I would carefully and safely measure the PV voltage at the inverter input terminals, and verify that the PV voltage actually goes up from what the panels are when not loaded. If the voltage does indeed go up, then the hardware is definitely faulty. Unless you exceeded the maximum PV voltage or current ratings (the current rating is easy to overlook), then I can't see how it could possibly be your fault.

Hello all,

Is there any command we can send on these inverters in order to bypass the power switch ?

I have found MNCHGC0497 (go standby) and MNCHGC0498 (go normal operations) commands available for some models when patched with DCLC Dynamic Current and Load Control but of course not working with this model.

Would it be feasable to add this feature to this model's firmware?

3 hours ago, boljul said:

Would it be feasible to add this feature to this model's firmware?

No. We don't have such a model, and the patching is complex enough that we'd have to have a machine to test on.

We're both not interested in owning any high PV voltage model that doesn't have insulation monitoring. The MKS IVs (and sadly 90% of the models that Voltronic make) are of this type.

On 2022/12/04 at 11:52 AM, Coulomb said:

Here is patched firmware version 174.03, based on factory firmware version 74.03. Will only work with '2809 processors. For Axpert MKS IV 5.6kW only, NOT FOR Axpert VM IV, which looks a lot like it. The VM IV uses a smaller processor, which doesn't have enough flash space for this firmware.

This patch implements only one of Georg's suite of patches to avoid the "stuck at 90V" bug (which is here a "stuck at 120V bug). It also fixes the premature float bug. It will show as U1 74 03 on the display, but should show as version 174.03 on monitoring software.

[ Edit: Note that the file dsp_173.03.hex is actually version 174.03, so it should have been named dsp_174.03.hex. Sigh. But with the old reflash tool, you don't have to choose the file name anyway. ]

Use at your own risk.

dsp 174.03 3 patched 2809.zip 1.76 MB · 4 downloads

I have an MKS4 (Branded Rebel Max). It has been misbehaving on the digital comms side. I was thinking of trying to reflash firmware. This would do?

15 minutes ago, Kalito said:

It has been misbehaving on the digital comms side.

That sounds more like the solar charge controller cables too close to sensitive electronics problem (specifically, the display, but it could also affect communications, which are on the same PCB). This seems to affect a number of models, including the MKS IV. 

17 minutes ago, Kalito said:

I was thinking of trying to reflash firmware. This would do?

As above, your problem sounds like it is hardware, not firmware. See for example (don't necessarily act on it) the PDF in the below-linked download.

If you do want to try it, there is still the issue of whether you have the right chipset for this firmware. Short of poking about inside your inverter, you should only consider this firmware if your existing main (U1) firmware has the version 74.xx, where xx is a two digit number less than about 20, and (for other readers) it's an Axpert MKS 4 (however named by the reseller), and is NOT an Axpert VM IV, and is not any on-grid model (some of those look exactly like the Axpert MKS IV). It should not be a Twin model either (those require different firmware, at least if you want to make use of the second output).

 

 

1 hour ago, Coulomb said:

That sounds more like the solar charge controller cables too close to sensitive electronics problem (specifically, the display, but it could also affect communications, which are on the same PCB). This seems to affect a number of models, including the MKS IV. 

As above, your problem sounds like it is hardware, not firmware. See for example (don't necessarily act on it) the PDF in the below-linked download.

If you do want to try it, there is still the issue of whether you have the right chipset for this firmware. Short of poking about inside your inverter, you should only consider this firmware if your existing main (U1) firmware has the version 74.xx, where xx is a two digit number less than about 20, and (for other readers) it's an Axpert MKS 4 (however named by the reseller), and is NOT an Axpert VM IV, and is not any on-grid model (some of those look exactly like the Axpert MKS IV). It should not be a Twin model either (those require different firmware, at least if you want to make use of the second output).

 

 

This makes sense as it suddenly started. Was working ok. I have tried getting another comms card from another unit and connecting on but same problem (Error 61). I know its not the cable as the same Pylontech works ok on a King.

Firmware U1 74.01 U2 3000 U3 0000.

21 minutes ago, Kalito said:

Firmware U1 74.01 U2 3000

Did you mean U2 36.00?

