Coulomb

Thanks guys, for the very kind comments. And sorry for the late reply; the local humidity commanded my attention!

Looks like I never got the time to thoroughly analyse the difference between 52.29 and 52.31. I now strongly suspect that 52.31 is the later of these two, though apparently one minor revision older than your odd inverter.

My suggestion is to update the two older inverters to 52.31, the second of the two firmwares at this post:

Quote

Will it be better to split the panels so that each inverter gets 4 panels?

You didn't say if you have a high PV voltage model or not. I'll assume that you have the now-more-common high PV voltage model (450-500 V max PV voltage).

For these, 4 panels in series will give you reasonable performance, except in cloudy conditions (you said no shading issues). With a Vmp of 41.5 V, I'm guessing at about 6/5 x 41.5 = 49.8 V Voc @ 25°C, say 7% more in winter or 426 V max. That should work OK for all 5 kW high PV voltage models.

There will be a slight advantage with 8 panels in series on cloudy days, but there would be a slight advantage for two MPPTs (in two inverters) with shading. Since you claim no shading, I think it would be a slight advantage to go for 8 panels in series. It's also a little less wiring, so cheaper and more convenient.

Imp should not be a factor here.

Quote

Can you please tell me at what hex address is the value for the voltage setting , in order to check what is value?

There are several places, and each firmware is different. But for all the 90.06 patches, you can find the voltage at address 3E2B2C. In the hex file, search for ":202B20" without the quotes and the two-byte word at offset C will tell you the voltage. For example, in the below:

the highlighted word is 07D0, which is the hex representation for 2000 (the voltages are represented internally in tenths of a volt, so 200.0 V is represented by the number 2000). So this is a 200 V patch.

I use the Windows program Nodepad++, which has automatic colourisation, and it helpfully bolds every second data word. You can just start at the end and work back; the last byte (here E4) is the checksum, then you have the word at address F, then E, then D, then C which is the one you want.

As a sanity check, it should always have the word 28A9 before it (the opcode for mov @AL, #), and be followed by the word 9642 (a store instruction).

Note that there will be at least two occurrences of the string ":202B20"; the second one pertains to address 3F2B2C, so you want the first one, about 40% from the start of the hex file.

Quote

Can you please send me a 90.09(or other version) version for Max 8kw with minimum voltage of 180 or 200v ?

Can you confirm that yours is an Axpert Max 1 (with the removable display) and the '2809 DSP chip? Details in this post:

If so, you could use one of the existing firmwares for that combination, e.g. the many patched versions of 90.06. There is an entire thread on patching that firmware. I believe that there are several versions with different minimum panel voltages. I'm hoping that we can eventually find a way to get one patched firmware work with all panel configurations.

Quote

Merry Xmas to you all down under!

Seasons greetings to you and yours also!

Quote

an error on one Axpert MKS 5K inverter

It would really help to know

* What exact model of Axpert MKS 5k inverter (e.g. high PV voltage model, or 145 V max PV), and

* What was the error? A fault code? A warning code? Beep? Just won't change the setting?

Quote

I am also having issues that this inverter does not want to charge using both utility and solar.

What is setting 16, charge source priority?

Quote

the last 4 blocks keeps failing

Interesting. I can't find any differences between 90.06 and 90.14 near the end of the hex file.

Was your 90.06 a patched one? But I didn't think that Georg added any new patch code, just changed code in-line as needed.

> now but still get 61 on both RS485 and CAN
What cable are you using? It can't be straight through, it has to be a special one. Note that there are different cables required for early and later Pylontechs, check with the battery supplier for what pinout is required.

> will you please send the 90.06 if you have it
This is available from this forum:
https://powerforum.co.za/files/file/82-axpert-max-8kw-main-firmware-version-90067z/

It is also available in patched form to reduce the severity of the "stuck at 90V" bug, but that's still a work in progress. See the long thread in the inverters forum for details.

Edit: These inverters only talk to Pylontech BMS via RS-485, not CAN bus.

Quote

I am trying to make my battery Pylontech work with the inverter.

Well, updating to 112.19 should enable that. The DSP doesn't have much to do with talking to the battery BMS; that's all done in the display firmware.

If 92.04 won't update, it's possibly because you have a '2809 DSP processor, in which case you'll need 90.06. Or maybe it's actually a clone and is looking for some different reflash command.

Quote

U1 51 00
U2 113 17

confirmed

Huh. My notes say that 51.08 has been seen in an Axpert King 3kW, but that's presumably an Axpert King one (with the 145 V max SCC). But that 13.xx has been seen in a PIP-5048MGX (Axpert MKS III). Generally, those display firmwares are different to Axpert King firmwares.

So sorry, I can't tell what it is.

What did you try to update it to?

Your best bet would be to try to resurrect as in this post, although in this post, it wasn't a completely black screen:

https://forums.aeva.asn.au/viewtopic.php?p=94053#p94053

Attempt to update with a recent 112.xx display firmware.

