Coulomb

Sorry for the very late reply. No, the Voltronic high PV voltage models (450-500 V max PV) boost the PV voltage (if needed, else pass through) and it's used as the DC bus voltage. To charge the battery from PV, this power is then pushed back to the battery by running the DC-AC converter in reverse (effectively making it an AC to DC converter). It also runs the DC-DC stage (bus to/from battery voltage) in reverse. The power flow can change dynamically as loads and available PV power go up and down. If necessary, the buck stage will reduce the voltage to the DC-DC converter stage when charging the battery. Only the 145 V max (and lower) Voltronic models buck directly to the battery voltage. The reason for this disparity is efficiency. Transformerless power converters (which in an Axpert is all of them except for the DC-DC stage) work more efficiently when the ratio of input to output voltage isn't too large. So bucking directly from say 400 VDC panels to 50 VDC battery is an 8:1 difference; it would not be efficient. 130 V bucking down to 50 V (2.6:1) or 120 V up to 360 V (1:3) is acceptable, but you get better performance when bucking down from 80 V or boosting up from 320 V.

The problem is that this is a clone inverter, using stolen firmware. Voltronic got sick of this and put traps in the firmware to cause fault code 90 to come up after 60 days of running, if it detects that it's running in a clone. How long the inverter is running as a clone is stored in the EEPROM. That's why erasing the EEPROM gives it another 60 days. But you also lose all your settings and calibrations.

It's usually panel leakage to earth that is the problem. Though it should not matter if the grid is present or not; only if the inverter is producing AC from any source. Try isolating the PV input to see if that stops the problem. Note that it can still happen at night, as it's the voltage from the PV wires to earth that is the problem, not the voltage from one PV wire to the other. In fact, many users find that this fault only happens at night, and some use a relay to isolate the panels automatically every night.

The service manuals don't appear in normal web sites. They occasionally appear in the same places that firmware can be found. I checked my present locations and don't find any for Infini VIIs, sorry.

Sorry, I know next to nothing about the Infinisolars. There are enough Axpert (off-grid) models to keep me busy and tizzy. Unfortunately, fault code 09 is usually not an easy or inexpensive fix, at least for Axperts. Especially after a burning epoxy smell. Usually it means that IGBTs and/or MOSFETs have fried, usually plus a handful of small components that if not fixed fry the next set of power transistors. Check the file section for Infinisolar manuals. Some Infinis, especially the Infini V series, use similar or nearly identical hardware to the Axperts; if so, the various Axpert schematics and guides can be helpful. Others are totally different in basic circuitry.

That means that the display firmware no longer talks to the main firmware. What was the main firmware before the update? You might be attempting to use the wrong firmware. It might also just be a glitch due to the cheap and sometimes slow opto-couplers. See If You Think Your Inverter is Bricked.

There is some functionality, I think to do with ECO mode, that goes away when Kings are paralleled. Read the manual carefully, in the section describing each setting.

King II firmwares are very confusing. My collection of King II main (U1) firmwares are only 50.xx and 70.xx. I have various display firmwares that are supposedly for King IIs, including one that seems to report as 51.02! So until I get some clarity, I can't make suggestions for what to update to. Sorry! Also sorry for the late response.

I'm amazed and pleased that you didn't need a 128-bit password to unlock the firmware. On many other models you certainly do.

Sorry for the very late reply. Firmware version 68.04 is for Axpert VM III 4 kW or 6 kW models, with Twin outputs. I'm not sure if 55.03 is a Twin firmware or not, but 55.06 is for non-Twins. I don't know what the effect is of running Twin firmware on a non-Twin model; I would not recommend it. I assume that you have found the 68.04 firmware update file; if not, it's in this post. You are probably better off updating to 55.06, the latest 55.xx that I know of, despite the known issues. The VM IIIs don't seem to be very popular, and so firmware update files are rare for this model. Or simply ask your supplier for a firmware update, quoting your problems and perhaps also those reported for 55.06. The same firmware runs both 24 V 4 kW and 48 V 6 kW models, automatically adapting to the hardware it runs on.

Yes. sorry for the painfully late reply. It sounds like actually you may need the "capacitor mod", either do it yourself if you can do it safely, or get someone to do it. There is a small chance that it's firmware related. Here is the latest firmware in case you want to try it. 52.10 is about 8 months more recent than 52.00. Your Kodak seems to be a rebadge of an Axpert MKS IV. This is main (U1) factory firmware version 52.10 for the Axpert MKS IV 5.6kW '28066 Non-twin. Do not use with any other model or version. Use at your own risk. The hex file is dated 10/07/2024. Firmware upload instructions for models with a colour 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. The RAR file password is "flash", without the quotes. MKSIV_Inverter_Main_Reflash_Ver52.10 pw flash.rar

Looks like 90.19 to me. No, it's "live" on this firmware. P.s. sorry for the very late reply, a whole month after you posted. I've been really busy on another project.

