Hello friends.

 

Two years ago I bought an Axpert King 5000w and so far so good. I have bought another King to put it in parallel but this one comes from the factory with version 72.00 and it seems to work even better. But the problem is that I need the same version in both, otherwise I can't parallelize them.

I could give the investors the version 71.97 that they have here, but I have a 2-year guarantee and I would prefer not to cancel the guarantee.

So I need firmware version 72.00, could any of you tell me how to get it?

 

can you help me? Thank you

Edited September 22, 20223 yr by apolo122
vamos

8 hours ago, apolo122 said:

I have contacted the store that sold me the first inverter and they only have version 71.90, they are going to ask Voltronic if they give them version 72.00

That's the right way to do it.

When you get it, please post it here, as it's a firmware that has important fixes, and I would like to eventually release a fully patched version. I've not seen this version available anywhere.

Edited September 19, 20223 yr by Coulomb

On 2022/09/20 at 1:55 AM, Coulomb said:

That's the right way to do it.

When you get it, please post it here, as it's a firmware that has important fixes, and I would like to eventually release a fully patched version. I've not seen this version available anywhere.

Edited September 22, 20223 yr by apolo122
vamos

hi guys.

What is the difference or benefits between this version and prior versions ?

if known .

 

12 minutes ago, razzor13bt said:

What is the difference or benefits between this version and prior versions ?

The differences between 71.97 (which has been around since March 2021) and 72.00 (August or September 2022) are confined to two functions. One of these is the largest, most complex interrupt routine, which does a large number of important, time critical things. The other is that strangely the fault code 90 logic has been defeated.

The changes to the interrupt routine are that several blocks of code were moved around. So it's performing the exact same logic, just in a different order, and therefore with slightly different timing. It's very difficult to figure out the effects of this without the facilities of the R&D section, and unfortunately, they don't routinely publish release notes. Occasionally, we find some download sites with very minimal notes, but not in this case. Unfortunately, I can't recall where I obtained this particular firmware update file.

All I can recall is that I had been hearing good things about 72.00 for a while, and wanting to find it so that a colleague and I can release the long-promised fully patched firmware, and being frustrated that it would not appear for months.

I wish it was otherwise, but that's the reality since the advent of the clones.

Weber and I are finally starting the fully patched Axpert King 1 72.00 firmware. On the AEVA 71.97a page, we say:

Quote

The main version should display as "BCa71.97" (non-LFP) or "BFa71.97" (LFP) on Watchpower, and possibly on other monitoring software that displays the full version number as a string.

Can someone running any kind of monitoring software and patched firmware version 71.97a confirm or deny this, please? It would be nice to know, and to document how it really works.

Thanks in advance.

Coulomb what is the main advantage of running the patch firmware. I am still running 71.97 on my King. It is running flawless but one thing is the power production from the panels. I have a 4kw array but will only generate say 2,5kw. Will the patch help with this as well?

 

7 hours ago, Jacques Ester said:

Coulomb what is the main advantage of running the patch firmware.

There are numerous advantages:

* Fixes the premature float bug. This can actually damage your lead acid battery, or just waste available solar power if you have a lithium battery.

* KettleKomp™, which to a degree compensates battery voltage readings for load and charging. The result is less frequent switching to bypass mode with minor overloads.

* When it does switch over to bypass mode, factory firmware leaves it there for a minimum of 10 minutes. Patched firmware makes that 2 minutes.

* IF you have an LFP battery, several of the internal limits are inappropriate. This can cause nuisance low battery warnings, switching to/from float stage too frequently or not enough, etc.

* IF you use the front panel, there are numerous improvements: "font", the ability to choose upper and lower display data independently, etc.

* There are some specialised commands, e.g. the ability to change the maximum charge current without writing to the EEPROM and needlessly wearing it out, and avoiding a 2 minute penalty with zero or very little PV power.

* IF you have paralleled machines, you can power one or more of them down to a lower power state with special commands.

There are a few more that I can't think of right now, see the manuals (near the end of the documentation section of this page) if you are really interested.

7 hours ago, Jacques Ester said:

I have a 4kw array but will only generate say 2,5kw. Will the patch help with this as well?

I doubt it frankly, unless you are caught by the premature float bug.

