Coulomb

Actually, only some Axpert settings (I think only 12 and 13, to grid and back to battery) are 1 V granularity. It's really hard to change the granularity, because the settings info is packed into bit fields to be sent over a small number of CAN packets between master and slaves.

My understanding is that the Windows software for the 3 kVA models with no RS232 port is different, and you can (only) update firmware in such models via USB.

No, you can't use USB unless you have a model that has only USB. I think it's just laziness on the part of the manufacturer, but that's the way it is. 5 kVA models have the RS232 port, and you have to use it. So yes, you'll need the serial to USB adapter, or an rare Windows computer with a built-in serial port.

The same post you got the beta version from now has the release version. http://forums.aeva.asn.au/viewtopic.php?p=67551#p67551

You will get minor improvements from 73.00c: a manufacturer glitch where you can't re-edit a setting is fixed, and there will be slight improvements to the accuracy of battery current reporting. But you'll also get the new fan behaviour; see a recent post about this. Perhaps try one inverter on 73.00c, and if you don't like the extra huffing, revert to 72.70c.

Again, a feature, not a bug, this time from the manufacturer. All main firmwares after 72.70 seem to have more aggressive use of the fan: minimum 30% speed at all times (this should be quiet, unless your fans are faulty), and 5/3 as much fan speed due to SCC charging. I can't recall if it's safe to use 72.70 firmware in a 5 kW model, and in any case you'd be missing out on the extra 1 kW of power output. If the inverter is for use in a noise sensitive environment, then it might be possible to replace the fans with quieter models, e.g. as in this AEVA post (but see other fan related posts linked from the index). But note that some models cause intermittent fan locked errors, so this is no longer recommended. I note that you won't be able to parallel the inverters with the firmware as they are. If your supplier tells you that the only option is to downgrade to 4 kW firmware and it's safe to do this, then when you go back to 72.70 or 72.60 (patched or not), then you'll get the quieter fan behaviour. Note that you can't parallel 72.60 and 72.70 (see the parallel compatibility post), so if safe to do so, I'd suggest 72.70c for both.

The latest patched firmware for 4kW/5kVA single MPPT models is out of beta now. No functionality change.

You can. It's possible that the charge bugs contributed to your old battery failing through chronic undercharging, so it would be good to get the best from your new battery with patched firmware. The latest firmware went from BF1 (Beta LiFePO₄) to LF1 (release LiFePO4) yesterday.

An excellent theory. I use pre-charge with a single 4 kW Axpert, let alone three 10 kW inverters in parallel. It costs a high surge power resistor, and a smallish DC rated contactor. The money seems well spent to me, but I never seem to hear pre-charge discussed.

I can't be certain. But my reading is that fault code 56 is not automatically sent to the other paralleled machines. So each of the machines must have measured less than 30 V at the battery terminals. My guess is that your temporary wiring had some sort of glitch whereby there was enough voltage to run the electronics and processor, but not enough to actually run the inverter and power loads.

I count four adjustments per year, one for each season. But I wonder at the usefulness of this. Why not leave it at the winter setting? The mismatch of pitch in the other seasons is partly offset by longer days. Of course, if you need to run air conditioning in summer, that could make it worth two adjustments per year.

It's in the files section of this forum: I see that there is also the firmware for a VMIII model. They almost certainly use the same reflash tool, so if you have trouble opening the .rar file (I use 7-zip, a free download, just Google it), you can open the .zip file for the VMIII by double clicking on it, and use the UPSReflashTool.exe from that zip file.

I've often wondered about that. Yes, there is no transfer time, but there may be a very similar effect. When the grid goes down, depending on where the disconnection happens, there could be a hundred houses of load at the AC input connection. So the AC input, which is initially still connected to the AC output, is close to a short circuit. So the output will go to zero volts until the inverter can disconnect the AC input. Disconnecting a relay might be a little faster than connecting, but it will still be of the order of 10 ms, I would think. But now, different to the non-hybrid setup, the inverter is working into a short circuit for that time, and the input relay opens under that overload situation. That has to be hard on those components. Meanwhile the load sees a similar drop in voltage. Have I got that wrong?

That's not a bug, that's a feature! You''re supposed to look at it like this: The "MS" stands for MaSter. As far as we can tell, "HS" is their attempt at making two seven-segment characters look like "MS", using their rather poor "font". It's impossible to please everyone, but the changes are all meant to be improvements. They should all be mentioned in the AussieView™ manual. If after reading the manual you think that something is unreadable or missing etc, please post specifics. In particular, the keys don't work the same in information display mode. All I can think of is that perhaps some monitoring software is running in the background (e.g. you might have tried WatchPower and thought that closing the window would stop it from running. You need to use the system tray to exit, as explained in the middle of this post (NOTE: the software in that post is not the latest).