My understanding is that it will work, but you won't call the calibrate function that is burned into your processor's Read Only Memory, and your readings may be inaccurate as a result. I don't know the consequences of that. So I don't recommend it, but my guess is that it won't be a disaster. Going the other way (attempting to use a '28066 firmware on an inverter with a '2809 processor) almost certainly WILL brick the inverter. It will attempt to call a calibrate function in non-existent (or at least unprogrammed) Read Only Memory, and this is very early on before even interrupts are enabled.

Your photo shows U3 02 00; I need the U1 firmware version. Sorry for the late reply; I've been really busy.

Yes, battery side capacitors could be dry. But the battery should absorb most of that ripple; I suspect either the battery itself or the battery cables have high resistance. Actually, with the ripple on the DC bus, the DC bus capacitors could be dry as well. Unfortunately, I have no experience with the 12 V models.

Sorry for the late reply; it's been busy here. That's probably the same as running 11 kW firmware on a 7.2 kW model. It will likely work, except that it won't protect itself from loads over 7.2 kW. Twin means exactly that: it has hardware and firmware to support an extra relay and output terminals. You can adjust settings for how and when the extra output comes on.

That's because the manufacturer stole the firmware and didn't contribute to the development costs. Voltronic firmware checks for clone-ness, and if so, brings up fault code 90 after 60 days of runtime. You might be able to rescue it. You should be able to find the existing firmware version by using the arrow keys until U1 appears at the bottom left of the display. Report any digits after that. I note that updating the firmware alone will not fix the issue. You need someone to patch the firmware to work around the clone test. Or get a legitimate firmware from whoever sold you the inverter, if they are still in business.

74.01 has not been patched, but the more recent 74.04 has. Patched firmware version 374.04 implements only the lowered current threshold. I've come to doing the "full set" of patches for later patched firmwares. It's still the original perturb and observe algorithm, but there is a special test for low PV power that they get wrong, so it gets stuck at the low MPPT voltage limit. You can try the above, and if you don't like it, you can revert to the original factory firmware version 74.04, or I guess use the hex file you posted.

Please post the main (U1) firmware version for the other inverter. The 90.xx firmware versions are all over the place. Sadly, the 7.2 kW models are not very popular any more, so there's not a lot of work done for them now. Generally, the latest factory firmwares don't fix the "stuck at 90V" issue. However, 69.70 (from March 2024) has the "presumed good" algorithm, so that's as good as you will get with factory firmware. You could try patched firmware version 269.61, but that's based on even older firmware (November 2022).

You haven't mentioned paralleling boards or cables. Are you using these? Some models have paralleling boards built in, or at least come installed from the factory. Without these boards and cables, the inverters can't co-ordinate properly. Also, some models aren't designed to be paralleled at all, e.g. the VM series. Please post your main (U1) firmware version so I can tell exactly what model these Kodak branded inverters are. Sorry for the late reply.

I'm not familiar with this model, but it sounds like a 24 V model similar to my 48 V models. These inverters have trouble with capacitors drying out. There is a thread on this, see for example the post below. But you or your repairer have to be familiar with electronics, obviously. https://powerforum.co.za/topic/14666-repair-of-axpert-inverters-a-journey-started/page/4/#findComment-150549

I certainly did with my ancient PIP-4048MS inverter (Axpert MKS 5K-48 PF1). Part of that is the fact that the transformer temperature sensor is now more directly in the air flow. But an iGrid model is quite different; I don't think you can conclude much from this.

At a very wild guess: Something corrupted the EEPROM, and now there are charge source settings set to take effect at a particular hour or set of hours. It could also be per-hour charge source priority as well. Yes, it happens at 31 minutes past the hour, but very likely the time was never set properly, and/or the time has drifted so now it thinks that 7:31am is on an hour boundary. Use whatever monitoring software you have, or the front panel, to reset the per-hour charge priority and output source priority. The "stuck at 90 V" issue is mostly fixed with patched firmware; few factory firmwares fix this problem. Voltronic seems to be blind to this and a few other problems with their firmware, for reasons that are beyond me. What main (U1) firmware are you running? There is patched firmware for most Max models to address this issue, and you can do it at home without the help of your supplier.

I don't believe that it matters for most inverters. The ones I'm familiar with have the same speed on both fans. VM models do seem to treat the fans differently, but if you swapped them I don't think it would make a real difference.

Firmwares for the smaller (≤ 1500 W) are scarce. Please post your existing main (U1) firmware version.