Coulomb

Sorry for the late reply. It could be that the "safety" relay has stuck on, possibly due to welded contacts. You could test this by powering down the inverter completely (including isolating of solar input) and checking for conductivity between AC-in active and AC-out active. You could also check for neutral conductivity. There should be none. Also perhaps check for conductivity from AC-out neutral to earth. For almost all models, this should be present.

I assume you mean C78 and C79. This is possibly because The Sun Pays is a clone, not a Voltronic inverter with other branding. In a standard Voltronic inverter, C116 (at the input to the 7912 -12 V regulator) is 1000μF 25 VW. The capacitors used in genuine Voltronic inverters are so-so, typically rated at 2000h at 105°C. What you'll find in clones will vary tremendously, most likely at the low quality end of the spectrum. I certainly agree that it's worth spending a little more and replacing with decent quality capacitors.

So that's the start of winter, or he start of summer, depending on which hemisphere you are located in (you didn't say). If you mean why is the production different, each string may be getting different solar input. If the strings are essentially identical, then it's possible that the battery voltage calibration of one inverter is different to that of the other. With low loads and low charging current, do the two inverters report the same battery voltage, within 0.1 V? If not, adjusting the calibration could help. If you mean why are they reporting production with no loads, solar input also charges the battery.

I'll presume that you mean connecting neutral-in to neutral-out. The main problem is that if you get it wrong, safety devices like Residual Current Devices won't operate correctly, so you could get shocked and the safety device won't save you. The other main problem is nuisance tripping of these RCDs. Finally, you may find earth conductors bearing some of the neutral current, and they are not supposed to carry any current. Often they will not be sized for significant current, so that could pose a fire hazard. I have no idea, sorry. You presumably have breakers in your switchboard. Do you have RCDs? These will have a "press to test" button on them. Without a proper earthing system, I struggle to understand how these would work. Voltronic inverters do connect neutral in and neutral out via relay contacts, for example when AC-in is supplying or supplementing the loads. They also connect neutral out to earth, when neutral out is not connected to neutral in. It may be possible to disable this; some people point to a screw on the mother board that can be left out for this purpose. I'm not convinced that this is intended by the manufacturer.

I haven't seen one of those in a long time. All I have is version 1.06 from 2015. This is remote panel (not removable display) firmware version 1.06, as pictured in the above post. Use at your own risk. IPSRemote_v1.06.zip

The problem with massive "over panelling" of Voltronic inverters is that they have poor control loops. The more panels that you connect, the more the current changes for a given change of the buck converter modulation. In effect, the system becomes too twitchy, and Voltronics' control loops are prone to overshoots even with nominal panel power. So I strongly recommend connecting no more than 120% of rated MPPT charging power. In this case, the limit will be about 4800W nominal. Even one more pair of 545W panels will exceed that limit. You can however connect more panels using an external MPPT connected directly to the battery. Set the bulk charging voltage in that external MPPT to be just under the inverter's bulk/absorb/CV voltage setting.

The "usual suspects" for fault code 51 seem to be the 10μF SMD capacitors; see PIP inverter repairs and hardware modifications and look for fault code 51. In case you want to save money repairing your existing inverter, or getting someone else to repair it.

This is a clone. Have you perhaps reflashed the main or display firmware with Axpert firmware? 3.6 kW is not a standard rating for a Voltronic model. For whatever reason, the display is no longer talking to the main inverter firmware. Or at least, that's what warning code 32 would mean for a genuine Voltronic.

The display firmware will flip from voltage to SoC% based on whether a BMS connection is found. Immediately after turning on the inverter, it won't yet have connected to the battery's BMS. You might have to wait until the battery icon starts flashing, indicating that the display is talking to the battery's BMS. For those settings, presumably you mean battery voltage.

Sorry, I have no idea. You would be lucky to come across a reader that has tried both models side by side. I would think that the Twin models are still in production, so they're not "old ones", just a slight variant with more relays inside and slightly different firmware. It's possible that the extra relays interfere slightly with the air flow, possibly resulting in the changes that you observe.

I note that 57.6 V for a 16S LFP battery (with nominal voltage of 3.20 VPC x 16 cells = 51.2 V) is 3.60 VPC (Volts Per Cell). This is the usual "absolute maximum, never exceed" voltage, and usually the BMS will cut off the load to save the battery. With inexpensive inverters and their tendency to overshoot voltage targets, it's best to stay well away from that. Personally I use 55.2 V max for my 16S LFP home-made battery. Above that voltage, you get very little additional energy stored, but run the risk of higher degradation and cutoff. Edit: Battery manufacturers love to quote the absolute maximum voltage, because that gives the battery the absolute highest capacity. If they don't use that voltage and some other manufacturer does, they stand to lose sales, and that's the biggest no-no in business.

The latest 60.xx firmware for VM IIIs and IVs 4 and 6kW Twin that I know of is 60.10, July 2024 versus March 2024 for 60.08. If you want the quietest fans, you likely have to go backwards a few months to 60.95 (January 2025). The latest 25.xx firmware for VM III touch button displays that I know of is 25.16 (August 2024), compared with 25.12 from November 2023. 60.10 and 25.16 are available here: Patched firmware versions x60.10 are available a few posts after the above.

