Hello,

i'am from Berlin / Germany and after some weeks of only reading this interesting forum, one question / problem occurs.

 

I set up an off-grid solarsystem with the Axpert VMIII Twin because it has a wide mppt Range from 60 to 455 V. So i remain flexible with the solar-panels.

An why 24V?  According to the datasheet, the Twin 4kW/24V consumes "only" 40 Watts without load, the bigger 6kW/48V needs 55Watts.

In my small system this ist probably better.

And i need maximal 2000W. 

 

Now my story:

In the last weeks i installed the Inverter "VM III TWIN 4k" + Pylontech Batteries + Solarpanels (2*455W,50V/9A).

At Friday,18.3.2023, i put the system into function. Everything works fine.

But in the night to monday,20.3.2023 shuts the Inverter down,after the battery capacity fell below the set value.  The grid voltage to the inverter was switched off.

So far everything seemed fine, the installed automatic transfer switch is switched to the "normal" grid.

But this morning i went to the basement, where the system ist and (one relay? in )the axpert  clicks regulary because the solar power is not yet sufficient for regular operation.

Unfortanely the display shows error "F09".  I looked it up on google and the power transistors were given as the cause of the error. However, the device did not generate any significant power and so this error seemed unlikely to me.

I swiched the grid voltage to the inverter and the pylontech's on, so that der inverter has enouch energy to start up and luckily i found him in the WatchPower - App. In this app i restored the parameters to default and everything looks fine.

Oddly enough, the serial number of the inverter ich changed to " 55355535553555"

Is that a problem for me? Can i restore the old serial number? I contacted the dealer of this device, "Solarpower24.it", he is located in italy, but i didn't get any reply.

 

I also build an Interface with an ESP32 to communicate with the inverter, but i found not the complete command set for this specific device.

I used "wireshark" to read  the communication between the Axpert (WIFI-Interface is inclusive) and the Android App, and extracted some commands. 

 

Thank you!

Peter

 

 

 

 

 

 

Thank you, Peter

14 hours ago, pintopf said:

In this app i restored the parameters to default and everything looks fine.

Oddly enough, the serial number of the inverter is changed to " 55355535553555"

I don't suppose you noticed if the serial number was normal before the restore to default settings?

14 hours ago, pintopf said:

Is that a problem for me? Can i restore the old serial number?

Almost certainly this is a problem. You have probably started a countdown that will cause fault code 90 in 60 days of operation. The serial number is part of the cryptographic calculation that decides if you have a legitimate Voltronic, or it's a clone running pirated firmware. Do you have the means of sending a QBOOT command? If so, what does it return?

The 553... serial number means that your eeprom was erased. Whatever caused the F09 fault also seems to have wiped or at least corrupted the eeprom. So settings "stick"? Say you bump the float voltage by 0.1 V, check that it's changed, power down and up again... is the setting still changed?

You can't restore the serial number yourself. You need a special "setserial" program that is unique to your serial number. You have to get that from your supplier. As far as I know, you can't use anyone else's serial number, though I've never tried. There are a few setserial programs floating available for download if you know the right places to look.

But you also have the F09 issue to fix. Hopefully, it's still under warranty, and hopefully you will eventually get a response from your supplier.

Edited April 5, 20232 yr by Coulomb

Thank you for your answer.

2 hours ago, Coulomb said:

I don't suppose you noticed if the serial number was normal before the restore to default settings?

On my screenshot you can see the original serial number. The Android-App creates a second device with the new serial number. It's the same hardware with a new number.

 

3 hours ago, Coulomb said:

Do you have the means of sending a QBOOT command?

Sendig QBOOT receives only "NAK".

 

I think, that the F9 error occurs, if only solarpower exists, without grid and battery power. Then is in the inverter no path to forward the energy in any direction.

i can reproduce this situation at any time.

With connected battery and/or grid, no error and the inverter works fine.

Yes, the device is under warranty, but it's from italy and it's a long way to italy.  I found the seriel number set program (AxpertSetTools), but i don't test it with the includet serial-number, because my inverter is under warranty.

Error 90 with shutdown sounds not so good, yesterday i send a second mail to my supplier. It was unfortanely the one and only supplier for the Axpert VIII Twin 4kW in Europe.

My hope was, than anybody can calculate for me the checksum for the AxpertSetTools, so that can i do this operation alone.