I think that this is a good idea. I believe that in Australia, it is mandatory.

It would be good to know how much you reduce that 50% winter penalty by at 45° tilt, compared with 27°.

Looking more through the code, either the behaviour has changed, or I mis-remembered. The main way for the load relay to be off is if you are in standby mode (the inverter switch is off), and you'll always see setting 28 (at least in 73.XX factory and patched firmwares). It silently refuses to change setting 28 if the output relay is on.

Correct. The Parallel Manual states: "*This setting is only available when the inverter is in standby mode (Switch off)." But my recollection is that if the inverter is switched on, you don't see setting 28 at all; it would go from 27 to 29 without stopping at 28. [ Edit: but see next post. ] That's presumably not the same as trying to set it to Sig and having it fail to "stick", or no having the Sig option (only PaL and the three-phase options). I'm surprised by not being able to change setting 28 to Sig with a parallel card installed. But as @ebrsa mentioned, I don't have a parallel system myself, so I can't speak from personal experience. Looking at the code, it seems that if the load relay is on, then you won't be able to change the output mode. That makes some sense, I suppose.

What tilt are your panels? I think that at least some panels should have a pretty large tilt, say 45° (for latitude 27.5°, Brisbane Australia), so production is better in Winter and worse in Summer. That would balance to some extent the longer days in summer. But it's far more convenient to have panels flat against roofs. Here, the roofs are often at less than 25° tilt. What's it like in various parts of SA? I imagine you don't need steep roofs to stop the snow from building up, same as in 99% of Australia.

Indeed, Plonkster. I often have difficulty forcing battery or bypass modes when I want to, but I'm too lazy to set up the Node-RED buttons to make it easier. I rarely need to do it, so the pressure to do it never builds. Basically, to get from bypass/line mode to battery mode, you have to change the priorities of source and charging, then wait for the battery voltage to exceed the Back To Discharge setting (setting 13) before the change actually happens. Maybe ICC changes all three settings to make this happen, then sets setting 13 back to what it was, I don't know. There is also a 10 minute minimum time before it switches (2 minutes with patched firmware), though there seems to be exceptions to this minimum that I haven't bothered to nail down as yet.

Yes, they conflate the terms "bulk" and "absorb", often confusing the reader by using the wrong term. You can set the float and absorb settings temporarily to be the same if you want, just don't have float higher than absorb. The diagram at the top of this forum page that you pasted from the Axpert manual is a bit of a copy and paste error; it's for another product entirely. Crisp correctness in the manuals is sadly not a priority. So for example where it says in small print that the absorb time (T1) is 10x the bulk time (T0), that simply doesn't apply to Axperts.

If the MPPT is blown up, and the seller won't replace it for a reasonable fee, then you may be able to purchase just the MPPT board and replace it yourself, or get the inverter repaired (so they replace the MPPT board for you and you still have warranty). The inverter part, which is the majority of the cost, should still be fine. But you way want to trade up anyway as Pilotfish has suggested. If the MPPT has died, the symptom would be that you won't see the PV panel icon appear on the LC Display, and you won't have a U2 firmware version number, even with correct-voltage PV connected. The inverter part (which drives the display) can't tell the difference between no PV input (e.g. night, not connected yet) and a completely blown up MPPT. So you won't see an error code related to a faulty MPPT / Solar Charge Controller (SCC). I'm speaking from 5kVA model experience here, but I believe that the smaller models behave much the same.

I'm surprised that others haven't offered an opinion. 15 VAC is about 6.5% of 230 V. A lot of electronics these days have what's called a Power Factor Correction (PFC) stage. This is a poor name, really it's just a circuit to make the device look more like a resistor than a diode and capacitor. These will struggle a little to follow the mains voltage jumping about, hence the flickering LED lights, but won't be stressed at all. Old-school devices that are a bridge rectifier followed by a large capacitor and no boost stage don't care about the mains voltage going down, but when it comes up, they'll have a bit of a surge in current. This will cause a slight hit on their life. But as others have pointed out, Eskom mains voltage can fluctuate pretty wildly anyway, so I don't think it's serious. Things like toasters, ovens, incandescent lights, and motors won't care at all, at least in terms of life. I wonder if smaller motors might have a slight hiccup. A stiffer battery and thicker cables can minimise the effect.

I'll just point out that the later Axperts have an inverter-output-neutral to earth relay inside of them, which (at least by default) turns on when the inverter is in battery mode, and off otherwise. These are the ones that came from the factory with firmware version 73.XX, and only happens when 73.XX (factory or patched) is running. (So you lose this facility if you regress to say 72.70 or patched versions based on that). Edit: And I've not seen this documented anywhere, though I could have missed it.

Yeah, my guess too. They must be shockers.