Thank you for the reply.

Just a few more questions

Must I update the display as well and what are the chance of putting my inverter into a state of brick?

Will I be able to install via the USB?

4 hours ago, Jacques Ester said:

Must I update the display as well

Some of the features will require updating the display firmware. But it should be possible to only update the main firmware, and miss out on some of the features.

4 hours ago, Jacques Ester said:

what are the chance of putting my inverter into a state of brick?

There is always the chance of bricking your inverter. With a fairly new inverter, the chances of bricking are quite small, and some situations that look like a bricking can actually be recovered from.

5 hours ago, Jacques Ester said:

Will I be able to install via the USB?

The main firmware yes, the display firmware no.

Hi all

Sorry to open this again but I am stuck ...

My Axpert King gave me the 32 error and after playing with options as given on the forums (disconnect PV, battery and connect only AC in) my display even died - nothing, just black.

Eventually I managed to update the firmware with mcu_BU2_02.66a - and the display worked again ! They said ONLY the MCU and NOT the DSP, so I stuck to it.

After rebooting the display still shows the error 32 ..

I tried both dsp_BF1_72.00a and dsp_LF1_73.00b - both says they cannot open serial port ?? WTF ???

I tried various combinations of PV off, AC off, Invertor On/Off but nothing works.

I even tried the USB option with the OTG cable - that gives U03 - Document inside the USB disk with wrong format.

Any suggestions ?

 

Edited June 7, 20241 yr by DerickS
spelling

Someone also posted : If updating the MCU (removable display or secondary) firmware, click on the "Update MCU" button. If updating the DSP (main) firmware, click on the "Update" button. DO NOT USE THE "Update DSP" BUTTON on the reflash tool that has two update buttons. Answer Yes to "Are you sure". Nothing will appear to happen for 10-25 seconds, so don't panic. Eventually, new text should start appearing in the box under the progress bar, saying that erasing has been successful and so on. It will take some 10 minutes to complete the flash programming. At the end, if prompted, click OK to exit the reflash application.

I need to update the main firmware - can I do it from this patch - how ? There is no Update button, just MCU and DSP ...???

17 hours ago, DerickS said:

I tried both dsp_BF1_72.00a and dsp_LF1_73.00b - both says they cannot open serial port ?? WTF ???

73.00x is for a completely different model, do not use this for your King!

The reflash tools are very different between the display and DSP; for updating the DSP, you have to use the reflash tool that comes with it.

They obviously had a go at combining the two, but they don't seem to have succeeded, at least last time I tried; that was many years ago.

The DSP reflash tool will have the Update button, and will know how to "tunnel through" the display to get to the DSP.

Thanks Coulomb, noted. Please provide me with the link to "if you think your inverter is bricked" - I simply cannot find it now.

As mentioned above, I made a bit of progress :

 

Eventually I managed to update the firmware with mcu_BU2_02.66a - and the display worked again ! With old PC and RS232 cable.

After rebooting the display still shows the error 32 ..

Then I tried dsp_BF1_72.00a with same RS232 cable - it cannot open serial port Step 1 ???

I also tried the USB option with the OTG cable (with ONLY dsp.hex on the flashdrive) - that gives U03 - Document inside the USB disk with wrong format.

Alternatively - is there any HARD RESET or Factory Default process ?

Edited June 9, 20241 yr by DerickS
Added text

23 hours ago, DerickS said:

Please provide me with the link to "if you think your inverter is bricked" - I simply cannot find it now.

It's near the end of the upload instructions page; direct link here:

https://forums.aeva.asn.au/viewtopic.php?title=pip4048ms-inverter&p=59897&t=4332#p59897

23 hours ago, DerickS said:

I also tried the USB option with the OTG cable (with ONLY dsp.hex on the flashdrive) - that gives U03 - Document inside the USB disk with wrong format.

That option seems to be fairly hit and miss. I suggest just use RS-232.

23 hours ago, DerickS said:

Alternatively - is there any HARD RESET or Factory Default process ?

No. The processor does a complete reset at power-up, though it doesn't wipe any settings (and other important data) stored in EEPROM. But no settings affect reflashing.

It's all about timing.