The Axpert works fine, and with my ESP32 ->MQTT interface and the corresponding Node-Red program i can control him nearly perfectly.

 

Greetings, Peter

 

2 hours ago, pintopf said:

the F9 error occurs, if only solarpower exists, without grid and battery power. ...

With connected battery and/or grid, no error and the inverter works fine.

Huh. It sounds like under these conditions, the solar charge controller (a boost converter) is shorting the bus, or at least diverting current from the bus somewhere so that the bus soft start power supply can't charge the bus capacitors fast enough. Maybe the bus voltage leaks out to the panels somehow.

3 hours ago, pintopf said:

My hope was, than anybody can calculate for me the checksum for the AxpertSetTools, so that can i do this operation alone.

I haven't studied it in detail, but the associated firmware code that checks the eeprom contents is very complex. I assume it's essentially cryptographic, so you need some magic number (a cryptographic key of some sort) that only Voltronic knows to generate the data in the setserial program.

Despite that, it's not very hard to defeat the fault code 90 check if you have a firmware update file for your main firmware. What is your main (U1) firmware version?

6 minutes ago, Coulomb said:

so you need some magic number (a cryptographic key of some sort) that only Voltronic knows to generate the data in the setserial program.

shit.😬

I take a screenshot from may app, so can you see all the numbers.

Thank you!

Greetings from Berlin!

U1: 56.05. It turns out that I have 56.11 for Axpert VM III 6 kW or 4 kW, so I should be able to patch you a firmware to work around fault code 90.

3 minutes ago, Coulomb said:

that I have 56.11 for Axpert VM III 6 kW or 4 kW

That sounds good, can you please patch the firmware for me?

Are also improves in the newer software?

Thank you, Peter

20 minutes ago, pintopf said:

Are also improves in the newer software?

I imagine so. Without two versions to compare side by side, or release notes from the manufacturer, it's impossible to tell.

In this moment, after 3 attempts, i received a answer from solarpower24, they answered me, that they where waiting for a response from voltronic.

Is it much work for you, to modify the firmware?  18 out of 60 days are already passed, and with your firmware i have a solution for my serial-number problems if voltronic sends no answer to solarpower24.

Greetings, Peter

 

here is the answer:

Edited April 5, 20232 yr by pintopf

12 hours ago, pintopf said:

Is it much work for you, to modify the firmware?

Sigh. My apologies. When I went to look for where to patch, I could not find the relevant code. Indeed, fault code 90 is not set anywhere in the firmware. It seems that they decided that it wasn't worth adding the fault code 90 logic in the VM series, as the flash space is quite small at half the size of the other processors used in Axperts.

I had a look in an older VM III firmware, and there it was: missing!

Sorry for the scare. It seems that you don't have to worry about that issue, at least.

Many thanks for your effort.
I'll wait and see what my seller says. If the errormessage only shows in the display, it is no problem for me.

Would it be worthwhile for me to import the new firmware?
Could you please give me a source for this?

I have another question:
Is it possible to implement in the software the powersave (shut off the DC-AC converter) command you mentioned in another thread for another Voltronic inverter?

Or is my hardware to cheap?

 

Greetings, Peter

 

 

 

3 hours ago, pintopf said:

Could you please give me a source for this?

I haven't gotten around to patching it yet, so there is no source.

If you mean factory firmware version 56.11, that might be a good idea, just to establish that it does run on your hardware. Alas, I seem to only have had a .hex file for this. Sorry, I have no energy to package it up tonight.

Edit: I see I got it from this post, in case you want to try it yourself:

 

Edited April 6, 20232 yr by Coulomb

In the linked thred, on february 8, user @lucass descibes the same problem, which my axpert has, as the EEPROM deletes.

Quote

I have bought this inverter  Axpert VM III TWIN but i have some issue with the mppt because at the morning oscillate from 0v to VOC 180v (4s 550w solar panel)

But i can't read in this thread, that this issue was solved with the new firmware.

My inverter is under warranty, an new firmware with unknown features is to risky for me. i loadet the firmware down from the link, if i need it later.

I solved this problem for me with an "shelly 1PM" remote switch, so i switch the grid for my inverter 1 hour before sunrise on, he awakes, ten minutes after (enugh time to connect the pylontech's) my node-red programm switches the battery on (set the shut-off capacity to 10%), an after 1 hour i swiched the shelly automatically  off.

