Ok thanks  I will try later.

So I manage to update DSP firmware with Georg594 version  (ReflashTool_Xseries) and maybe same as add1c7ed55, the tool open Com port UPG yes flash with USB sign on the inverter panel but after a few second it return to main screen doing nothing. I have been waiting for long Time and neither reflashtool or inverter moved, nothing happened. Exact same U1 version.

I suspect I need to use ArenaINV to flash my firmware too but I can't retry for now

Edited October 25, 20223 yr by sharkut

Add1c7ed55 : how perform firmware Coulomb have done for you ?

6 hours ago, sharkut said:

Add1c7ed55: cum a făcut firmware-ul Coulomb pentru tine

Modified 92.04 with 200v that can flash with arena 

Yes this one

On 2022/10/18 at 7:58 PM, Georg594 said:

@Coulomb

I think the original code to reduce mppt voltage when current is below a limit is to archive if one or two modules get shade to pull down the voltage until the bypass diodes will get active and so the other modules can bring full power.....

 

But the code is so bad that there is no check if this routine makes it more worse .....

 

On professional Inverter it's shade management and really highly complex 

 

Maybe we can bring back this function but with intelligent logic 

Hello Georg.

Is it difficult to modify firmware?

Is it necessary to have some special equipment? (software, hardware)

I would like to be able to edit it myself.

I think you need skill in assembly langage no ?

19 minutes ago, sharkut said:

I think you need skill in assembly language no ?

Yes, and DSP assembly language is arguably the most difficult of them all, written for C compilers or hard core Texas Instrument library code designers, not humans. Though possibly VLIW (Very Long Instruction Word) would be worse.

Over the years, a handful of people have asked me to help them become firmware patchers. All have balked at the difficulty when I presented them with examples. Georg started patching on his own, though we collaborate a little now; he's the only new patcher so far. Even my colleague Weber, who does a lot of the hard core firmware patching (e.g. rewriting the data display code for Axpert MKS 1 models) leaves all the disassembling work to me. But he's also an assembly language expert, writing large chunks of code in assembly language, which gets assembled to hex code by TI's Code Composer Studio and hand pasted into the dsp.hex file. He makes life slightly easier for himself by using structured control-flow macros, so we get IF-THEN-ELSE-ENDIF, DO-WHILE-ENDDO, a form of switch statement (ends up as a skip chain of compare and branch if not equals, much like the TI C compiler does), and so on. You have to be aware of processor limitations (which the C compiler automatically guarantees), such as keeping the stack double-word aligned. We learned that one the hard way, and the only debugging realistically available is "cave man" print statements, which usually aren't even print statements (occasionally we can send strings to a serial port). So we'll turn on a display segment, or toggle an LED to say we've reached a certain point, etc. There is the luxury of JTAG debugging, but firmware is very real-time, and if you single step through certain parts of the code, you can leave IGBTs or MOSFETs driving inductors into saturation, etc. This can result in loud bangs. So we usually avoid that.

I don't want to write articles on how to patch, because people will almost certainly get into trouble, and also it might help clone makers to steal Voltronic's intellectual property (the firmwares, which must be very costly to develop) and profit from using it in their clones without paying for the development. But I'm happy to work with anyone who I judge to have enough skills to actually do their own patching. Uncontrolled patching has the potential to not only damage hardware, but potentially to export power to the grid, which could result in some lineman's death. That's an extreme example, but the consequences have to be taken seriously. Changing on-grid firmware could violate regulations, etc etc.

46 minutes ago, sharkut said:

I think you need skill in assembly langage no ?

I updated now the 92.04 of Cuolomb 200v min, and it works perfect for now 

I will give it a try as soon as possible with 92.04 and ArenaINV. If it's work I will report

34 minutes ago, Coulomb said:

Yes, and DSP assembly language is arguably the most difficult of them all, written for C compilers or hard core Texas Instrument library code designers, not humans. Though possibly VLIW (Very Long Instruction Word) would be worse.

