Hello,

I would like to ask about the possibility of modifying / patching the firmware for my Axpert MAX / MAX II 8000 inverter.

My current firmware versions are:

U1: 72.03
U2: 112.15
U3: 00.00
Model name: MKS2-8000

What I am mainly looking for is a modification of several battery-related parameter ranges, because I am using a custom battery pack with a higher voltage range than the standard 48 V settings allow.

The first parameter I would like to change is the low DC cut-off voltage range. At the moment, the inverter allows this setting only from approximately:

42.0 V to 48.0 V

For my custom battery, I would need to be able to set this value to around:

52.0 V

and ideally up to:

54.0 V

The second part is the charging voltage range. Both bulk/absorb and float charging voltages are currently limited to approximately:

48.0 V to 62.0 V

For my battery, I would need to extend this range to at least:

66.0 V

and ideally up to:

68.0 V

Of course, I understand that the related protection thresholds would also need to be reviewed and adjusted accordingly. I am not looking to simply disable protections, but rather to move the relevant voltage limits consistently, if the hardware is actually capable of operating safely in that range.

I have inspected the PCB and, from what I could identify, I did not find any obvious component in the battery voltage measurement / control path that would be clearly underrated for these voltage levels. That makes me suspect that these limits may be mostly firmware-imposed, but I would like to verify this with people who know these Voltronic/Axpert firmware branches better.

My questions are:

1. Is this type of parameter range modification feasible in the U1/DSP firmware for this generation?

2. Is U1 72.03 / U2 112.15 a known branch, and is it related to the newer MKS2-8000?

3. Are the battery voltage setting limits stored as simple constants/tables in the DSP firmware, or are they tied more deeply into the control/protection logic?

4. Would it be possible to patch only the U1/DSP firmware and leave the U2 display firmware untouched?

5. Is there any known rollback package for the original U1 72.03 / U2 112.15 firmware combination?

6. Is anyone here able to help with such a patch, or point me to the correct firmware branch / tools / person to contact?

I understand that increasing these voltage limits may have hardware and safety implications, so I would like to approach this carefully and not flash anything without first confirming compatibility and having a rollback option.

Thank you for any guidance.

Update – successful battery voltage range patches and current DC BUS investigation

Hello,

I would like to add an update to my original post, as I have since made significant progress with the U1 firmware analysis and performed several real tests on the inverter.

The inverter is:

Model name: MKS2-8000
U1: originally 72.03
U2: 112.15
U3: 00.00
Firmware branch used for testing: U1 72.07
DSP: TI C2000 / C28066

The U1 firmware is stored as Intel HEX, using C28x word-based addressing.

Battery parameter range patches

I was able to identify several battery voltage limits in the U1 firmware. The voltage values are generally stored in tenths of a volt, for example:

48.0 V = 480 = 0x01E0
54.0 V = 540 = 0x021C
62.0 V = 620 = 0x026C
65.0 V = 650 = 0x028A
68.0 V = 680 = 0x02A8

Bulk and float charging voltages

The menu/configuration limits for:

Program 26 – Bulk / C.V. charging voltage
Program 27 – Float charging voltage

were found in the area around:

0x003E38D6
0x003E38E7

The original upper limit was:

62.0 V = 0x026C

I first changed this to:

68.0 V = 0x02A8

However, changing only the menu/configuration table was not sufficient. Additional shared validation constants had to be patched in several other places.

After changing the corresponding shared validation values, higher bulk and float voltages could successfully be written through RS232 and correctly read back from the inverter.

The currently preferred test variant reduces the maximum from 68.0 V to a more conservative:

65.0 V = 0x028A

Low DC cut-off voltage

The original range for:

Program 29 – Low DC cut-off voltage

was approximately:

42.0–48.0 V

The menu/configuration candidate was found around:

0x003E38F1

Changing only this value was not sufficient.

Several additional copies of the 48.0 V validation limit were found in other branches of the firmware. After changing these together from:

48.0 V = 0x01E0

to:

54.0 V = 0x021C

the inverter accepted low DC cut-off values above 48.0 V through RS232.

