Coulomb

I'm not familiar with any firmware for King IIs that is 60.xx, but then again I don't seem to have any that are rated at 6 kW. Are you certain that it's not a VM III model? The VM models are notorious for the fan noise, though there are some quieter firmwares out there.

They are not cross compatible in that sense, no. Yes, it should still be possible to parallel them. It's best if the firmware versions are fairly close to each other, in the sense that they were compiled from largely the same source code, but unfortunately Voltronic don't help with this. Just updating to the latest you can get is probably fine. But if you are concerned, ask your supplier(s) for the latest firmware update files.

The 325 VA is about what I'd expect to see. The 2000 VA is certainly not normal. I'd say that you have a problem with the board that implements the UPS (versus APL) AC input range (I can't recall the exact name, perhaps they call it a parallel board, which is very confusing). I think it's worth getting that board replaced.

Ok, thanks. So that's a King, not a VM III. Hopefully the calibration procedure will fix or at least vastly improve your problem.

That appears to be an InfiniSolar E 5.5kW, at least according to one search result. I no longer collect these firmwares, as I'm swamped as it is with Axpert (off-grid, these are on-grid) models. So I'm no help, sorry. I can usually help with Axpert firmware updates.

I may have misspoken. Now that I think about it, I think that the BMS to inverter comms is all one way, BMS to inverter. Sorry for the confusion. The inverter isn't interested in being a data hub; it just wants what it needs. You should be able to get a program that runs on a PC or Pi that communicates separately with the battery's BMS, often via an RS-232 port (as opposed to the RS-485 or CAN bus port for the inverter comms). I'm 99% sure that this won't work. Watchpower is only for talking to the inverter, not to the battery/BMS.

In the 5kVA MKS models, it comes from the main power supply (that provides ±12 V and +5V); one of the windings is half wave rectified and feeds the main power supply's UC3845 chip via a diode. (Another diode provides a way for the main power supply to bootstrap itself). Edit: It could well be different in the VM III and other models; the power supply sources seems to be one area where the various models differ from each other.

A lot of the earlier models only supported a maximum of 6 inverters in parallel/3-phase. Please post the main (U1) firmware version so I can check that they support 9 paralleled units or only 6. If they are only configured for 6, do not be tempted to try 7, as each inverter requires some memory in all the others. So it only specified for 6, the seventh will not be properly synchronised, which will lead to massive problems (a Big Bang). If only specified for 6, it might be worth keeping the 7th as a spare. Unless you have upgraded the capacitors and MOSFETs, these old models will be approaching the end of useful life (whether used or not, though time left unused would extend their working life somewhat). I presume that you are familiar with the MKS Parallel Guide:

Yes, Coulomb knows 😎! Main and DSP firmware are the same thing. The main/DSP firmware version is also known as the U1 version number. You didn't say what chipset you have, or give an existing main firmware version. I assume you have the Axpert Max E 7.2 kW non-twin, even though this is a King thread. For '28066 chipset (69.xx main firmware), the latest is 69.70, dated 2024-3-06, and this has the presumed good MPPT logic (no stuck at 90 V issue). For the '2809chipset (45.xx main firmware), the latest is 45.13, dated 2023-9-05, and unfortunately does not have the presumed good MPPT logic. As for display firmware, you are presumably running 12.xx, and the latest for that is still 12.28 (a year old now).

Yes. That sounds like the Solar Charge Controller isn't MPPT at all, but rather PWM. These connect the panels directly across the battery, only pulsing the power when the battery becomes full. If so, the panels should be reconfigured so that there isn't such a huge difference in voltage. That difference is costing a lot of potential solar power.

The VM III and King models are very different. Perhaps you could post a photo of the sticker on the side of the inverter? Are those voltages as measured by the inverter, the battery's software, or by a multimeter? It may be that your inverter's battery voltage calibration is way off. That can be corrected with the appropriate series of commands.

It's tricky with so many inverters and so many battery modules. It's best if you don't have the positive and negative of the inverters or battery modules "lined up" (same distance from the left edge of the bus bars). If you do, then the battery modules towards the edge do less of the work. If no one module is at the edge (maybe the positive connection is, but the negative connection is closer to the middle or even near the other end), then the current should share better.

We can't tell, because the only available firmware update file for 30.07 has all the same date stamps, meaning we can't trust any of the date stamps. My guess is that 30.08 is a little more recent than 30.07. The xx.yy firmwares, where yy < 20, seem to be "release" firmwares (not experimental), and seem to get released in sequential order.

My apologies if I'm misunderstanding you, but you do realise that you can't change the DSP by firmware upload? The '2809 versus '28066 refers to a Texas Instruments part number, that relates to how the millions of transistors are hard-wired. As for your other issues, I'm sorry I can't help as I have so little experience with the Infini / on-grid models' hardware and firmware. On 2024/11/04 at 6:23 AM, Piscanc said: > what is even more bugging me, that i still have intermittent alarm 61 on slave inverter (on 56.14). Assuming that these models work similarly to the Axpert on-grid models, warning 61 relates to a lack of connection to the battery's BMS via either RS-485 or via CAN bus. This connection is entirely handled by the display ("remote panel") firmware.

We don't know. If you have an early 52.0x main firmware, then it's probably worth trying 52.04. It's available in my second post in this topic.

It's possible that you need a resistor in one of the leads of the RJ-45 to D9 cable, as detailed here. I do mention this towards the end of the upload instructions. It's also possible that the USB to RS-232 adapter isn't one that the reflash tool works well with. You could try others; the firmware upload instructions mention known-good bands available from South Africa. Finally, it could possibly be your computer or laptop going to sleep before the reflash process is complete. I was going to suggest trying the USB stick reflash method; that avoids the RS-232 adapter altogether. But you probably can't get to the reflash by USB menu when the display is showing warning code 32.

