Coulomb

Inverters are often picky about generator quality, and Axperts certainly are. See FAQ #7. Unfortunately, no firmware update or patched firmware will help.

As noted earlier, these 55.xx firmwares actually have the "stuck at 120V" issue, rather than "stuck at 60V" as I originally incorrectly posted. But I believe that many owners will need the patched firmware anyway, as merely changing the minimum MPPT voltage is only part of the solution in general. These are patched firmware versions x55.81, based on factory firmware version 55.81 for the Axpert VM III 6kW/4kW/1.5kW '28062 non-Twin. Do not use with any other model. The original hex file is dated 18/Nov/2025. This firmware works around the "stuck at 120 V" problem by changing the current threshold from 0.50 A to 0.05 A. It also fixes the premature float bug, and changes the reported version number to x55.81. The removable display will continue to show version U1 55 81. Use at your own risk; I cannot test this firmware as I do not own an Axpert VM III. It has been "sanity checked" by disassembling the firmware and confirming that the intended changes are present. Choose the minimum MPPT voltage that is higher than your working panel voltage even in poor light. dsp_155.81_patched.hex: Minimum MPPT voltage 100V. For very low panel voltage. dsp_555.81_patched.hex: Minimum MPPT voltage 150V. dsp_255.81_patched.hex: Minimum MPPT voltage 200V. dsp_955.81_patched.hex: Minimum MPPT voltage 90V. For extremely low panel voltage. Firmware upload instructions: https://forums.aeva.asn.au/viewtopic.php?p=87767#p87767 dsp x55.81_patched.zip

I'll just add that I now believe that this is mostly because SOME VM III firmwares come with a minimum MPPT voltage of 120V, while others come with a minimum of 60V. You updated from a 60V one to a 120V one. In some cases, perhaps many cases, this will largely solve the "stuck at 60V" problem. However, I think that it may simply transform into a "stuck at 120V" issue instead. Not as bad, but depending on your panel configuration, it may still be far from optimal. I'd love to hear from actual users' experiences. So it looks like many VM III owners will still need patched firmware to get proper MPPT performance.

Yes. Ok, these do seem to be popular models.

Yes, 55.81 is the latest firmware compatible with models that come with 55.xx, including 24 V models. But no, it does not fix the "stuck at 60 V" issue, as far as I know. If there is enough interest, I can patch 55.81. Edit 23/12/2025: I assumed above that all VM III firmwares use a minimum MPPT voltage of 60V. In fact, some, including all the 55.xx firmwares, use 120V. Perhaps this is Voltronics' workaround for the "stuck at 60V" issue. But I don't believe that it's a complete fix.

I assume you meant 81.xx. No, both of those are for the '28066 DSP, and yours is a '2809. They would cause instant bricking. Unfortunately, the 8 kW models don't seem to get a lot of firmware updates. It's OK to use, but it's a few months older than 54.06, and also doesn't have the required command support. The .6x and higher are "experimental" versions to fix a specific issue, which often does not find its way back to the "production" versions (.0x, .1x etc). That's probably because you forgot to mention how many units per month you want to buy. Voltronic doesn't deal with customers, except if absolutely necessary and then only via resellers. Sadly, that's the only official path for support.

Sorry for the late reply. No, your model is a high PV voltage model. All the fully patched firmwares are for 145 V max PV models. Your model has "III" in the name. It may be a VM III or an MKS III. Depending on your main (U1) firmware version, there may be an existing lightly patched version to work around the "stuck at 60 V" MPPT issue.

