked.

On 2022/01/07 at 2:26 AM, Jopie said:

It "writes" 2 blocks and stops as per attached photo.

That's strange. Perhaps the flash memory in the DSP has somehow become damaged. You seem to know what you're doing with flashing, so I'd say you've just been unlucky.

It looks like you'll need a new control board. There are many versions of the control boards, with different processors, and I suspect many different subtle changes to the rest of the circuit. So you'll have to get exactly the right replacement one. It will also have to have the correct version of boostrap loader; a totally blank chip won't work at all. You'll have to start with your supplier, or if you're extremely lucky, you might be able to get a third party control board with all the right elements. But that's win the lottery type lucky.

Sorry to be the bearer of bad news.

Edited January 15, 20223 yr by Coulomb

Thanks, I surely hope it is not that bad, I only bought it less than a month ago🤢
I'll see what the supplier has to say, they sent me firmware via email, but can't open it....
I'll update once something good happens 😀
Thanks👍

@CoulombTurns out, it was no bad news at all...

I found an old desktop PC in the attic with an actual COM port. Still had XP loaded on it, just needed a power supply and voila, it worked.

Inverter is up and running again🥳
I loaded 90.19 and 12.06. First thing I noticed was it constantly back fed approx 200 watts.... So I just tripped the inverter's supply. No running smoothly from sun/batteries. Today is the first day in a couple of weeks where we have no clouds😀
I will wait and see what version the supplier can offer me, and I'll post feedback here.

Moral of the story... Never get rid of old computers, they might be handy in the future.

5 hours ago, Jopie said:

I found an old desktop PC in the attic with an actual COM port. Still had XP loaded on it, just needed a power supply and voila, it worked.

😃 I should have thought that: it looks like you have a dodgy / less compatible USB to serial adapter.

5 hours ago, Jopie said:

Inverter is up and running again 🥳

I'm glad my doom and gloom was wrong 🙃

5 hours ago, Jopie said:

I loaded 90.19 and 12.06.

I believe that 90.19 is moderately old (at least 6 months, probably a fair bit more). 12.06 is similarly fairly old. 45.07 (also available from the Czech site) is about 4.5 months later than 90.19. [ Edit 2: Actually, that's going by Last Modified Date; that might not indicate which one is actually the more recent. Ugh! ]  The Voltronic version numbering scheme is infuriating! They seem to have switched from 90.xx with 45.xx being the same firmware but when running on 120 V hardware, to now "unifying" the two version numbers to 45.yy (with yy being less than xx)!  [ Edit: But maybe they started out both being 45.xx and recently have split. That would actually be more consistent. Sigh. ] There is nothing like consistency, and these firmware versions are nothing like consistent!

5 hours ago, Jopie said:

I will wait and see what version the supplier can offer me, and I'll post feedback here.

That will be interesting to see. Though unfortunately it's not unknown (sadly) for firmware via the official supplier (which I believe has to come from engineers at Voltronic in Taiwan) to be many versions out of date. Sigh.

Edited January 9, 20223 yr by Coulomb
As annotated.

OK, this issue of which firmware version is later than which other version is really bothering me. So I set out to compare what I thought were the leading contenders, and lucked out: there was only 4 bytes different in just one function. That made it easy to compare. My conclusion seemed unlikely though: the newest function seemed to be the shorter one. Usually, there is "creature feep" (programmer's slang for feature creep), which only ever increases the size of the code 🤓 But every now and then, the right way to fix a bug is to delete code. In this case, it would have been part of one line of source code. I don't actually see the source code, but I occasionally hand decompile small pieces of code to document it.

So... was this one of those rare cases? While I was pondering it, I came across a comment I added months ago giving the date for one of the firmwares. I started wondering where that date came from. I checked the Czech site, and my comment disagreed with the Last Modified date on the Czech site by one month. Then it hit me!

💡

Coulomb (I said to myself), firmware update files come in archives (.zip, .7z, or .rar usually). And archive files have date stamps stored in them. I usually ignore those dates, indeed I ignore the archive altogether as soon as I have extracted the files. But of course they are precious information. Sure enough the one that I thought was the latest (90.28) shows the dsp.hex file as modified on 2021-09-22, and the other contender had its dsp.hex date last modified in 2021-03-26. Sanity check: 45.04 (numerically the lowest firmware number, so presumably the oldest) showed dsp.hex last modified 2020-09-17. Duh! That looks to be a fairly safe and sure way to find out when a firmware update file was created. Once it's in the archive, nothing on my computer system is going to change that date and time.

