Jump to content

Axpert MKS II firmware - 71.82 - August 2020


Recommended Posts

I've flashed my 2x Axpert MKS II inverters with the latest firmware (71.82).  I've attached the firmware to this post for interest sake.  Whatever you do with it is at your own risk.

I used this cable and it worked the first time on both. Note this cable goes from your computer to the serial cable that came with the inverter (see picture):
https://www.takealot.com/ce-link-usb-2-0-to-rs232-db9-serial-male-a-2m-cable/PLID51855970

The inverters remained connected via parallel cable during upgrade.  I turned one on and the other off, which sets the one turned on to master.  I then flashed the master.  Did the same to the other one.  Note sure if this process made any difference I'm just posting it anyway.

Mostly followed the instructions from @Coulomb from this link. Didn't reset all settings with watchpower afterwards. Note files on this link should NOT be used with MKS II, the instruction is just the same: https://forums.aeva.asn.au/viewtopic.php?title=pip4048ms-inverter&p=64096&t=4332#p64096

Note it seems like the MKS II doesn't have SCC firmware to upgrade.  On the inverter display I can see the inverter version (U1) but scrolling up or down doesn't show a SCC version (U2).

 

WhatsApp Image 2020-08-11 at 21.39.52.jpeg

WhatsApp Image 2020-08-11 at 21.38.23.jpeg

MKS II Firmware 71 82.zip

Edited by pierre.
Link to post
Share on other sites

Hi Pierre

Are these the SOL-I-AX-5NB Axpert MKS II models with 450Vdc capability?

What would you say is the most noticeable changes / advantages to upgrade? (I have two in parallel on version U1-71-50) 
 

Regards
Jaco P

Edited by Jaco P Bloem
Link to post
Share on other sites
Posted (edited)
6 minutes ago, Jaco P Bloem said:

Hi Pierre

Are these the SOL-I-AX-5NB Axpert MKS II models with 450Vdc capability?

What would you say is the most noticeable changes / advantages to upgrade? (I have two in parallel on version U1-71-50) 
 

Regards
Jaco P

Yes it's the 450V models.  I upgraded them a few hours ago so I don't know how they perform yet.  I'm hoping it solves this issue.

 

 

Edited by pierre.
Link to post
Share on other sites
15 hours ago, Jaco P Bloem said:

What would you say is the most noticeable changes / advantages to upgrade?

From a binary code perspective, the differences from 71.80 to 71.82 appear to be:

1) In SolarVolt1Chk(), the "tolerance" for PV low loss and return are multiplied by 10. That means in effect that instead of needing to see a deviation for 5 calls in a row (0.1 seconds) with no exceptions, it now wants to see 50 calls in a row with no exceptions (1 second). My guess is that this is the change in 71.81.

2) In FaultMode(), there is a new case to a large switch statement; servicing fault code 11 (PV over volts). In this case, new code sets bRestartFlag (making the fault restartable), if GetBusOverCount() < 3 and also _g_uwSolar1HighLoss is false. This seems to allow some PV over-voltage events to no longer be "fatal" faults; after 10 seconds, it's possible to exit fault mode and continue operation, unless it was a "bad" overvoltage.

3) There is extra code in FaultMode() [ edit: just before the main loop in FaultMode() ] to count a fault code 11 [ edit: in addition to fault code 08, as before ] as a bus-over-voltage (i.e. GetBusOverCount() will return 1 more).

I have to dash to an appointment now; I just wanted to get this down before I forget it.

Edited by Coulomb
Link to post
Share on other sites
18 hours ago, Coulomb said:

2) In FaultMode(), there is a new case to a large switch statement; servicing fault code 11 (PV over volts). In this case, new code sets bRestartFlag (making the fault restartable), if GetBusOverCount() < 3 and also _g_uwSolar1HighLoss is false.

Either I'm misunderstanding the code (quite possible), or this can never do anything. In order to be processing fault code 11, _g_uwSolar1HighLoss has to be true (=1).

Edit: from the post below, looks like I'm misunderstanding 🙂 

Edited by Coulomb
Link to post
Share on other sites
2 hours ago, Jaco P Bloem said:

Pierre, just to make sure, do you have 2 PV sets consisting out of 9x 360W Panels in Serie - 1 set per Inverter?

@Jaco P Bloem I have 9x 340W in serie per inverter.  My PV voltage generally goes between 310V and 390V.  What do you have?

Feedback in terms of the upgrade: I can confirm that my problem has 90% been solved at the cost of 5% less PV generation.  As far as I can tell the inverter now tries to squeze out less power from the panels, but is much more stable.  The PV voltage doesn't have sudden drops and my AC out doesn't drop below 220V, where it previously dropped to 190V.  There is another forum topic where a user also reported less PV generation when upgrading from 71.70 to 71.80.  I've upgraded from 71.71 to 71.82.  I think the big change came when going to the next major version 80.

When looking at the chart below, pay attention to where the blue and yellow lines meet, at those points the voltage tended to become unstable and then the yellow and blue sticks to each other for larger periods by getting stuck in the unstable state.

Before:

image.png.7927e530093588490fcb7a7e8462ed8b.png

 

After: At the one point just before 12pm I increased the load slightly on purpose to see what it does when PV generation and load is equal.  It became slightly unstable but output voltage remained above 220V.

image.png.1cc428c642c21c61dbb79139396c3469.png

Link to post
Share on other sites

..one thing I do not have is good monitoring software! Are you using ICC on the Raspberry Pi?

My first installation (2019) is 12 x 330W Canadian Solars, the array is two parallel strings of 6 in serie each.