Please tell me there isn't yet another display firmware series 😮 .

Mine has U2 30.06 and showing 3+1 dots after U2
Lots of problems with loosing USB connection...
Can / should i use the 36.01 from the downloads section?

2 hours ago, mjmx8k said:

Mine has U2 30.06 and showing 3+1 dots after U2

Ah, yes, 130.06. I've actually see that one before.

2 hours ago, mjmx8k said:

Can / should i use the 36.01 from the downloads section?

I don't think so. I think you need 30.xx. 36.01 is possibly older than what you have now.

Perhaps visit the capacitor mod document in the downloads section instead. The MKS IV is known to have intermittent interference to the display and/or comms.

@Coulomb

i will try the capacitor mod.

please could you make me for the MKS4 a 74 04 where the float bug is fixed and where the MPPT amps is set to 0,20, no changes in the MPPT startup voltage.

7 hours ago, mjmx8k said:

could you make me for the MKS4 a 74 04 where the float bug is fixed and where the MPPT amps is set to 0,20, no changes in the MPPT startup voltage.

I like a customer who knows what they want.

Also that pays (in coffees) in advance; thanks 🙏 . That should be no problem, I just have a few things to get out of the way first.

@Coulomb

no hurry, its not urgent.

about the capacitor mod, MKS4 models since 08/2022 should have this mod ex factory, my supplier told me.

7 hours ago, mjmx8k said:

MKS4 models since 08/2022 should have this mod ex factory, my supplier told me.

Good to know, but there can be a lot of older stock about. Inverters are probably heavy enough to need transportation by ship, which can take months.

Is yours later than that?

On 2023/04/10 at 4:33 PM, mjmx8k said:

could you make me for the MKS4 a 74 04 where the float bug is fixed and where the MPPT amps is set to 0,20, no changes in the MPPT startup voltage.

It turns out that this requires only 3 one-word patches, and I have a new technique for automating much of the hard work of patching.

This is patched firmware version 274.04 for the Axpert MKS IV with '2809 DSP. This firmware does NOT supersede any other firmware, it's a one-off for mjmx8k. If standard MPPT voltage and an MPPT current threshold of 0.20 A suits you, then you are welcome to try it, as long as you have the matching model.

As usual, do not use this firmware with any other model, even the right morel but the '28066 DSP. Use at your own risk.

Firmware upload instructions.

 

dsp_274.04_patched_0.2A.zip

thank you a lot @Coulomb

274.04 installed well, behaves well - its cloudy today 🙂

my MKS4 invereters are production date 08/2022 and should have the capacitor mod already ex factory - based on information from the supplier - yet in progress to be verified.
 

Sticky again on 274.04:

before String #2 off/on

                     Total            #1          #2
PV power     650 W     546 W     104 W
PV voltage                   280.9 V 124.2 V
PV current                      1.9 A     0.8 A
it should theoretically go to higher voltage from 0,2+ A but does not.

after String #2 off/on

                         Total     #1            #2
PV power     737 W     578 W     159 W
PV voltage                 276.8 V     251.9 V
PV current                      2.1 A     0.6 A

5 hours ago, mjmx8k said:

it should theoretically go to higher voltage from 0,2+ A but does not.

You did specifically ask for the MPPT minimum voltage to not be changed. So I'm not totally convinced now that this should work in theory, When I get some time I'll look into this.

Edited April 13, 20233 yr by Coulomb

Yes, you made exactly what i wanted to have as this "change only startup amps" works well for me on both inverter models.:-)

Version 274.03 started to increase Voltage as soon as 0,25A was exceeed this worked well without higher MPPT voltage set.
Version 274.04 start to increase Voltage after MPPT reaches 0,8A for any reason, even you set 0,2A in firmware parameters.
when it passed the 0.8 amps once and then for any reason go back to 0,3A it keeps the higher MPPT voltage nicely.
so maybe thy have added in 74.04 a new / additional parameter for the 0,8A as startup value...

btw, 274.04 seems to be better in terms of daily production. it makes more watts a day.

Edited April 16, 20233 yr by mjmx8k

5 hours ago, mjmx8k said:

Version 274.03 started to increase Voltage as soon as 0,25A was exceeded; this worked well without higher MPPT voltage set.