I have an "outlier" firmware, 45.84, that doesn't seem to fit the usual numbering scheme. Because of its high 45.xx number, I assumed for a long time that it was the latest, and put more effort into studying and commenting it. But its dsp.hex file is dated 2020-09-10, the oldest so far.

Ok, so how about firmware version 90.19, as mentioned above? It sounds like it has to be fairly recent, just a little older than 90.28, right? Wrong. Its dsp.hex file is dated 2020-12-11, earlier than 45.07, but not as old as 45.04.

I hope that this isn't some scheme to confuse the clone manufacturers, and I've just blown the secret. I certainly hope not; I have no wish to make life easier for the illegal cloners. If it is, then it's one of those schemes that confuses customers way more than it confuses the cloners.

The theory that the 90.xx versions are experimental and the 45.xx are the official stable versions might still be true. But 90.28 reports as such only if the machine it's running in is a 230 V machine; if it's a 120 V machine (North American market), then it reports as 45.28. Another firmware file reports as 90.11 or 50.00. 45.84 reports as 50.75 when running in 120 V hardware. I suppose that doesn't necessarily invalidate the theory; it just means that the numbers are all over the shop.

Forum member @Dirkie reported that version 90.28 fixed something that none of the 45.xx firmwares did. That seems to support the idea that the 90.28 firmware is an experimental one, designed to fix one problem. Happily, it's one of the ones that I compared tonight and found to differ only my being 4 bytes shorter in one function than another firmware. So I know the change that was made! It's 3am here now, so I'll have to leave investigation of that to later.

Edited January 10, 20223 yr by Coulomb

On 2022/01/06 at 6:26 PM, Jopie said:

Good day

I had a problem today with the MAX 7.2kW where it didn't want to charge the batteries at all.

It was a very cloudy day, so I changed it to SUB mode with the hope that it will charge the batteries as "first priority" (like my old VMiii did).

After no success with that, I changed the "solar only" charging to both other options but still it would only charge when there were excess sunlight (load was first priority)

So I did the right thing and attempted to load new firmware, by far the best decision this year.... (only 6 days at least)....

Original firmware taken down before disaster strike: U1: 45.07,  U2: 12.13,  U3: 00.00 (don't know what this U3 is for).

After turning the inverter into a brick, I eventually got 12.03 flashed and then the screen came on at least. But trying to get 45.04 or 90.19 loaded doesn't work at all.

I've tried through the coms cable and via USB stick, no luck.

When I power it up, it just goes through the normal countdown 6,5,4,4,3,2,1 and the 32 flashes. Tried with batteries only and AC only, makes no difference...

It "writes" 2 blocks and stops as per attached photo.

Also, with the old (working) firmware, I did notice that it backfed into the grid on SUB setting... Another reason why I though it would be brilliant to load newer firmware.

Any help would be much appreciated 😀

The 90.28 firmware fix the sub bug for me

Thanks Coulomb for the detailed explanation. I have tested 90.19 on Sunday, did not charge batteries in SUB mode, same as original 45.07 I bought the inverter with... I'm still waiting for firmware from supplier, I think they "ignore" me due to my last email to them I stated that the inverter was bricked😆

@DirkieIf they don't come to the party, where can I find this beautiful 90.28?😀

Hi @Jopie sorry for the confusion the problem that it fix for me was not in sub but I think it is still the same problem my problem was that I set the inverters in sbu and then I set setting #16 so that the inverters wil charge the battery from grid at night when solar is not available. With the old firmware it dit not charge it but with 90.28 it does charge the battery I wil try to upload the firmware for you

Hi @Jopie here is 90.28 for Axpert Max 7.2Kw

90-28 OG-7.2.zip

Thank you very much, I'll give it a try😀

 

Good day

I managed to flash 12.13 and 90.28 just now... I was very happy at first to see that it charged the battery in SUB mode, until the AC kicked on again... Then it draws 0.5A

I'll see if it feeds back to the grid tomorrow and report back... But at this moment it seems like the same problem still exists... No charging of the batteries in SUB mode no matter what charge settings are used if there is insufficient solar power...