I have now added the second inverter with 8 x 360W Panel array all in serie to the second inverter - running in parallel with the first Axpert.

Hats of to the mppt chargers referencing different input voltages (223V DC on the one and 317V on the other), and still maintaining equal power drawn. When the load exceeds the capability of one of the PV arrays, then only will the first array ( my old set) supply more than the other - to a point where both give max for the time of day.. 

Just by watching the Displays and WatchPower I do not see major fluctuations (227 to 233 Vac which is 1.3%) but I will have to record the Voltages to see if I have a similar drop - that you and others have noticed.

It's difficult to comment on your graphs as one must compare them with the Voltages and also I am sure you have Battery usage and charging as well (?). 

Thanks for sharing! 

((PS: I see a 10% PV power increase in my old array each time I wash the panels - it's dusty in Bloem! 😅))

Jaco P

Link to post
Share on other sites

Hi @Pierre and all

After measurements using WatchPower ("no cloud, poor sample rate - put laptop in 'no sleep mode' and export to Excel" method).. the output Voltages on my 2 parallel Axperts inverters do seem to be stable at around a 1% fluctuation, for this day at least. The batteries were below 47V (setting 12, to give app. 20% left on the LiFe's) just after 6am so the inverter switched to line (grid) mode.
The rest of the day consisted out of battery charging (20Amps x 2), app 1kW from each array.. back to battery mode just before 9am..and normal home loads (washing day)..

 

549505779_WatchPowerVolt13Aug2020.JPG.9515fa3c533d0e73a351f9edc23b5c23.JPG

(sorry about the picture quality! Output Voltage 1 and 2 is the same as they are bridged on the Inverters.)

714028727_WatchPowerPower13Aug2020.JPG.06dd1171c931bda8add56f15dd86e3cf.JPG

If zoomed into the Voltages during high and low loads:

1683814985_WatchPowerVoltdev13Aug2020.JPG.3e65a40329cf02bc3ed8a301819b4d9b.JPG

The fluctuation is exaggerated i.t.o the scale on the right, but the actual (orange and green) outputs are stable in both Line and Battery mode.

One should expect greater stability in line mode as the fault level (firmness) of the Utility is much greater than the installation (you cannot beat the massive rotating machines of Eskom 😁 ) but to be fair the variance in both cases seem similar around 1%. The home loads are quite 'rouge' especially fridge compressors! The power spikes (related to voltage dips) are noticeable during the night (and day) - also they are more noticeable to the naked eye in battery mode - lights dim/dip. In line mode you will seldom 'see' a fridge compressor switch on - again due to the fault level, or the firmness of the network.

This is also the reason why more battery packs are more stable - they add more firmness. If you run the Inverter without a battery (as these models can), you run in all sorts of Voltage issues and fluctuations especially when the grid falls away (i.e load shedding) as the fault level from only the panels is very low (poor), so low firmness. 9amp from a serie PV string vs. +100amp from a single battery (example). (I've witnessed this at an installation close by). Similar to you, I have 2 x 100a/h Lithium Iron batteries - wish they were 4 but it is expensive as you know.

Trust your installation is running smooth and stable after the firmware upgrade!
Kind Regards
Jaco P

Link to post
Share on other sites

Thank you for sharing your results.  Although I can see the inverters is having a hard time handling that specific problem scenario, my problem is fixed because the output voltage remain stable.

On 2020/08/12 at 6:41 PM, Jaco P Bloem said:

..one thing I do not have is good monitoring software! Are you using ICC on the Raspberry Pi?

I've created my own software that I will probably make available for free.  It runs an embedded grafana which is what you see in the screenshots.  It runs on a Pi and in future I will probably have a closed beta if you are interested.

Link to post
Share on other sites

"Wat jy meet is wat jy weet "!

Thanks @pierre. I will keep an eye on your posts. Would love to get a Pi in future and learn - the complete 'plug & play' set with ICC is now above R3000 and I am not sure if that is worth it..

I am going to Flash my 2 inverters tonight or tomorrow with the 71.82 version. (I am still on the Old 71.50!) Since I took measurements it turns out I am suffering from the premature float issue. I am using 52.5V Charge and 51.5V Float values (15s LiFeP) - thanks @plonkster I have learned a great deal from your posts and the 'correct but low as possible' settings that I have seen @Coulomb propose, works well for me.

The Axpert Inverters gets the estimated SOC totally wrong - During the day with PV available and there is a low load while the batteries show 100%, the Inverter display battery Voltage (and WatchPower) will jump to over 60V!? I have seen a few ' battery Voltage high' faults recorded and that explains my infrequent early afternoon trips + resets. I am confident this is a measuring issue and not the battery BMS. 

My supplier agrees and Voltronic send a link with the exact same file as posted by Pierre. 

PS: I have an 'old faithful' i7 laptop with a serial port, so will use the Serial to RS232 inverter cable - hope it will run smoothly..

Keep well!

Jaco P

Link to post
Share on other sites
10 hours ago, pierre. said:

  It doesn't (and can't) set metrics such as battery voltage in the inverter itself as those are read only in the inverter so ICC won't resolve that error in case you were wondering. 

Thank-you, I was wondering yes!

The update to 71.82 went well - all seem normal and stable this morning.. Although now it is snowing (possibly just  'ice rain') in the Free State - go figure!

But the PV is doing well, still charging (slower) and supply load through the clouds - no experience with other panels but the Canadian Solar performs well during overcast conditions. Will monitor..

I agree with the ICC issue, I monitor my batteries manually now and use WatchPower, at least to identify faults..

Kind Regards!

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...