Coulomb

I used to believe that, but the models with '28066 processors have a newer reflash tool that has the USB option as shown above. My understanding is that with a suitable cable (e.g. USB-A to micro-USB), you can use this option and it should communicate directly with the micro-USB port in the removable display or the inverter. I've never done this, and I'm quite hazy about how it would work, but I believe that some readers on this forum have reported success. That zip file has had some 85 downloads, so if there is a problem with it, we should well and truly have heard about it by now. @Ale70, you have posted in Italian again. Even your quote of my post comes out in Italian. Please sort this out. Are you sure that your inverter has a '28066 DSP? The best way to know is to check the presently working main (U1, DSP) firmware version. Note that the 9x.xx firmware versions are all over the place. If your present firmware is 72.xx or 91.04, then you have this processor. Otherwise, please report what you have.

Thanks for the report; it's invaluable! None of the VM II/III/IV firmwares have what I had come to presume was the best MPPT logic, yet here you have proof that the issue can be solved without it. This prompted me to re-examine my analysis of the MPPT code, and I've come across numerous errors in my analysis. Sigh. So I have a lot of work ahead of me, but hopefully soon I'll have an idea of what's going on, so that I can patch other firmwares if necessary, or at least identify which new firmwares are likely to have fixed the issue.

That is a little strange. Though the power supply will start up at a pretty low voltage, and below a certain voltage, the battery voltage would show as zero, as on your photo. I don't know if those can overlap. I assume that the top photo above is from below. The melting of the fuse's heatshrink tubing strongly suggests that there is resistance around that fuse. The soldering may be imperfect. Can you perhaps measure the resistance from the battery positive terminal to the other end of the fuse? It might be best to run some 3A from a bench power supply through there, and measure the voltage drop on a multimeter millivolt range. Another possibility, though it implies two faults, is problems with the battery measurement resistors. R29 in the top photo is one of them. There should be 6 or 8 of them near each other, all marked 104 or similar (indicating 10^6 or 1.0M ohms each). Check that these measure about the right value (you should not need to take them out of circuit). Bad soldering there might have triggered the Battery not Present warning. It's possible that heat from the fuse resistance has affected some of the measurement resistors.

I don't believe that it makes any difference, but I'm not sure. Sorry.

It's hard to tell from the photo; it looks to me that the darkness / tarnish / burned PCB track / capacitor gunk gets worse the further from the fuse (the closer to the set of 3 battery capacitors). Is that the case in real life? If so, one or more of the capacitors may have let some gunk escape. Though it's possible that this is merely flux from where they solder the power transformer (the big thing at the top of the photo). The thing that usually generates a burning smell and carbon residue like that is a semiconductor, or several. I see a heatsink just to the right of the battery capacitors; can you see any burned devices (MOSFETs or IGBTs) that are discoloured, cracked, or have legs missing? I find it unusual that the inverter doesn't display any fault codes if there was a catastrophic failure in the battery circuitry. Does it display bP anywhere, perhaps with a ⚠ warning sign nearby? That would indicate a Battery (not) Present warning. What value does it show?

That seems a little high; 5 kVA models draw about 50 W with the DC->AC converter running, and 15 W otherwise. Add 50% because it's a 6 kW model. Actually, my figures are based on a now quite old PF0.8 model, i.e. the DC-AC converter is rated at 5 kVA, but the DC-DC converter is rated at 4 kW. So a true 6 kW model might draw a little more than 75 W when the DC->AC converter is operating. In these later models, now that I think about it, might have the DC->AC converter active at all times, unless you are in SBU output source priority. And maybe also remove PV input as well.

It might be because you hadn't made your first post. Try again now that you have posted once. Failing that, does the 72.05 link work? The link is near the top of the first post this page.

Get updated firmware from your supplier (Elba Power?), or wait until a recent King II 6kW firmware appears.

Please post in English on this forum. It's hard to say what's going on with your clone. It seems to be running a Voltronic Max E firmware, but it's clearly a Max II. Any genuine unmodified Voltronic 46.xx firmware should accept the 90.06 based patched firmwares. Which patched firmwares have you tried? A few clones can't accept any firmware updates, but that was a long time ago; cloning technology has advanced a lot since then. Here's my guess: The clone is using a Max E firmware (designed for removable displays) because Max II firmware was not publicly available until 2023. This clone is made in 2022, if they follow the Voltronic serial number convention. They may or may not have changed the firmware; if they changed the reflash command, then you will not be able to use standard Voltronic software, patched or factory. If this is the case, you will have to get new firmware from your supplier (good luck with that). If they have a version based on say 46.71, then this will largely fix the stuck at 90 V issue. There is also the chance that you are not updating correctly. Are you using a USB to serial adapter? If so, what brand/model? There are many that simply don't work with the Voltronic reflash tools. Can you post the error message from when the reflash failed?