If i don't switch the battery on, with grid connected, the inverter sucks 2,5 A from the battery for nothing.

When enough solarpower is generated,  i can swich the battery off, is neccesary ( shut-off capacity to 90%), an the inverter works fine in "charging mode".

On 2023/04/06 at 11:34 AM, pintopf said:

Is it possible to implement in the software the powersave (shut off the DC-AC converter) command you mentioned in another thread for another Voltronic inverter?

Now i have realized this feature with an optocoupler  between the contacts 7 & 8 in the connection inverter <--> display, which emulated the on/off switch.

My ESP32 controls now also the optocoupler via MQTT, so i not longer need reprogramming the battery shut off level everytime i want to switch the DC-AC converter off, to save power. 

I measured 1,6 A (25V) battery current at no load . If the inverter in standby ( the switch (optocoupler) released), the current is near 0A.

Unfortunaly, i dont have a complete list of commands for this inverter, have anybody in this forum such a list?

In the status flags (QFLAG) and in my app exist a point called "enable/disable power saving", but i can't set this bit. 

Is it functionally in newer firmware-versions?

 

I'm still waiting too an answer because of my serial number problem  from "Solarpower 24".  50% of the 60 days are over.

 

2 hours ago, pintopf said:

i dont have a complete list of commands for this inverter, have anybody in this forum such a list?

Check the files section for protocol manuals. The VM IIIs have slightly different commands, e.g. QMN instead of QGMN, but most of the commands are the same as the other models.

 

2 hours ago, pintopf said:

Is it functionally in newer firmware-versions?

The VM models are cut-down with half the firmware size of other models, so it may be that power saving mode is one of the features that was cut. Your manual should tell you; if power saving mode is enabled, it will appear in one of the first 5 or so settings.

On 2023/04/05 at 6:35 AM, Coulomb said:

You can't restore the serial number yourself. You need a special "setserial" program that is unique to your serial number. You have to get that from your supplier. As far as I know, you can't use anyone else's serial number, though I've never tried. There are a few setserial programs floating available for download if you know the right places to look.

But you also have the F09 issue to fix. Hopefully, it's still under warranty, and hopefully you will eventually get a response from your supplier.

Good day @Coulomb,

I have the same problems.
I have 2x11k MAX Plus HV 48MPPT working in parallel since 6-7 months which were flashed with original TommaTech (TT) Firmware 73.06.
Suddenly one good morning one of Inverters change its SN to 55355535553555. Few day latter other Inverter change its SN to 55355535553555 too.
I tried to flash but unsuccessfully with Error Code 32.
I read somewhere that if You cannot complete flashing Inverter Firmware You have to re-flash first the Display and after that to flash Inverter.
I try to re-flash the Displays but also successfully up to 50% of process only.
Briefly, an Specialist here in Bulgaria was able to repair somehow my Displays and after that I was able to re-flash my Inverters to 73.06 again successfully. 
Unfortunately, few days latter I got shut down of Inverters with Error Code F90 of one of Inverters. I re-flashed that Inverter with 73.06 again and so far everything is working let say fine BUT problem with incorrect SNs remains. Speaking honestly I thinking that main problem is in Displays themselves. 

Year ago when I was working the Systems with 2x7.2k MAX Plus HV 48MPPT and I had some problems with Inverters the TT Support send me Software Tool for restoring Serial Numbers of that 2x7.2k MAX Plus HV 48MPPT Inverters. I can send You that software to You.

My questions is: Can You find out How that Software configure the Serial Numbers? so that to recover Serial Numbers for my Inverters? 
That will be really Highly appreciated. Thank You very much in advance!...

On 2024/10/14 at 1:25 AM, aamilev said:

Suddenly one good morning one of Inverters change its SN to 55355535553555. Few day latter other Inverter change its SN to 55355535553555 too.

That's the serial number that results from the EEPROM losing its contents. Unfortunately it will probably lead to fault code 90 eventually. I don't know what causes the EEPROM to lose contents; possibly glitches on the power supply due to capacitors drying out. 

On 2024/10/14 at 1:25 AM, aamilev said:

Unfortunately, few days latter I got shut down of Inverters with Error Code F90 of one of Inverters. I re-flashed that Inverter with 73.06 again and so far everything is working... 

I'm surprised that reflashing fixed it, even temporarily. Perhaps the EEPROM is gettting very bad. How long till fault code 90 is going to happen is also stored in the EEPROM.