Tomorrow my new batteries arrive then I'll have 16x 105Ah deep cycles, should be enough to just keep it in SBU mode permanently... It's just those cloudy days that worries me a bit... I usually would put it on SUB mode before I leave for work on the old VMiii, then I knew my batteries would be fully charged and access solar would not be wasted.

Thanks for all your help @Dirkie and @Coulomb

PS Still no feedback from supplier wrt their firmware yet, I'll post here if they send it😀

23 hours ago, Jopie said:

Good day

I managed to flash 12.13 and 90.28 just now... I was very happy at first to see that it charged the battery in SUB mode, until the AC kicked on again... Then it draws 0.5A

I'll see if it feeds back to the grid tomorrow and report back... But at this moment it seems like the same problem still exists... No charging of the batteries in SUB mode no matter what charge settings are used if there is insufficient solar power...

Tomorrow my new batteries arrive then I'll have 16x 105Ah deep cycles, should be enough to just keep it in SBU mode permanently... It's just those cloudy days that worries me a bit... I usually would put it on SUB mode before I leave for work on the old VMiii, then I knew my batteries would be fully charged and access solar would not be wasted.

Thanks for all your help @Dirkie and @Coulomb

PS Still no feedback from supplier wrt their firmware yet, I'll post here if they send it😀

Im going to flash 90.28 and report back also
hope it stops the grid feed in in SUB  

Good day

Bad news again with the 7.2kW MAX...

Seems like all 3 versions of software does the same thing, no charging of the batteries in cloudy conditions and it feeds back when the sun comes out...

Below I attached a screenshot, orange is load, green solar, red grid. Around 12:30 the geyser switched off and then it pumped 800W back to the grid, luckily I saw this and switched it back to SBU mode... Same thing just before 16:00, I switched it to SUB and immediately it fed back to the grid, switched back to SBU again...

The bottom graph shows battery charged nicely (SBU mode) until the geyser switched on at 10:00. At about 10:20 I switched it to SUB which worked as it should (red line) until it switched off as per above. But one can also see that no charging happened while the load was more than what the PV could provide.

It's a pity it doesn't read the battery current, that's why "battery watts" and "Current" is zero.

Today is very cloudy again, I'll be brave and leave it on SUB for the next 24 hours and have a look at those graphs again 😒

Thank you very much for your detailed reports.

4 hours ago, Jopie said:

Good day

Bad news again with the 7.2kW MAX...

Seems like all 3 versions of software does the same thing, no charging of the batteries in cloudy conditions and it feeds back when the sun comes out

I don’t think that they will be able to fix the feed in to grid with firmware but the charging problem must definitely be able to be fix with firmware.

Hi guys,

I also have the problem of no charging on cloudy days (my inverter has got firmware version 45.07). I haven't connected utility power, so can't tell about the unwanted feed.

My observation is that in cloudy conditions, the PV voltage is stuck to around 70-100V, pretty much independent of how many panels are connected in series. I have two strings with a nominal max power voltage of 200V. No matter if I connect only one string, or both in series, or parallel, the MPPT of the inverter does not seem to find the correct voltage under cloudy conditions. There is no shadow on my roof, and I can compare to another two strings on the same roof that I connected to a simple PWM regulator using a fixed voltage. While the PWM regulator achieves about 100-150W at 200V, the inverter shows mostly around 10-20W at 80V (Not sure how accurate these values from the display or from the app are, but I'd trust the voltage more than the power reading).
Only when the sun comes through and the power exceeds about 150W, the inverter increases the solar voltage also to around 200V.

I did not yet dare to flash a different firmware version, but if I understand Jopie right, it wouldn't improve the situation.
Do you think I should try nevertheless?

/nullmodem

 

Good day

I accidently figured out what to do the get the batteries charged.......

Turns out USB setting must be used on the MAX inverters. USB setting on the MAX is the same thing as SUB on the VMiii.

2 nights ago I switched over from battery mode (SBU) to utility (SUB) as I planned to use the aircon and don't want it running from the batteries. I accidently put it on USB instead of SUB. The next morning it charged the battery as "first priority". 

The MAX is perfect, just a little feeding back to the grid, but I can live with that.

Overall I'm very happy now

6 minutes ago, Jopie said:

The MAX is perfect, just a little feeding back to the grid, but I can live with that.