Over the years, a handful of people have asked me to help them become firmware patchers. All have balked at the difficulty when I presented them with examples. Georg started patching on his own, though we collaborate a little now; he's the only new patcher so far. Even my colleague Weber, who does a lot of the hard core firmware patching (e.g. rewriting the data display code for Axpert MKS 1 models) leaves all the disassembling work to me. But he's also an assembly language expert, writing large chunks of code in assembly language, which gets assembled to hex code by TI's Code Composer Studio and hand pasted into the dsp.hex file. He makes life slightly easier for himself by using structured control-flow macros, so we get IF-THEN-ELSE-ENDIF, DO-WHILE-ENDDO, a form of switch statement (ends up as a skip chain of compare and branch if not equals, much like the TI C compiler does), and so on. You have to be aware of processor limitations (which the C compiler automatically guarantees), such as keeping the stack double-word aligned. We learned that one the hard way, and the only debugging realistically available is "cave man" print statements, which usually aren't even print statements (occasionally we can send strings to a serial port). So we'll turn on a display segment, or toggle an LED to say we've reached a certain point, etc. There is the luxury of JTAG debugging, but firmware is very real-time, and if you single step through certain parts of the code, you can leave IGBTs or MOSFETs driving inductors into saturation, etc. This can result in loud bangs. So we usually avoid that.

I don't want to write articles on how to patch, because people will almost certainly get into trouble, and also it might help clone makers to steal Voltronic's intellectual property (the firmwares, which must be very costly to develop) and profit from using it in their clones without paying for the development. But I'm happy to work with anyone who I judge to have enough skills to actually do their own patching. Uncontrolled patching has the potential to not only damage hardware, but potentially to export power to the grid, which could result in some lineman's death. That's an extreme example, but the consequences have to be taken seriously. Changing on-grid firmware could violate regulations, etc etc.

Ok. Now I know, that I don't realy want to make my own patched firmware. I don't have enough knowledges. 

I have to rely on your knowledge and your willingness to help.

Thanks.

On 2022/10/17 at 2:33 PM, Georg594 said:

Hi,

changed both mppt min values to 280V

named version to 90.09

 

 

8KW.zip 1.42 MB · 5 downloads

Hi.

Ihave tried 280V version. This version worked pretty well, but I also tried the 200V version and I found that during cloudy skies the MPPT goes lower than 280V. The lowest voltages in my graphs was 261V. (current 5,2A)
I beg you. Please, can you adjust the MPPT lower voltage to 260V for me?
I think this will be the right compromise for my solar system.
Thank you sincerely.

ArenaINV is working for me. Coulomb modified firmware was uploaded well and inverter restart correctly. I need to wait tomorrow now to see change

Arena update is pretty fast, only 6 minutes and fans was at full speed during this

 

12 hours ago, sharkut said:

ArenaINV is working for me. Coulomb modified firmware was uploaded well and inverter restart correctly. I need to wait tomorrow now to see change

Arena update is pretty fast, only 6 minutes and fans was at full speed during this

 

The same to me.

Thank you very much Coulomb and Georg594, it's awesome !

Only few time since sunset and i can already notice great improvements.

I wait a whole day and I will post comparison graph

 

We are at least 2 now who need 28066 flash command, and I can confirm that the old reflash tool never worked for me too since the first time so it seems not to be a mistake of Easun/Voltronic engineer.

 

3 hours ago, sharkut said:

Thank you very much Coulomb and Georg594, it's awesome !

Only few time since sunset and i can already notice great improvements.

I wait a whole day and I will post comparison graph

 

We are at least 2 now who need 28066 flash command, and I can confirm that the old reflash tool never worked for me too since the first time so it seems not to be a mistake of Easun/Voltronic engineer.

 

Thank you Coulomb and Georg594, is working perfect now. And whats is awesome is that never stays on limit 200v, the voltage and amperage grows up together. Only arena works.

Hello Georg594.

Please, if you find some time. Create a version for me with the limit set to 250V. I've been watching the graphs for a long time, and under some weather conditions, the voltage goes below the 280V limit of the 90.09. version.
Thank you very much.

Made a 250V version on both mppt inputs,

 

 

 

can someone confirm if eg. solar input is 1KW , the load is 2 KW on the inverter ( the missing 1 KW comes from battery)

then the mppt regulation cant work if the voltage is higher as 380V .

 

(i think the diode of the boost converter will become active because the inverter from the battery (48v to 360-380V)

will provide maximal 380v (Limitet by firmware) , the boost converter can only make the voltage of the solar inputs higher than it was before .

 

when now solar power is not enough for the load , internal BUS voltage drops to 380V and the mppt cant regulate in this condition ,

only between 380 and the low limit (original 90V)

 

I change my configuration of solar moduls for testing and have now mppt point at 410V ........it drops to 380v if the solar power

is lower than the load on the inverter .

 

if please somebody can test this on his inverter and share results

this seems to be an hardware limit because of the boost converter diode.

8KW.zip

Hello Georg594.

Thank you again.

After check of the datas from last days, Ican tell that you have right.

If solar power is lower that active power (consumption), than bus voltage drops near to 380V (379-385V).

Hope it helps.

