Axpert MKS 5K (64V): avoid bad sine waves

[teba]

Hi everybody.

I own two rebrandet Axpert MKS 5K (64v) inverters.
In a parallel thread I read with horror about the problem with the quality of the outgoing sine wave form of the 64v version with 48v batteries (lead acid or 15s LFP) under certain circumstances.

I'd like to start a discussion on how to prevent such situations.

I'm located in Germany and we have 400V 3P AC grid. So one phase to ground is 230V AC. The tolerance is given with ±10%. So all devices should work with voltages between 207V and 253V.

The problem in the inverter is told to be the limited bus voltage. For a proper sine wave we need a voltage of:

output voltage × √2 + 12v

I have seen these bus voltages on my inverter so far (5 month):

Code: Select all

# NumSamples = 205021; Min = 323; Max = 377
# Mean = 354,573980; Variance = 154,802743; SD = 12,441975
 323 - 325 [    10]: ∎  0,00%
 325 - 327 [    16]: ∎  0,01%
 327 - 329 [    59]: ∎  0,03%
 329 - 331 [   111]: ∎  0,05%
 331 - 333 [  1290]: ∎∎∎∎  0,63%
 333 - 335 [  3070]: ∎∎∎∎∎∎∎∎∎∎  1,50%
 335 - 337 [  5456]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  2,66%
 337 - 339 [ 11887]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  5,80%
 339 - 341 [ 13558]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  6,61%
 341 - 343 [ 14194]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  6,92%
 343 - 345 [ 13845]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  6,75%
 345 - 347 [ 14172]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  6,91%
 347 - 349 [ 10464]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  5,10%
 349 - 351 [ 12837]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  6,26%
 351 - 353 [ 10342]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  5,04%
 353 - 355 [ 10076]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  4,91%
 355 - 357 [  4884]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  2,38%
 357 - 359 [  2565]: ∎∎∎∎∎∎∎∎  1,25%
 359 - 361 [   859]: ∎∎∎  0,42%
 361 - 363 [  1692]: ∎∎∎∎∎∎  0,83%
 363 - 365 [  4489]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎  2,19%
 365 - 367 [ 11502]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  5,61%
 367 - 369 [ 21228]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎ 10,35%
 369 - 371 [ 18723]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  9,13%
 371 - 373 [ 12887]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  6,29%
 373 - 375 [  4577]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎  2,23%
 375 - 377 [   228]: ∎  0,11%

You see about 4.9% bus voltages are below the needed voltage of 337V.

My idea is to switch to 220V AC output voltage. I then only need 323V:

220V × √2 + 12v = 323 V

I'd like the experts to make some statements about the pro and cons.

Will the inverter work this way and produce 220V output when the grid delivers 230V or more? I saw 241.3V on the grid.

I'm unsure if it is really bad for attached devices when the inverter switches from battery to grid with a voltage jump of about 20V (from 220 to 240V).

Thanks
Lars

[coulomb]

I'd like the experts to make some statements about the pro and cons.

Incandescent lights will be dimmer, and heaters will be cooler. Your kettle will take longer to boil. But it's a small effect. You'll also get slightly less maximum power output (the current limit remains the same, but the voltage goes down 4%).

Will the inverter work this way and produce 220V output when the grid delivers 230V or more?

I believe so. I have the opposite situation: I have 16S LFP, and an 8:1 transformer, so my bus voltage is high. Also, my mains voltage is nominally 240 V (supposed to go down to 230 V eventually), and it's commonly at 253 V or more.

I'm unsure if it is really bad for attached devices when the inverter switches from battery to grid with a voltage jump of about 20V (from 220 to 240V).

So I have this problem when switching from 230 V output to the mains if it's at 256 V (rare, but possible). I worry about the step in voltage, and how hard it might be on loads, but also the relays inside the inverter doing the switching. The switches are usually pretty smooth, but sometimes they make a moderate thump when switching. I have no idea if this is due to arcing from inductive loads, or perhaps some weird magnetostriction effect due to the suddenly changing current.

So far, I've had no failure of the relays. I've never had computers reset after a switch. As of a few months ago, I now run two inverters in parallel, which at least shares the switching load.

I considered switching to 240 V output to minimise the voltage difference, but our mains voltage is supposed to eventually fall into line with international standards, i.e. be closer to 230 V.

[teba]

Let me ask another question.
What is so bad about these clipped sine waves?
What kind of device has problems with it?

Maybe I can live with these waves 5% of the time.

It seems half o the year I will not see these waves.

Greetings,
Lars