This is main (U1) factory firmware version 64.69 for the Axpert Max 11kW '2809 Twin. Do not use with any other model or version. Use at your own risk. The hex file is dated 18/10/2023. Firmware upload instructions for models with a removable display. This firmware processes the QDOP command, so with suitably recent display firmware, it should support SoC% for settings 12,13, and 29 when connected to a suitable battery's BMS. MAXE11K_TWIN_6469.7z

This is patched firmware version 378.08, based on factory firmware version 78.08 for the Axpert max E 11kW '28066 Twin. Do not use with any other model. The original hex file is dated 2024/6/24. This firmware has two sets of patches: the "derate at 450V" patches, and the "stuck at 90V" work arounds, with minimum MPPT voltage set to 200 V. So this will suit only a few people! This firmware works around the "stuck at 90 V" problem by changing the current threshold from 0.50 A to 0.10 A. It also fixes the premature float bug, and changes the reported version number to 378.08. The removable display will continue to show version U1 78 08. There is only a 200V version for this particular combination of patches. Also in the zip file are the original factory firmware, and firmware only patched for derating at 450V, for convenience. Use at your own risk; I cannot test this firmware as I do not own an Axpert Max. It has been "sanity checked" by disassembling the firmware and confirming that the intended changes are present. Firmware upload instructions. https://forums.aeva.asn.au/viewtopic.php?p=87767#p87767 dsp_378.08_patched_200V_derate450V.zip

You might also have a clone. Clones are notorious for leaving off parts that don't appear to have an immediate function, or not copying the Bill of Material properly (not noticing that one of the three relays has one extra pin, due to the extra contact). It doesn't sound safe to me, without thinking about it at length. Edit: It might be as simple as replacing one relay with a double throw version; some clones have the extra PCB traces and hole for the extra contact copied faithfully from the original, but they populate that position with a slightly cheaper relay with the single throw contact. Thanks to @fhocorp for pointing this out to me.

Ah, yes. I didn't realise that 68.13 is for 4.2 kW or 6.2 kW. Some resellers seem to get special firmware with an extra 200 W of rated power advertised (I haven't checked to see if it actually would get delivered). This seems to be one of those "special" firmwares. Alas, there are few other 68.xx firmwares. The other one I have is 68.04, and it's 4.2 kW or 6.0 kW. So perhaps 68.xx is the "special" version of 60.xx. So you might be better off with 60.10, or possibly 60.87 (older than 60.10, but newer than 60.95, and is another "experimental" firmware version number).

Right. But with the switch off and no charging sources (PV or AC-in), it should switch off after about 30 seconds. To be revived if PV is present next morning. I assume that your problem is that it doesn't stay off when the DSP commands the power supply to go off. I've had a lot of problems with the opto isolators on the 145 V max Solar Charge Controllers. There is one opto dedicated to waking the inverter at dawn; if this is shorted or very leaky, it would restart immediately after shutting down, I think. For example, U4 at the bottom of the below traced schematic (parts designators like U4 may differ, especially with 24 V models):

For other readers, I have answered this on another forum (55.xx main firmwares for the Axpert VM III 6kW(etc) non-Twin).

Ah! I overlooked that bit. Oops. Yes, the capacitor mod does seem like it won't explain that problem. It could be something to do with the buck converter; that "works harder" in those conditions. I can't think what exactly about the buck converter, though. Perhaps a weak IGBT? Weak gate drive?

The spikes from the fast changing potential of the PV wires seems to interfere with communications. The display talks to the DSP via serial lines. It's possible that this could also disrupt the BMS RS-485 commands as well. The capacitor mod seems to apply to many high PV voltage models, which is most if not all of the models made in the last several years.

My remaining question is why I was never able to connect to the chip via COM when the inverter was working. I could connect to the display, but that was it.On these inverters, you only ever talk directly to the display. The display is supposed to forward your commands to the DSP, amongst its own regular commands which try to keep the display up to date. When warning code 32 is in effect, that means that the DSP is not responding to commands sent from the display (either forwarding commands from the RS-232 port, or its own regular commands). With the wrong firmware in there, the DSP was never going to respond to anything, unfortunately. Yes, it's a pretty crazy setup. As for before the warning 32, I don't know.