5 hours ago, Georg594 said:

A făcut o versiune de 250V pe ambele intrări mppt,

 

 

 

poate cineva confirma daca de ex. aportul solar este de 1 KW, sarcina este de 2 KW pe invertor (1 KW lipsă provine de la baterie)

atunci reglarea mppt nu poate funcționa dacă tensiunea este mai mare ca 380V.

 

(Cred că dioda convertorului boost va deveni activă deoarece invertorul de la baterie (48v la 360-380V)

va oferi maxim 380v (limitat de firmware), convertizorul de amplificare poate face doar tensiunea intrărilor solare mai mare decât era înainte.

 

când acum energia solară nu este suficientă pentru sarcină, tensiunea BUS-ului intern scade la 380V și mppt-ul nu se poate regla în această stare,

doar între 380 și limita inferioară (original 90V)

 

Îmi schimb configurația modulelor solare pentru testare și acum am punctul mppt la 410V ........ scade la 380v dacă energia solară

este mai mică decât sarcina pe invertor.

 

dacă vă rog, cineva poate testa acest lucru pe invertorul său și poate împărtăși rezultatele

aceasta pare a fi o limită hardware din cauza diodei convertizorului de amplificare.

8KW.zip 1,42 MB · 1 descărcare

Makes the same.to.me, i have 430voc, and goes under 380v when i have consummer. This is a problem? How can solve it ?

Here the result, Hazy sun today and an equivalent day before patch

I also notice something since firmware update (DSP and MCU) The LED Bar doesn't respond anymore. Personnaly I don't care but It's just to know if the same thing occurs for you

8 hours ago, sharkut said:

Here the result, Hazy sun today and an equivalent day before patch

I also notice something since firmware update (DSP and MCU) The LED Bar doesn't respond anymore. Personnaly I don't care but It's just to know if the same thing occurs for you

Dont know, mine is disabled since i installed it

Hello @Georg594, i have checked the bus voltage, indeed with the sfotware changed it goes lower than before.

With the 46.08 software it also goes down when load is greather than pv production but the lowes i saw it was around 407V, but it totaly depends on how high the PV production is

With the last software which you made for me with the limit of 175V , i have a print screen of today, lowest i see is 393V, but than again production is low, check the graphs, i will make another one for today a little later .

Hello @sharkut, i have checked about the led bar, by me it is not affected, when the solar PV is active i have white, when on battery i have green, and when on grid i have magenta and it is working still by me.

11 hours ago, sharkut said:

Here the result, Hazy sun today and an equivalent day before patch

I also notice something since firmware update (DSP and MCU) The LED Bar doesn't respond anymore. Personnaly I don't care but It's just to know if the same thing occurs for you

If led bar desnt respond, try to restart inverter (completely off).

It happened to me also few month ago, and restart helps.

On 2022/10/27 at 12:01 AM, Georg594 said:

can someone confirm if eg. solar input is 1KW , the load is 2 KW on the inverter ( the missing 1 KW comes from battery)

then the mppt regulation cant work if the voltage is higher as 380V .

 

(i think the diode of the boost converter will become active because the inverter from the battery (48v to 360-380V)

will provide maximal 380v (Limitet by firmware) , the boost converter can only make the voltage of the solar inputs higher than it was before .

Let me see if I understand the situation.

As you say, PV power is 1 kW, load is 2 kW, battery is providing the difference, so the bus voltage is 380 V. But the MPP (Maximum Power Point) for the panels happens to be say 420 V, and at 420 V, it could be producing say 1.2 kW. So if things were different, you could be needing only 800 W from the battery, so in a sense, 200 W of potential PV power is being wasted.

So because of the MPPT design, it may turn out to be more efficient to configure your panels such that Vmp rarely exceeds about 380 V. In some cases, it might pay to move some panels from one MPPT input to the other. In other cases, you might not get much value out of the last few panels in a string, in sunny conditions. But that's often the case; since available solar power is so variable, it's often necessary to "waste" some potential PV power in good conditions, so that you have more PV power in poorer conditions.

Maybe I should sleep on this one. It does seem wasteful to have 200 W of battery power coming from the battery, when taking off one panel in a string might mean you don't waste that power. Though it might mean that in some even poorer conditions, losing that panel might mean using more energy overall than leaving it in.

Yet if you had a 145 V max PV inverter model, you might get the best of all conditions. I wonder if all string inverters have this limitation? It doesn't seem to be anything silly that Votronic have done.

If the above makes no sense, I claim it's because it's after 2am here. ⌚

[ Edit: "strings in a panel" → "panels in a string". Sigh. ]

Edited October 27, 20223 yr by Coulomb