Good point.

I thought perhaps the difference might be something that was changed in 74.04 compared to 74.03. I have written a script to compare two binaries that are very close to each other (for technical reasons I can't run BinDiff). It found that the only significant change between the two factory firmwares is in one function, and it's all to do with judging the quality of the AC input waveform. If you are using the UPS AC input voltage range setting, it will have no effect at all. So the change in factory firmware is utterly unrelated to anything MPPT, as far as I can tell.

That leaves whether the patch was applied correctly. So I checked that again, and it's correct.

OK, I suppose that there might have been something different that I actually did for 274.03. Looking at the text file, I see that I actually changed the threshold to 0.05 A, which is ¼ the threshold that 274.04 has (0.20 A).

Would you be interested in another patch with a target of 0.05A? But really, I think its performance would be identical to 274.03, as far as MPPT performance goes. I suppose actually doing the flashing would remove all doubt.

On 2023/04/16 at 10:36 PM, Coulomb said:

OK, I suppose that there might have been something different that I actually did for 274.03. Looking at the text file, I see that I actually changed the threshold to 0.05 A, which is ¼ the threshold that 274.04 has (0.20 A).

I just noticed that IF 274.03 is really reacting at 0.25A, that's 5x the current that was targeted. 274.04 seems to be reacting at about 0.8A, which is 4x the targeted value. Maybe the units that @Georg594  and I have been assuming aren't actually hundredths of an amp, but in reality are fiftieths or perhaps fourtieths of an amp.

Then I realised that it's trivial to check; PV current is reported in QPIGS in whole amps, at least according to the protocol documents, and QPIGS values do seem to be accurate. The value is divided by 10 before printing, so that means that the units are in fact tenths of an amp. This agrees with many other "decimal scaled" values. But that means that 274.04 is actually targeting 2.0A, and factory firmware will do its "plummet to 90 V" thing all the way up to 5.0 A! Something else must be happening to rescue the PV voltage from the 90 V zone.

That does vindicate patched firmware such as 274.03, which are actually targeting 0.5 A, which seems like a reasonable figure to me for when the target voltage should no longer be artificially pulled low.

Duh. I wish we had known that a long time ago.

Edited April 18, 20233 yr by Coulomb

@Coulomb yes, lets try it with 0,05 to see the difference.

here is from today and it even looks to be 1A instead of my reported 0,8A

Edited April 17, 20233 yr by mjmx8k

On 2023/04/17 at 8:03 PM, mjmx8k said:

lets try it with 0,05 to see the difference.

To be clear, we're now trying with a constant of 05 instead of factory firmware's 50 and 274.04's 20. We now know that these figures represent 0.5, 5.0, and 2.0 amps respectively.

This is patched main firmware version 374.04 for the Axpert MKS IV with '2809 processor. Its only differences from factory firmware version 74.04 are:

* Displays as version 374.04 in monitoring software

* Fixes the premature float bug

* Uses a current threshold of 05, compared with the factory firmware's threshold of 50. We now know that these units represent tents of an amp.

Do not use with any other model, not even an Axpert MKS IV if it happens to use the '28066 processor. Use at your own risk.

dsp_374.04_patched_05.zip

installed well, went immediately to 200+ volt (end of day)
i will see more tomorrow, but looks very promising

Hello @Coulomb @mjmx8k,

I have the 374.03 patched version installed and having this kind of graph pattern almost every day.

Should I consider this as a sticky at 120v situation and patch to 374.04 ?

Start of a day:

End of a day:

Thanks

 

Edited April 23, 20233 yr by boljul

5 minutes ago, boljul said:

Should I consider this as a sticky at 120v situation and patch to 374.04 ?

Your graphs are looking good to my eye; when they come off 120 V to shoot higher, the current is so low that you're only shooting from ~20 W to ~40 W. A doubling, yes, but still very little in absolute terms.

That said, I'm coming around to the idea that the (actual) 0.5 A threshold (which we thought was 0.05 A) is the "right" way to patch these MPPTs. Fortunately, it's also a minimal change from the original, which makes it quicker and easier to patch. But I don't own one so I don't watch the graph every day.

BTW, did you mean 274.04? I can't find 374.03.