If you can set those settings in Soc%, then I believe that the inverter (display) firmware is reading SoC% from the battery. As to whether those values make sense, I think 13 is a bit too low. You must really hate using utility energy. That's my understanding, yes. As for whether the BMS can be so accurate, I don't know. It would have to see very near 100% SoC fairly often to be able to calibrate itself. I suspect that SoC estimation for an LFP battery is often ±5% or so, due to the many factors that affect the apparent capacity of the battery, particularly temperature. And if course if the inverter is estimating SoC based on battery voltage, then it's ±50% at least, but that should not be the case if it the units for settings 12, 13, and 29 are in SoC%.

Ah. That's not what your first post said. There is a separate setting (number 11) for the maximum utility chsrge current, so my guess is that a maximum utility charge current setting command (MUCHGCR) is being sent somehow. Perhaps check the monitoring software. You might also try sending your own such command with say a 50A limit (around 2.5kW).

Please post in English on this forum. Yes, it's best to invert the fans. I run mine "uninverted" too.

I assume that this is at the high voltage end of the DC-DC converter. [ Edit: Sigh. It seems I didn't consider the waveform photo carefully. ] My wild guess would be some of the push-pull transistors between the SG3525 and the isolating transformers. Or possibly those small MOSFETs that are supposed to turn the DC-DC converter off when not needed.

It looks like it's near the 7912 -12 V voltage regulator. Perhaps the diode feeding it is shorted. Electrolytic capacitors don't like being fed AC. Edit: I'm assuming it's the green smallish one, not one of the two giant bus capacitors.

Measured by a multimeter, or reported by the inverter, perhaps in some sort of log? If the displayed battery voltage is significantly different from that of a trusted multimeter when the loads are steady, then either the battery voltage calibration is off, or there is a hardware fault.

When the battery is powering the loads and the battery voktage falls below the "back to grid" voltage, AC-in (utility or generator) powers the loads. This prevents the battery from over-discharging. When the battery voltage rises above the "back to battery" voltage, then the battery becomes the preferred power source to power the loads. This saves utility energy cost. These two thresholds are settings. So you can adjust them to suit your battery's needs, and also your personal preference for battery longevity or minimal utility use. Also, for example you might know or predict poor weather coming up, so you might be prepared to spend for a little more utility use now to keep the battery full in case you need that during the bad weather. In that case you could increase the back to utility setting, or if already using utility power, instead increase the back to battery voltage setting.

You haven't said what brand or type of inverter. For the Voltronic models (Axperts and Infinis), they have to match in model completely. You certainly can't parallel different Voltronic power levels. For other brands, it may be possible, I don't know.

Sorry for the late reply. My guess is that your model is a high PV voltage one (as almost all inverters are these days). If so, there is a high AC (but not sinusoidal) potential between the PV input terminals and earth. This is an unavoidable side effect of the popular transformerless designs with high PV voltage. I'd say you are seeing leakage, or capacitive coupling, or both, from the panel wiring to the roof. It could also be the panels themselves, especially when they are wet. It may be that you happen to get approximately battery voltage at the roof. Be thankful that it's not high enough to be a shock hazard, at least so far. Earthing the panels should provide a path for the leakage to earth. If the leakage is large enough, the inverter should detect it and stop with a fault code, before it damages its electronics.

I had a superficial look at some 39.xx display/BMS firmwares. Both 39.21 and 39.27 process and send on the PMAXDISCHG command to the DSP. But in DSP firmware version 76.60 (the only 76.xx for VM IV that I have), this command is not processed. The VM models, including the VM IV models, have a smaller processor, so they may have to skip some commands to get the code to fit into flash memory. I checked 39.21 first as it is supposed to be the first display/BMS firmware to support BYD protocol. It seems to replace Shoto protocol! Worse, the BYD support is missing in display/BMS firmware version 39.27, which is the latest one that I have. Also, LIA is replaced by BAK again. I have no idea why they do this. There seems to be plenty of flash space available in these processors. I didn't look terribly hard, but I did not see any place where a BMS function set the bit that would transmit a PMAXDISCHG command. So it looks like the pipeline is blocked in two places: the display/BMS code may not be heeding the discharge limit values, and is not sending a PMAXDISCHG command. But the DSP is not processing that command anyway (second blockage). That's assuming that the PMAXDISCHG command is the only mechanism for limiting the battery discharge current.

Ah! Sometimes the number on the chip's sticker is the firmware version number, sometimes, especially for older hardware, it is not. One firmware I have that comes close is 03.27 for a dual MPPT model. But it's compiled for a TI TMS320F28027 (often abbreviated '28027) processor chip, and that comes in a 48 pin package, not 44 as per the photo. I also have SCC firmware version 03.20, but that's for PWM models, and your PCB and sticker both state "MPPT". Sorry again.

No, it must be true! Because: 😉🤔

That's my guess also.