Coulomb

Ah! Yes, it would be. The last two digits are the flash size in kB. I was thinking of the 145 V max Solar Charge Controller firmware, which they squeeze into 48k. Considering that the TI DSP firmwares are 128 kB or 256 kB (parallelable models), they do well to fit it into 60k.

The only 15.xx firmware that I have is 15.61. This is firmware for an MC9S08AC48 processor (not using a TI DSP). It handles 1-3 kW inverters with 12 V and 24 V batteries. Available from a link in this post.

It's the 14 pin header with pin 6 missing, near the DSP chip. You purchase an XDS100 compatible JTAG interface from Ebay, Ali Express or the like. Generally these come with an easy to use flashing program. They are reasonably priced. But the bad news is that genuine Voltronic firmware is password protected, and we don't know the password for these models. It's possible that a clone inverter won't bother with the password, but that's by no means guaranteed. You can erase the entire flash area, but this will erase the boot loader as well. If you have a JTAG interface, you could give up on the bootloader and replace it with a single jump instruction, and always update the firmware using JTAG. But then the Voltronic firmware will detect this and stop working after 60 days. You could patch the firmware to prevent this, but you have to really know what you are doing. It's also not guaranteed that genuine firmware will run on clone hardware. They may have cut corners, or changed some DSP pin assignments to avoid copyright violation. So I'm sorry to say that things are looking grim. That's the downside of buying a clone. I discuss many of the JTAG issues here:

Yes, though the latest available versions are 72.05 and 02.87. They are both available at the link below, or the latest fully patched firmware for 145V max PV Kings is here.

You might need a resistor in your upload cable. Details here. It might be better to try the USB drive update method. See the upload instructions here: https://forums.aeva.asn.au/viewtopic.php?p=87767#p87767

The latest "mainline" firmware for this model is 72.08, available here. The latest "experimental" firmware, some 3 months older but with the "presumed good" MPPT logic, is 72.75, available here. That depends on your electronics expertise. This forum and others have pretty good traced schematics, "usual suspects", and so on. I mentioned some parts to check last post. If you don't have the expertise, then yes you'll need a repairer; fortunately Mecer seem to be reasonably good with repairs.

Fault code 55 is excessive DC in the AC output. There are some components for detecting this connected to CN9 pin 7. Specifically, a resistor chain, a capacitor (making an RC filter to average the inverter output), and back to back zener diodes to limit the voltage in case something goes badly wrong. A fault in any of these could falsely trigger this fault. Of course, it could be legitimate, caused by a leaky IGBT, or bad gate drivers on one or more of the IGBTs, or possibly a few other things. The firmware upload procedure is documented here. To find the right firmware update file for your model, you have to tell us your existing U1 (main, DSP) firmware version.

Actually, on all the control boards I'm familiar with, [ edit: which admittedly are quite old now, ] they are roughly where you have D27-D31. So these control boards seem to differ from the ones I know slightly. There is proper stuff called conformal coating stripper; I have a bottle made my MG Chemicals. But it's probably not worth chasing for this project. I generally leave it in place, and use sharp multimeter (etc) probes or pins and clip leads. Even then it can be hard. For stubborn pads I just reheat that pad with a soldering iron.

It looks to me that the metal parts that the fuse is bolted to have not soldered properly on the bottom of the board. So that would put extra strain on the plated through holes, hence this area of the board overheated. It might just need cleaning up the tarnished metal, and resoldering the bottom of the board around the fuse metalwork. If you can find a replacement fuse, that would be good, as that one looks like it's been partially melted. Certainly the heat shrink tubing has burned. It may well be high resistance, and that might well be the reason for the poor battery voltage measurements. Of course, if you get a new main board, you don't have to worry about that, assuming that the battery terminals, lugs, and wiring are not tarnished or otherwise damaged. But I assume that a new main board would be at least half the cost of the whole inverter.

You're asking for release notes. Since the advent of the clones a decade ago, Voltronic have not produced anything like these, sadly. Some of the early King firmwares had significant bugs, and it seems that 52.01 is fairly recent (a bit over a year old at this point), so I think you'll find it considerably better than say version 51.80. One of these bugs was when the battery was fully charged by solar, the firmware decided that there was no further need for solar, and didn't use any to support the load. I vaguely recall several other weirdnesses to do with solar power.