Therefore, this part of the patch was successful.

U1 versus U2/display behaviour

The U1/DSP firmware accepts and stores the extended values through RS232.

However, the U2/display firmware still appears to contain its own original validation rules and display mappings.

Examples:

• Bulk and float values above the original range can be written through RS232.

• Low DC cut-off above 48 V can be written through RS232.

• The front panel still restricts the original ranges.

• Program 13 can store values above 62 V through RS232, but the display may show FUL instead of the numerical value.

• RS232 readback still returns the actual stored numerical value.

This strongly suggests that the parameter handling is split between:

1. U1/DSP validation and runtime logic,

2. U2/display validation and user-interface representation.

Battery high-voltage protection candidates

I also identified several likely high DC battery protection thresholds.

The firmware contains values matching the manual specifications:

66.0 V high DC / overcharge threshold
64.0 V high DC recovery

Likely runtime candidates include:

0x003E1DA8 – 66.0 V
0x003E1DC8 – 66.0 V
0x003E2138 – 64.0 V

There are also matching profile/default values near:

0x003EFD35
0x003EFDC9

A test variant changed:

66.0 V → 68.0 V
64.0 V → 65.0 V

The purpose was not to disable protection, but to maintain a reasonable margin above the extended charging range.

Current DC BUS investigation

The next issue is the internal high-voltage DC BUS.

From inverter telemetry, the BUS voltage rises with battery voltage and then becomes clamped at exactly:

475 V

The measured relationship is very strong. During charging, the BUS rises gradually with battery voltage, but once it reaches 475 V, it remains at exactly that value for long periods.

This is not an occasional peak. It forms a completely flat plateau in the logged data.

The BUS therefore appears to be actively limited, regulated, or at least reported with a hard upper clamp at 475 V.

Direct 475 V constants

The firmware contains two repeated groups with values:

515.0 V = 0x141E
475.0 V = 0x128E
514.0 V = 0x1414
476.0 V = 0x1298

at approximately:

0x003E71D6
0x003E7255

Because the values looked like possible BUS thresholds with hysteresis, I created a test firmware changing:

475.0 V → 485.0 V
476.0 V → 486.0 V

in both repeated blocks.

However, this test had no observable effect.

The measured BUS still stopped at exactly 475 V.

Current interpretation

At this point, the 475 V BUS limit may be implemented as:

• a raw ADC threshold,

• a PI-controller output limit,

• a maximum PWM duty-cycle value,

• a fixed-point coefficient,

• an analog hardware limit,

• or a limit stored in another representation rather than directly as 4750.

It is also possible that 475 V represents a 5% safety margin below a nominal 500 V hardware limit:

500 V × 0.95 = 475 V

but I have not yet found direct evidence of this calculation in the firmware.

Questions

I would now especially appreciate help with the following:

1. Has anyone mapped the DC BUS regulation section of the 72.xx firmware for the MKS2-8000 / original Axpert MAX 8 kW?

2. Is the 475 V limit known to be:

   • a regulation target,

   • a reporting clamp,

   • an ADC threshold,

   • or a hardware limitation?

3. Do the repeated values around 4750 / 4760 / 5140 / 5150 have a known meaning?

4. Does anyone have:

   • a commented disassembly,

   • known function addresses,

   • a partial firmware map,

   • service documentation,

   • calibration documentation,

   • or schematics for the MKS2-8000 / Model 44 / U1 72.xx platform?

5. Has anyone compared the 72.03, 72.07, 72.08 and 72.75 firmware branches specifically in the DC BUS or MPPT regulation areas?

I can provide:

• the original 72.07 firmware,

• exact Intel HEX changes,

• all patched variants,

• measured BUS, battery voltage and PV power logs,

• screenshots of the 475 V plateau,

• and the complete patch notes.

Any information about the internal DC BUS regulation, ADC scaling, PWM limits or relevant firmware sections would be very helpful.

Thank you.