Ok, I've never heard of that brand, so I was expecting the label to indicate it's a clone or work-alike. The sticker does look genuine. Unfortunately it's not clear whether it's an MKS III or an Axpert King II. What are the present firmware versions (U1 and U2)? Does the display look like an Axpert, with the hexagonal icons? For example:

Can you post a photo of the sticker on the side? 

Sorry, I have no contact with Voltronic at all. 

Quote

What is the advantage of using this opposed to my current version 71.70?

Early Axpert King firmware seemed to have a few issues with solar charging; I'm not convinced that they are all fixed in 71.86. They may be fixed by 72.00, but it's not available for download as yet.

71.86 is obsoleted now by Axpert King main firmware version 71.97.

My brief notes in the files section mention possible improvements with battery BMS, but most of that will be due to progress with the removable display firmware. It may be that the later removable display firmwares make use of or perhaps require functionality that was introduced in 71.86.

Sorry, I don't know the Deyes. Just Axperts, by their various names. 

Quote

I have 2 axpert type inverters

It really helps if you say what model of Axpert. There are over a dozen, each with subtle differences.

Quote

and 2 LIFPO4 48v batteries (15 cell) connected together. 

Are you using a cable direct to the battery's BMS? That makes a huge difference.

Quote

but the excess solar is not used to charge the batteries. Is this the premature float bug as mentioned?

Probably not, depending on many things. The main one is, is your battery's BMS saying "do not charge"? It's unfortunately not easy to know this. The other things are:

* What is your charge source priority setting?

* What is your output source priority?

* What are settings 12 and 13 (back to battery and back to grid)?

Quote

The inverter does have equalisation settings, but would these be applicable to lithium batteries?

It depends. If you have your float voltage setting the same as the bulk voltage setting (this happens if you are using "the cable" and battery type = PYL, for example), then no, this doesn't matter, and the premature float bug doesn't matter.

Quote

I have Voltronic VM2 Infini Mode 3KW 48 Volt

I think you mean an InfiniSolar V II 3K-48.

Quote

one of them has U1 13 21

I have some InfiniSolar firmware, but I know so little about them I can't even figure out where the version number is returned.

Sorry.

Quote

I have a three phase installation with Growatt SPF5000es inverters.

Alas, I don't know much about 3-phase inverters, and not much about Growatts. But they do seem to be based on Axpert designs, so there is some hope of me making sense.

Quote

Two updates went well. However, one inverter didn't accept the 040.06 FW, but did accept the 041.06.

All I can say is, maybe, just maybe they operate like the Axperts, and you can retry the firmware send. It sounds like you already did that, but maybe you need to turn on AC-in power at just the right time, like with Axperts.

See here for all I know about restarting Axpert main firmware updates.

The difference between pins 3 and 4 should be proportional to the AC output instantaneous voltage. The top half of the two voltage dividers are on the main board, the bottom half of the dividers is on the control board.

You won't be able to measure anything there until the full bridge (DC-AC converter, inverter proper) is running.

Yes, or the isolated power supply components, or even the measurement system, including the resistor strings above, but also some op-amps and other components on the control board (follow pins 3 and 4 of CN9 on the control board).

I would have thought that a person having ordinary skill in the art would know these components pretty well. But here are a couple:

As well as the labelled current sensors, the four full bridge IGBTs are at the left, the LC filter are at the extreme left and the large black box near the middle, there is a green common mode choke to the right of the filter capacitor, and two relays can be seen at the lower right.

Under the board, the IGBT gate driver components:

The 5 8-pin driver chips can be seen in the middle, with DSP-side components below, and gate-side components above them. A "transformer" (TX7 on most models) supplying isolated gate drive voltages is to the right. The resistor chains for the output voltage measurement are at the lower left.

See also the traced schematic for a 2013 main board. Also see the relevant service manuals in the files section of this forum.

It so happens that I've just started investigating the inverter soft start process.

It looks like the inverter starts with a low target voltage, and they increase the target voltage by 5 volts every so often (possibly 50 times per second). I'm unclear what the relays are doing while this happens, but I would guess they should all be off. If the inverter voltage fails to reach the nominal voltage less 5 V in a certain time, and a few other conditions, then fault code 53 results.

In other words, the inverter can't reach its intended voltage, as measured by the output voltage op-amp and associated parts.

So fault code 53 could be caused by any failure in the full bridge (e.g. the 4 IGBTs and their gate drivers), the output LC filter, or the current transformer primary is open circuit, or a fault in the output voltage measuring circuit. Some of the EMI filtering capacitors or the MOVs could also be indicated.

"it starts and supplies under voltage"

What you you mean by "under voltage"?

Are you saying that the AC output produces less than 230 VAC?

Or doesn't it even get that far? (It sounds that way).

How long is the display on, after the "lamp test" that turns on all LCD segments?

Do the various power supplies (+12V, -12V, +5V) come on with appropriate voltages?

If you remove the control board, do the power supplies come up and stay up?

There are many Inverex models, and their firmware update files are hard to find. I don't have any, sorry.

My understanding is that they have a different, customised display, so that standard Axpert firmware won't work.