My only suggestion is to change the battery type to USE, set appropriate limits for float, bulk/absorb, and cut-off voltages, and ignore the SoC% reading on the inverter (it guesses based on voltage, as if the battery being lead acid). Hopefully the battery itself gives some indication of its true state of charge.

That's outside my area of knowledge. I only have 68.04, presumably older than 68.06, and 68.04 is indeed for a 4.2 kW model (as well as 1.5kw and 6.0 kW 48V). My guess is that this is a hacked firmware, and the hardware may not safely handle 4.2 kW continuously. So sorry, I have no upgrade path for this. The 28.01 display firmware is even more mysterious. My notes say that the only 28.xx display firmwares I have seen are for ESS (batteries)! So sorry, I've struck out on both counts. You'll have to ask your supplier for updates.

It could be that the available PV power was such that one just barely passed the current threshold, and the other just barely did not. The factory firmwares all set the current threshold too high (0.5 A); patched firmware reduces that to 0.05 (earlier patches) or 0.1 A (later patches) to help with this situation.

In poor light, you may find that the Maximum Power Point is under 200 V, so perhaps the 150 V version would work better in cloudy conditions. Your USB to serial adapter may not be on COM1, it might be COM5 for example. If you have a compatible USB to serial adapter and the right cables, then yes, that's all you do. The reflash should take about 10 minutes. There is no need to update the display (secondary) firmware in this case. The PV control is all handled by the main DSP firmware.

I believe that these need the LiB or LiC protocol, which is probably not implemented (or possibly implemented but incompletely) in older display firmware such as 02.68.

Warning code 61 is "communications lost" (to the battery's BMS). Why that would happen only at a highish State of Charge I don't know. That sounds like something weird is going on with the BMS to me. Is it possible to update the firmware in the BMS itself? It could be worth asking you inverter supplier for the latest (if any) firmware for the display, which also handles BMS communications. Per a manual I happened to have open, in the section describing warning code 61: "After 10 minutes, inverter will stop charging and discharging to lithium battery." In other words, after 10 minutes of warning code 61, the battery will disconnect, so that explains the fault code 04. Though I though that these models could operate batteryless, in which case I'd expect it to continue powering loads via AC-in or PV if there is sufficient PV power available, and instead merely flash the BP (Battery not Present) warning.

Yes. Each of my patched firmware sets has the original factory firmware from which it is derived, in this case 64.62. Or you can go back to 64.09 that your inverter came with here. No. That requires special equipment and a 128-bit "password" that only Voltronic knows. Yes; you have to choose one of the three patched firmware versions, based on the voltage that your PV string rarely goes below even in poor light. Or you could try all three to see which one seems to work best for you. But unless your PV string is very low voltage (few panels in series), either the 150 or 200 volt variants will likely be best.

Assuming that it's an EASun (legal rebrand of Voltronic) and not Powland (work-alike with different firmware), you should be able to run patched firmware versions x64.62.

Fault code 04 is battery voltage too low. Is this happening only after charging the battery? 60.10 is the latest 60.xx firmware that I'm aware of. 60.95 available from the link below is about 5 months older, but is supposed to quieten down the fans (too quiet for operation in Syria, according to one poster). Perhaps worth a try. If that doesn't fix the fault code 04, then you could always run that battery with battery type USE, and not attempt to connect to the battery's BMS. The only 57.xx display firmware I have is 57.16.

Assuming that this is a rebranded Voltronic inverter: What can happen is that the EEPROM can lose the serial number somehow. People have had success sending a series of commands to re-write the serial number. Monitoring software such as Solar Assistant can then recognise it again. Are you able to send commands to the inverter? You will need a Windows computer, the Voltronics communications tool, and a USB to serial adapter that can connect to your inverter, probably via a cable that came with the inverter (D9 female at one end, and RJ-45 at the other end).

3.5kW is an unusual power rating. So this is possibly a clone. None of the 20.xx firmwares that I have is rated at 3.5 kW. But please tell us what brand of lithium battery you want to connect to. Since it's a communications issue, it could well be the display firmware. 02.68 would be about 4 years old. The latest I know of is 02.87, available from the below linked post. Ignore the King 1 main firmware; King 1 and VM III (older) share the same display/comms firmware.