99.9% of recent installations would be for high PV voltage. My guess is that the installer assumed the usual 90-450 V MPPT range, which isn't going to work with this inverter. Quite possibly the solar charger has been damaged if this is the case. I'm a little surprised that this inverter was made as recently as mid 2019.

This is patched firmware versions x65.08 for the Axpert Max II 11 kW '2809 Twin, based on factory firmware version 65.08. Do not use with any other model. Use at your own risk. They fix the premature float bug, and change the value seen by monitoring software to 00x65.08. They implement all of Georg594's MPPT patches. Both MPPTs are set to the same minimum voltage. 265.08 has a minimum MPPT voltage of 200V. 165.08 has a minimum MPPT voltage of 150V. 965.08 has a minimum MPPT voltage of 90V (no change from the original; use only if you have a very number of panels per MPPT). The reflash tool only reads from the file dsp.hex. To use one of the variants other than the 200V MPPT minimum, delete dsp.hex; copy and paste the desired hex file; rename the resultant to dsp.hex. Start the reflash tool again. Firmware upload instructions for models with a round display. dsp_x65.08_patched.zip Edit 18/Jan/2025: Added MSVCRTD.DLL to the zip file. The 1.25MB version had this file missing. Sorry for the hassle.

I note very belatedly that the above are for the Axpert Max E 11kW '2809 Twin; the OP likely has an Axpert Max II 11kW '28066 Twin. So 2 out of 4 variables correct (11kW, Twin), but 2 incorrect for the OP (E versus II, '2809 versus '28066). I have not seen any firmwares for the Axpert Max 11kW '28066 ('2809 or '28066, Twin or non-Twin).

Like I said above, "Wherever you found the U1 version, the U2 version is for the display." You can just press the down button on the display until you get to a "page" with U2 at the top. Just report the two numbers under that. Example: This inverter is running display firmware version 02.83. I don't really know, but I suspect that it doesn't matter.

The MOSFETs in these models run right along the side, hidden under an overhang. That's where you usually see the worst carnage. In this case, you should also check the IGBTs on the middle heatsink. Sometimes the damage is fairly subtle, like a crack in the epoxy. Unless you are handy with electronics, fir example you have access to a desoldering station, it's probably time to retire this inverter.

Ok. The MKS II 5 kW is quite an old model by now. So I'm now guessing that the capacitors have dried up inside. If you are handy with electronics, it can be a fairly economic job to replace the usual suspects, and that might keep it going for several more years. The relevant information is in this post (scroll past the first photo), for you or for someone that you call in. Otherwise, yes, it might be time to replace that inverter.

You need a main firmware that handles the QDOP command. 46.05 is way too old for that. The latest is 46.83, available here. Yes, the display firmware also needs updating. Wherever you found the U1 version, the U2 version is for the display. Either in monitoring software or on the display itself. It's probably 12.xx or 122.xx. If so, the latest version is 12.28, available here. The linked posts should have links to the firmware upload instructions.

Is this on the display of the inverter itself, or via monitoring software? If monitoring software, which one? If it's Solar Assistant, what is the "driver" for this model? I'm thinking now that yours is a clone, inexpertly put together. I'd say it's a VM III clone, hence the noisy fans. 60.11 is a Twin main firmware (for VM III new, but there could well be firmware numbering overlap with Axpert King 6kW), yet U2 19.18 is a non-Twin display firmware. If I'm right, it's surprising that it works at all. But this is all highly speculative on my part.

If you mean what is the best firmware to update to, usually you should not update firmware unless you need a particular bug fixed or feature only available in a later version. In the 56.xx firmware version range, the latest is 56.72, available from the link in this post: Firmware upload instructions for models with a removable display.

Does it show a non-zero value in the battery voltage part of the display? Has it been working and suddenly stopped? Has it been working but without a battery up to now? Is the clicking sound from the inverter or the battery? What sort of battery is it? Is it known good? What is its capacity? It sounds to me that you have no pre-charging on the battery input. That's usually OK with modern batteries that have a built-in BMS, as they use large internal MOSFETs to limit the current drawn when the battery is initially connected. But if it's just a bank of cells, the inrush current can be several thousand amps for a fraction of a second, which is very hard on the battery breaker and also on the capacitors in the inverter. Another possibility is that the battery capacity is too low to handle the charge of the battery capacitors, and the battery BMS shuts off. This may leave some charge on the battery capacitors, which is perhaps what you are reading. The clicking that you hear is possibly the inverter getting ready for battery mode, but this requires a surge of current, and if the battery capacity is too low, it may disconnect and/or the voltage might "brown out". Can you tell if the battery voltage dips at about the time of the clicks?