It will be best to get the patched firmware running. Presumably, if you can reflash with factory firmware, then you should be able to run the patched firmware, unless I screwed up with the patching. It's a quite simple patch, so it's hopefully unlikely that I screwed up. 

It's actually not all that hard to send a new serial number to EEPROM, it's basically one command, though you need ATE commands either side to force it through the display. Unfortunately, that on its own isn't enough to prevent fault code 90. But it's nice to be able to distinguish between the two inverters. 

On 2024/10/14 at 1:25 AM, aamilev said:

My questions is: Can You find out How that Software configure the Serial Numbers? so that to recover Serial Numbers for my Inverters? 

No. As above, the serial number on its own is easy enough, but to avoid fault code 90, you need other values set, and those other values are cryptographically hard. So the best way around that is the patched firmware, unless you have a setserial program for your particular serial number(s).

One of my inverters (installed in 2016) had this problem, and I was able to change the serial number back to the original. I run patched firmware all the time, so fault code 90 was never a problem. That was just after I changed some power supply capacitors. It was so bad that the inverter would restart every few seconds. It's been running fine now for 15 months. 

4 hours ago, Coulomb said:

I'm surprised that reflashing fixed it, even temporarily. Perhaps the EEPROM is gettting very bad. How long till fault code 90 is going to happen is also stored in the EEPROM.

Yes, they are working fine for the moment but with that fake 55355535553555 Serial Numbers.

4 hours ago, Coulomb said:

It will be best to get the patched firmware running. Presumably, if you can reflash with factory firmware, then you should be able to run the patched firmware, unless I screwed up with the patching. It's a quite simple patch, so it's hopefully unlikely that I screwed up. 

@Coulomb - Patch of which Firmware You think will be the best to install?
73.06 - currently installed and I have not find a patch of this Firmware
73.63 /273.76 / 373.76
73.78 / 173.78 / 273.78 / 573.78

On 2024/10/15 at 10:15 PM, aamilev said:

Patch of which Firmware You think will be the best to install?

As per PM, I'd say factory 73.78, since it seems to fix the MPPT issue and will avoid fault code 90 (I have no idea why that version avoids FC90 and others don't).

But that goes against what you reported in your PM. So I await confirmation on what version had the stuck at 90 V issue for you.

42 minutes ago, Coulomb said:

As per PM, I'd say factory 73.78, since it seems to fix the MPPT issue and will avoid fault code 90 (I have no idea why that version avoids FC90 and others don't).

But that goes against what you reported in your PM. So I await confirmation on what version had the stuck at 90 V issue for you.

Maybe I should try 273.78 as I had again 90V staking issue with Original 73.78?

10 minutes ago, aamilev said:

Maybe I should try 273.78 as I had again 90V staking issue with Original 73.78?

Sure, if you want, and if your panels rarely go below 200 V even in shade.

But if you didn't mix up version numbers and 73.78 really doesn't have the good MPPT logic, then it means that what I thought I knew about the good MPPT logic is wrong.

So you're certain that the stuck at 90 V behaviour that you saw happens in 73.78? If you mixed up 73.78 with 73.06, then that would make life much easier for me 🙂

On 2023/04/05 at 1:11 PM, pintopf said:

shit.😬

I take a screenshot from may app, so can you see all the numbers.

Thank you!

Greetings from Berlin!

Hi,
Thanks to @Coulomb advice I was able to set my Serial Numbers back to originals with Command ATE1 and SID14XXXXXXXXXXXXXX000000 after that, where XXXXXXXXXXXXXX is the Orginal 14 digit Serial Number from Inverter Label.
Just be aware that ATE0 to close connection to Inverter is not working at all and Maybe You need to re-flash Your Firware.

6 hours ago, Coulomb said:

Sure, if you want, and if your panels rarely go below 200 V even in shade.

But if you didn't mix up version numbers and 73.78 really doesn't have the good MPPT logic, then it means that what I thought I knew about the good MPPT logic is wrong.

So you're certain that the stuck at 90 V behaviour that you saw happens in 73.78? If you mixed up 73.78 with 73.06, then that would make life much easier for me 🙂

No, I don't mixed, I am sure it was 73.78. Please see below Print screen I made with 73.78 FW.
Tonight I will install 273.78 and will share results too.

This is Results from this morning with 273.78