That's a tough question. All the 11 kW firmwares seem to have this limit. Many of the latest Max firmwares for other models have it, though it becomes less common with the lower powered models (e.g. the latest 7.2 kW firmware that I have does not have this limit). I just don't know enough about how these models regulate the bus voltage to be able to say. It's safer to keep the limit, obviously, even though in your situation it means that you can't charge your very large battery very quickly.

Actually, this is not true. The full bridge can puff a little energy back into the AC-in source to trim the bus voltage, if of course an AC-in source is present. So far: firmware version 90.06 (for an 8kW Max) limits the charge current to 80A if PV voltage is >360V and AC-in is not present, and does not limit otherwise (based on PV voltage; there are several other kinds of derating). Firmware 78.08 (for 11kW Maxs, 24th Jun 2024): if the PV voltage exceeds 360V: charge current is limited to 100A if AC-in is present, and to 90A if AC-in is not present. It also limits the charge current to 100A if the AC-in voltage is greater than 250 VAC, presumably because puffing in energy to that would make the voltage really high. Now that I can follow this part of the firmware code, I can check some more firmwares to see what they do. 90.06 is from 2022; I use it because it's my best commented Max firmware disassemly. Firmware 63.04 (for Max E 11kW non-Twin, 2023): Same as for 78.08. Firmware 46.82 (for Max E 8kW non-Twin, 2023): Same as for 90.06 (both of these are 8kW firmwares). Firmware 72.08 (for Max E 8kW non-Twin, 23rd Oct 2024, latest I have): Same as for 90.06. Firmware 81.08 (for 8kW Max II non-Twin 23rd Oct 2024, latest I have): Same as for 78.08. So it looks like the 8kW firmwares for the E models don't restrict the maximum charge current at all if the PV voltage is > 360 V and AC-in is present. Yet the one Max II model I looked at does! And two firmwares with hex files dated the same day have different algorithms. I'll need to look into this a bit further. Edit: Firmware 82.63 (For Max II 10kW non-Twin, May 2024): Same as for 78.08. Firmware 63.04 (For Max II 11kW non-Twin, 5th Sep 2023, earliest 11kW firmware I have): same as 78.08. Firmware 45.14 (For Max E 7.2kW non-Twin, 5th Nov 2024, latest 7.2kW firmware I have): No derating based on PV voltage that I can find at all! Firmware 50.17 (For King II 5kW 5th Nov 2024, latest King II firmware I have): The firmware is very different, and does not appear to limit the charge current based on PV voltage. Firmware 52.10 (for MKS IV 5.6kW 10th July 2024, latest MKS IV firmware I have): Different, less so than the King II firmware, but also does not seem to derate the charge current based on PV voltage.

It could be that the two inverters are seeing different battery voltages. That could be because of poor battery wiring (e.g. daisy chained battery connection), or because the battery measurement of one or both inverters is poorly calibrated. With low load and low charge, do the two inverters read the same (± 0.1 V, you can always get one Least Significant Digit differing by one) battery voltage measurement? If not, see Calibrating your inverter's battery voltage reading .

That's true. But to me the dangerous situation is when there is high PV power, then suddenly there is a large decrease in load that the PV was supplying, either in whole or in large part. Now as the control system tries to "turn off the tap", that very VI curve means that as the power is reduced, the voltage is increasing. So if the bus voltage is getting dangerously high, then the only option to lower it is to dump some of the energy presently stored in the bus capacitors into the battery. [ Edit: Not quite; see next post. ] I have no idea how models with no battery cope with this. Maybe that's why they want a lowered battery charge limit when the PV voltage is high, so that they always have the option of exceeding their self-imposed limit without exceeding the limit set by the customer. If we allow full current at high PV voltage, then that option is no longer there; to make the bus voltage safe, the firmware would have to exceed the customer set limit, which might damage the battery. The effect of exceeding the limit whenever a large load turned off would be pretty small, I think, and would have negligible effect on battery life. That's the big question. Did they limit the current because they rushed the firmware, or because there is an important safety / reliability / other reason? Unfortunately, I don't know the answer. If you watch your bus voltage in the inverter with the disabled derating after a drop in load and the battery current surge isn't excessive, then it seems to me that it's safe. But there may be other scenarios that I haven't thought of. I still want to go through some other firmwares to see how common this derating limit is. I haven't gotten around to that yet.