On 2024/06/10 at 7:46 AM, Coulomb said:

It's near the end of the upload instructions page; direct link here:

https://forums.aeva.asn.au/viewtopic.php?title=pip4048ms-inverter&p=59897&t=4332#p59897

That option seems to be fairly hit and miss. I suggest just use RS-232.

No. The processor does a complete reset at power-up, though it doesn't wipe any settings (and other important data) stored in EEPROM. But no settings affect reflashing.

It's all about timing.

Hi Coulomb. Sorry man, me again ...

I have tried several versions of FW, all of them "can not open serial port" or "can not open serial port step 1"

I am using a straight RS232 cable from an old PC with Win10.

I also tried direct commands. Some like QPI, QPIGS etc does work (gives weird returns), but others like QBOOT or PF returns NAKs.

I am now at the point where I consider replacing boards - would the culprit be the display unit, comms board or control board ?

BTW - I see a suspicious looking connector (black with clamps) on the control board, suggesting some diagnostic tool that can run complete diagnostics and reflash from that level. Any knowledge of such an animal ?

1 hour ago, DerickS said:

I am using a straight RS232 cable from an old PC with Win10.

Old enough that it has an actual RS-232 port?  Yet running Windows 10? Impressive if so.

1 hour ago, DerickS said:

I also tried direct commands. Some like QPI, QPIGS etc does work

That sounds encouraging.

1 hour ago, DerickS said:

(gives weird returns),

Can you post a picture or type an example?

1 hour ago, DerickS said:

others like QBOOT or PF returns NAKs.

Presumably you are providing the CRC characters, e.g. via the Voltronic Communications tool?

Ah. These may not be passed through the display. Try using the ATE1 command, documented here.

Quote

I have tried several versions of FW, all of them "can not open serial port" or "can not open serial port step 1"

Considering your success with some commands, it could still be one of the many gotchas, like not running the correct reflash tool (always use the one that comes with the hex file in the firmware update file), not exiting Watchpower completely, not selecting the correct COM port, something else using the comms port (e.g. the Voltronic communications tool), etc. You can check that a comms port is not in use by running a comms program such as TeraTerm; it tells you which ports are available.

Edited June 13, 20241 yr by Coulomb

17 hours ago, Coulomb said:

Old enough that it has an actual RS-232 port?  Yet running Windows 10? Impressive if so.

That sounds encouraging.

Can you post a picture or type an example?

Presumably you are providing the CRC characters, e.g. via the Voltronic Communications tool?

Ah. These may not be passed through the display. Try using the ATE1 command, documented here.

Considering your success with some commands, it could still be one of the many gotchas, like not running the correct reflash tool (always use the one that comes with the hex file in the firmware update file), not exiting Watchpower completely, not selecting the correct COM port, something else using the comms port (e.g. the Voltronic communications tool), etc. You can check that a comms port is not in use by running a comms program such as TeraTerm; it tells you which ports are available.

Yes, I am using Voltronics communication tool. Watchpower closed copletely. Tried all possible options of Baudrate and different Reflash versions (each with its own file, as unzipped according to version). Here is an example of the Comm toll result :

 

Just now, DerickS said:

Yes, I am using Voltronics communication tool. Watchpower closed copletely. Tried all possible options of Baudrate and different Reflash versions (each with its own file, as unzipped according to version). Here is an example of the Comm toll result :

 

Also, the "potential diagnostics connector" I referred to, is this one :

Axpert control board.bmp

23 hours ago, DerickS said:

Here is an example of the Comm tool result :

The idea is to try QBOOT and PF after the ATE1. When ready go back to normal, use ATE0. I hope I got those round the right way.

[ Edit: For many commands, you need the "Use CRC" checkbox ticked. Maybe that's why QPIGS gave a QPI response. ]

Quote

the "potential diagnostics connector" I referred to, is this one :

I don't know what that connector is for. It's not mentioned in the service manual either, at least not the King 1 service manual I looked at.

The 14 pin header with pin 6 cut off is the JTAG port. They probably use that at the factory to initially program the bootloader and firmware. It flashes the image in seconds compared to 10+ minutes. But you need either the password or the bootloader image to do that. We have neither.

Edited June 15, 20241 yr by Coulomb