SBU takes care of back feeding issues. 😀

On 2022/01/07 at 1:01 PM, Coulomb said:

Removable display firmware version 12.13 is the latest I have (as of this post).

Anyone has a link to the 12.13 version and know what changed?

3 hours ago, birdibird said:

Anyone has a link to the 12.13 version and know what changed?

@Coulomb might know wat change 

On 2022/03/29 at 5:35 AM, Dirkie said:

@Coulomb might know what changed

[ Edit: re changes in removable display firmware version 12.13: ]

It seems to be minor. Perhaps some changes to the settings options, and a few icons (warning, clock) set or not set here and there.

Nothing in the BMS code as far as I know.

Edited August 23, 20222 yr by Coulomb

2 hours ago, Coulomb said:

It seems to be minor. Perhaps some changes to the settings options, and a few icons (warning, clock) set or not set here and there.

Nothing in the BMS code as far as I know.

Thanks.

Do you have a link to it? I don't see it in the files section.

edit: never mind, I found it in the 90.28 zip

Edited March 29, 20222 yr by birdibird

I have been with my and someone elses systems on firmware 90.28, but their system now 2 times had a bug that suddenly the batteries were not being charged anymore. (A 2x Max 7200 parallel system).

They also had code 71 in the screen of one of the inverters (the one with the pyl bms setting and the RS485 cable to the Dyness pack).

Confusingly code 71 is used twice, you can find these 2 in the manual

-Firmware version not the same (this is not the case)

-If battery status is not allowed to discharge
after the communication between the inverter
and battery is successful, it will show code 71
to stop discharging battery. (sounds like the code 69 message they canceled out in newer display firmwares)

My single system hasn't had this problem.

In the graph below you see we cold restarted the inverters in the evening and charging started again.

I am going to downgrade the 2 parallel Maxes to the 45 series firmware and see if this happens then.

Edited May 15, 20222 yr by birdibird

On 2022/05/16 at 4:54 AM, birdibird said:

but their system now 2 times had a bug that suddenly the batteries were not being charged anymore.

If it's when the battery is 100% charged then it's expected that the battery doesn't get charged, but there is a bug in many of the latest firmwares such that solar doesn't help power the loads either. Does that sound the same?

Quote

(A 2x Max 7200 parallel system).

Paralleled MAXs would be relatively uncommon, so it would not surprise me if there are bugs in the parallel code that haven't been discovered yet (at least by Voltronic engineers).

 

On 2022/05/16 at 4:54 AM, birdibird said:

They also had code 71 in the screen of one of the inverters (the one with the pyl bms setting and the RS485 cable to the Dyness pack). ... (sounds like the code 69 message they cancelled out in newer display firmwares)

Except that it's kind of opposite to 69: 69 says the battery is full, don't charge; 71 says the battery is empty, don't discharge.

BTW, fault code 69 is back in the latest King firmware 02.70, but to prevent panicking users, it only comes on if the SoC is < 90%.

On 2022/05/16 at 4:54 AM, birdibird said:

In the graph below you see we cold restarted the inverters in the evening and charging started again.

Was the reset about 6pm, when the solar stopped following load, or just before 7pm, when solar charging jumped up? I'm guessing the former, with the sun coming out from  light clouds before 7pm. Impressive to have sun till 8:30pm, BTW. Even with daylight saving (so it's really 7:30pm as the sun moves relative to the earth).

Do you know what the battery SoC would have been at 6pm? It looks like the battery was at near 100% since the start of the graph (9am), since the solar power tracks the load. But if it wasn't 100%, then that's the problem, right?

On 2022/05/16 at 4:54 AM, birdibird said:

I am going to downgrade the 2 parallel Maxes to the 45 series firmware and see if this happens then.

The MAX 7.2 kW firmware numbers are a bit chaotic. From newest to oldest, I have:

90.28 45.07 45.06 90.19 45.04 45.84 90.11. The last three (earliest versions) are only about a week apart from each other, in September 2020. 90.28 is 22/9/2021 (almost 8 months old as I type), the next 2 are March and January of last year. I'd be interested to know the results of your test. It's possible that the problem is actually in the removable display firmware, since BMS interaction is probably involved.

[ Edit: 90.29 → 90.28 ]

Edited October 16, 20222 yr by Coulomb