Have you checked all the transistors, diodes, resistors, and zeners in that area? Especially for that MOSFET where the two resistors failed open circuit. Unless it was a soldering problem perhaps.

It looks like both the gate series resistor (47Ω) and the gate pull-down resistor (47 kΩ) were blown. To blow the pull-down resistor, there must have been very high gate to source voltage, which I would expect to cause more problems earlier in the gate driver circuit. But I would not expect it to affect PV measurements. Yes, as long as the part numbers are the same, it's fine to pair them up even if from different batches. That's probably the buck diode, if it came from the long heat-sink, not a small separate one. [ Edit: Duh! The photo clearly has a separate heatsink on it. Sigh. ] It's probably fine to just clean up the soldering, but replacing it won't hurt. I've actually never designed circuitry with MOSFETs or IGBTs; when repairing I've always been happy to leave the series gate resistor the same value as originally used. If you can find a recommended value for that resistor in the datasheet specs, then you could use that value. My feeling is that the Voltronic inverters are designed with fairly wide dead-time, so that they can just use whatever parts they can get for production, without having to match them to the MOSFETs.

I've never succeeded with that method, but I've not tried hard. My understanding is that for this model, all you need is the hex file, renamed if necessary to dsp.hex, and it must be in the root directory of the USB stick. Is it in the root directory? I assume that's a USB-A to micro-USB adapter between the stick and the inverter? You end up with the UPD YES on the display and press some key? BTW, 64.62 is slightly older than 64.09, but it does have the "presumed good" MPPT logic, unlike 64.09.

I do not have an update file for 20.45. 20.44 from 2017(!) is in this sister topic: There is also version 90.95, also for the Axpert VM II non-premium (there is a premium VM II model with presumed different firmware, as of about mid 2024). This firmware update is apparently from 2023. It supports PF1 and PF0.8, although the link it came from says PF1. The link also mentions 100A, suggesting that the maximum charge current has been increased to 100A. Indeed, 10-100A is allowed for setting 02, maximum total charge current. Though the QMCHGCR command only returns strings "10" through "80"; the author presumably forgot to change this command. It would be wise to use this firmware only if the label states that the maximum charge current is 100A. I have not otherwise compared this 90.95 firmware with 20.44. This is main (DSP, U1) factory firmware version 90.95 for the Axpert VM II non-premium. Do not use with any other model. Probably best to only use with models that indicate a rated maximum charge current of 100A on the ratings sticker. Use at your own risk. Firmware upload instructions for models with no removable or round display. VMII 100A PF1.0 ver90.95 reflash.7z

Welcome to the forum. That will be reactive power for sure, i.e. 2000 VA not 2000 W. But that 8 A will heat the cable just the same as a 2000 W load (at unity power factor). But that's about 5x the reactive power than I expect to see. Perhaps this model has a particularly large smoothing capacitor, compared to 5 kVA models. I'm not surprised that the firmware update didn't fix this issue. Your model uses the same firmware for 4 kW and 6 kW models; it decides which it is running by measuring the voltage at a particular input pin on start-up. But the display has to ask the DSP what model it's running as well. If the DSP gets the model wrong, catastrophic problems are likely. If the display gets it wrong, then that's just a nuisance. It looks like it's just the display getting this wrong. It's possible that a firmware update to the display will fix this problem. Be brave! As a point of interest, were you using a USB stick and On The Go USB cable to update the main firmware? That seems to be a bad translation. Could you re-word, please? I'm curious, especially if you have stumbled onto a method of making firmware updates more reliable.

This is factory main (DSP, U1) firmware version 78.08 for the Axpert Max E 11 kW '28066 Twin (with removeable display). Do not use with any other model or chipset, i.e. only if your existing main firmware is version 78.xx. This version does NOT have the presumed good MPPT logic, so it you are bothered by the stuck at 90 V issue, see the patched 78.07 firmwares. Use at your own risk. Firmware upload instructions for models with a removable display. MAXE11K_TWIN_DSP_78.08.7z

I assume that you mean you have 6 48 V nominal 100 Ah battery modules connected in parallel. By the looks, they are 15S LFP. This could well be your problem. That first battery module seems to be getting most of the charge current, but it is therefore also doing most of the discharge work. So you effectively have about a 150 Ah battery, instead of a 600 Ah battery. My guess is that your cables to the inverters are both at one end of the busbars. If you change them to connect to either end of the busbar, or preferably a little closer to the middle (say closest to modules 2 and 5, or between modules 1 and 2, and between modules 5 and 6), then the battery modules should share the load current and charge current more evenly. Since you have the three inverters, make sure that you have separate cables from each inverter to the bus bars. If you daisy chained the inverter wires, that may have contributed to the problem with the third inverter. After this change, it will likely take weeks for the modules to balance reasonably well. I note that once an 04 warning is triggered, it takes a battery voltage rise of 4.0 V before the warning goes away! This is so lead-acid. Finally, your battery charge voltages (55.0 and 53.7 V) are appropriate for 16S LFP battery modules, but other aspects of your system suggest that you have 15S LFP modules. Consider lowering these voltages to 52.5 V and 51.8 V respectively.

Quite true in general. However, in this case, it's fanciful to expect to get 200 A through the thin leads on these TO-220 devices. So yes, a 180 A part should be OK. It should be rated for at least 80 V; 100 V will give better margins.

I'd say it's an RS-232 port; RS-485 ports usually come with removable displays, or with models with the round colour display. Almost certainly you'll need a special cable to connect it to a real RS-232 connector, e.g. with this pinout. Genuine Voltronics inverters (this seems like a clone) come with this cable. It's not a standard cable.