Fair point. But has it been rainy recently in your area? We had a rainy start to summer here, dry and hot for the last week, but Chile is a looong way away.

Interesting, thank you for your comments. Unfortunately, this is at odds with what I was hoping was leading to a consensus on what the best MPPT algorithm is. Oh well, hopefully a pattern will emerge soon.

These inverters have high (hundreds of volts) AC waveforms between the PV inputs and earth, even at night, if the DC->AC converter is running. Your theories about leakage to earth could possibly explain it. If you isolate (disconnect) the PV panels, do you still see this measurement? If so, it's an inverter fault, possibly noisy or poorly soldered measurement resistors. But that doesn't explain the PV current measurements.

The firmware likely won't fix a complete lack of communication via the PC (RS-232) port. * Are you using the supplied RJ-45 (this is NOT an ethernet port) to RS-232 cable? * Is this plugged into a known-compatible USB to RS-232 cable? As for updating via the USB port, yes, the main firmware can be updated; you need a suitable cable and a USB stick with the firmware on it, and no other files. But only the main (U1, DSP) firmware can be updated using this port. To update the display/communications firmware, you need to use the PC/RS-232 port and cables as indicated above. Firmware update instructions for models with a round display.

There is a DC-DC converter inside the inverter, between the ~50 VDC battery, and the ~400 VDC bus. The bus voltage gets chopped into a sine wave for the AC output. It can vary somewhat, between a minimum of about 350 V and a maximum of about 500 V. As the bus voltage goes up and down, the DC->AC converter adjusts to keep the AC output voltage fairly steady. Your problem is mostly when in battery mode, it seems; so the battery is supplying the load. In this state, the DC-DC converter has a fixed voltage ratio; lets call it 8x for ease of maths (is likely actually around 7.0-7.5). So as long as the battery voltage is less than 54.0 V, the bus voltage will be less than 432 V, and all is well. The problem is if the battery BMS decides to disconnect the battery from the inverter; it has MOSFETs to do that. It does this to save the cells from damage. Suppose that the battery is badly unbalanced, and one cell is lower voltage than all the others. When this cell reaches say 2.8 V, the battery BMS would usually be sending an urgent command to the inverter to shut down. But if you don't have the BMS to inverter cable, either direct or via a monitoring Raspberry Pi etc, then the only thing that the battery BMS can do is to disconnect the battery from the inverter. That saves the cell, but what happens at the inverter? It had a nice stiff battery to keep the battery voltage at a steady level, now suddenly it does not. The battery terminal voltage will likely plummet. At this point, it's not clear what happens. Usually it should stop with fault code 04, battery low, or BP warning (battery not present). But it might attempt to blend in power from AC-in. If it does, it could overshoot, and put way too much voltage at the battery terminals. If the battery terminal voltage goes over about 62.5 V, then the bus voltage could go over 500 V. That's my guess. Though if the battery is getting charged, then the buck converter comes into play, and things get more complicated. As for the bulk and floating voltages, these are merely settings. The bulk voltage setting is the value that the inverter aims to keep at or below when bulk charging (which is most of the time at the start of a charge). When the battery voltage reaches this level for a certain amount of time, the so-called absorb stage, then the battery is considered full, and the inverter just wants to keep the battery nearly full despite varying loads and varying solar. That's the so-called float stage, when the inverter attempts to keep the battery voltage at or below the float voltage setting. Usually this value is lower than the bulk voltage setting, but you have them at the same value. Earlier in this thread I suggested that you reduce the float voltage setting to about 51.8 V; the battery lives at this voltage for many hours a day, and this lower voltage will allow the battery to last longer (lower degradation), at the expense of a half percent or so of run time each day. At this lower voltage setting, it's also less likely that the inverter will overcharge the battery. for example after a large load comes off. It's possible that these situations could trigger bus voltage errors, though it doesn't seem to match your case. Yes, good idea. It could disprove or strengthen my guess above. Or as I suggested earlier, disconnect the solar input at night for a night or few. Especially if the solar panels are wet. Electrical leakage to earth on these models somehow leads to high bus voltage. But do this separately to the utility disconnection, otherwise you won't know which change led to the change (if any). This solar panel electrical leakage is my other guess.

I doubt that these bus voltage errors would be fixed this way. You get no control if the BMS calls the shots. I would first experiment with battery voltage settings, especially the bulk/absorb/CV and float voltage settings. Often these are too high, as the battery manufacturers tend to recommend. They are more interested in making their products appear to last longer than whether it causes problems with real-world inverters.