That's "Ads" (equalisation Activation DiSabled). It's because you're not declaring a flooded battery; equalisation only makes sense for flooded batteries. So just ignore settings 30-36. Yes, it's confusing. Yes, there are different firmwares for the main (U1, DSP) firmware and the display (U2) firmware. There is also different firmware for the touch and button removable displays, as well as the colour versus removable displays. The ability to set three settings in SoC% depends on both the main and display firmwares. The latest 54.xx firmware I know of is 54.06, available here. For the display firmware, I'm guessing it's 22.xx, but please post your present display (U2) firmware version to be sure. Updating both will likely fix most of your problems. Sorry for the late response. Edit: Oops; I assumed that the latest 54.06 firmware would support the QDOP command, but in fact on checking just now, it does not. To get that to work, you will need an even later firmware; ask whoever sold you your inverter. Sorry for the false hope.

There is a link to Voltronics' communications tool in the link; search for the keyword "tool" without the quotes. It's likely, but not definite, that Sumry is a reseller, getting Voltronic to manufacture the inverters with the Easun branding on it. Or perhaps Easun do that subcontracting, and Sumry are reselling the Easuns. That's Voltronic's chosen approach: let someone else deal with customer service. There are dozens of such resellers and brands, as well as many illegal copies. The Easun SMG II models do look quite Axpert like, so the chances are good that this will work. Unfortunately, Easun seems to have started off as a clone manufacturer, so some of the older Easun models may not be genuine. But Easun have been around for many years now.

I and others have made many patched firmwares for working around the "stuck at 90V" issue. See What Patched Axpert Firmware is That?

This means that you have an inverter with a '2809 DSP chip. Unfortunately, 81.xx firmwares are only for inverters with the '28066 DSP chip. The 81.xx firmwares will call a calibrate function in the firmware very early on, in the first few milliseconds of operation. Unfortunately, the '2809 is an older chip, which does not have ROM memory at the address that the firmware calls. The result is that the DSP crashes. The only hope is to load a 54.xx Max II firmware (54.06 is still the latest available for your model) and start the upload at just the right instant so that the bootloader detects the start of the reflash before it runs the main firmware. See the If you Think Your Inverter is Bricked for clues on how to do this. Or use a JTAG device to flash the firmware, but that has its own issues as the firmware is password protected. Or buy a new control board for your inverter, but the hassle is getting the exact right card with suitable firmware on it, when there are so many variants that look nearly the same. If you buy it from whoever sold you the inverter, they should get the exact model number from you and hopefully supply the right card. I try to make this difference clear in my firmware posts, but I get that the whole thing gets confusing. Voltronic don't help themselves much with this; they could do a simple test before calling the calibrate function. Sorry for your difficulty, and I hope you can get it resolved soon.

Perhaps you mean that when the inverter switches from battery powering the load to AC input powering the load or vice versa some appliances drop out or lights flicker. This can happen because of the brief changeover period (some 10-20 milliseconds usually, i.e. one or two hundredths of a second) when there is no power to the loads. Axpert inverters have a special circuit that attempts to make this changeover quicker, closer to the 10 ms than 20 ms. But this only happens with certain settings and in certain circumstances. If it's really annoying, it might be worth running a small Uninterruptable Power Supply (UPS) just for the sensitive loads.

Another possible reason for warning code 32 is if the last firmware update was incomplete. In that case, the bootstrap loader program detects this situation, and rather than allowing the faulty firmware to crash, it stays running, waiting for another firmware upload. The bootloader is very simple, so it can't even let you know that it's running. So the inverter looks like it's totally bricked, other than the display, which runs totally separate firmware on another CPU, and it shows warning code 32 because all the commands it sends are timing out with no valid response. If this is the case, merely restarting a firmware upload will often start the reflash process again, and assuming that it eventually succeeds, the inverter will spring back to life. See If you think your inverter is Bricked.

I believe that it's a clone, which is not the same thing. They could have left off all sorts of things. The shutoff circuit is quite simple; it basically has to be U8 (per the schematic trace) or what drives it. Try shorting that opto collector to emitter; the power supply should shut off and stay off. Restarting continuously sounds like bad capacitors in the power supply. You could try replacing the usual suspects.

The later display firmwares relented and removed the warning altogether. The latest available 25.xx display firmware is 25.16:

It could be this issue: