My SolArk 12k install with Enphase Micros

[BentleyJ]

The micros went offline but the good news was the frequency was stable at 61 Hz. I am encouraged that I am going to get this working.

Before UL 1741 added the Supplemental Appendix (SA) provisions. Grid-Tie or Distributed Generation (DG) inverters would have to shut down at 60.5 Hz. The new rules require several ride-through (RT) ranges for both voltage and frequency. Frequency range was increased up to 62 Hz BUT between 60.5 and 62.0 Hz its only 299 seconds of RT before shut down. Based on the above, you will probably need to either get the Sol-Ark frequency down below 60.5 Hz or set the micoinverters to an Off-Grid profile if they are to stay on and work reliably. The attached is PDF has some good technical info about the particulars of UL 1741 SA.

 

[Ampster]

Based on the above, you will probably need to either get the Sol-Ark frequency down below 60.5 Hz or set the micoinverters to an Off-Grid profile if they are to stay on and work reliably. The attached is PDF has some good technical info about the particulars of UL 1741 SA.

Thanks. I remember see grid profiles on one of my Enlightent apps, but now can't find it. I want to see if I can get it working first without messing with grid profiles. It has been cloudy the last few days so I have not tried another test.
EDIT: Actually a cloudy day might be a good day for a test because a kW or two of the AC coupled solar might be easier for the SolArk to manage than all 6 kW of production.

 

[Ampster]

Several of the reasons I upgraded to a SolArk were closed communication and higher charging current for my pack.
Recently some change has caused me to reevaluate the importance of having a higher current charging rate. My pack is 42 kWhs so I am not even close to a 05C rate at 175 Amps (9.4 kW).
I often refer to part of my system as behind the meter and that is because I have added some capacity beyond the amounts allowed and specified in my NEM agreement. As a result I am mindful that my export does not exceed about 7.5kW because that is the only external way my power provider would ever know my system is larger. I have two EVs, changed my Heating source to a mini split so my overall Net generation would never be significant. In fact, last year I was a Net consumer of about one mWhr of energy but had a dollar credit. That is my sweet spot because I only paid the minimum charges.
When I connected my Orion BMS to the SolArk, the Orion began controlling the charge current and the result was that my batteries take longer to completely charge. Before that change they would often be fully charged before Noon and now they often take up until 3PM. The benefit to me is that means I will have a constant load on the system of at least 3 kW at least until 3PM. So even if all the solar on my system is putting out 9kW I have no worries of exceeding the theoretical limit of 7.5 kW. That is one unintended consequence. Once the weather warms up I won't have the overnight heating loads and I will be back to my batteries charging to full by Noon or before. I will then need to find another way to control loads or curtail some of my DC solar. Some of my plans for future panel additions are to orient them in a more westerly direction which will skew production later into the afternoons to better capture generation credits at peak rates.
These are just some of the reasons that make this is an interesting hobby.

 

[SolarFlares]

I just requested a firmware update. I am on 6.2.1.6. The latest is 7.2.2.2
Here is a link for how to request an update.

Software Update

Does your hybrid inverter need a software update? Use the Sol-Ark® software update request form to get the latest updates for your inverter today.

www.sol-ark.com

As posted on another thread there are no changelogs posted so we are on our own as to what to expect.

Anyone else weary of doing an upgrade to such a critical piece of your house with not even a few sentences for a change log? I expect a detailed list with changes, resolved bugs, and still unresolved bugs but would settle for at least a few sentences.
Since I am not experiencing any issues, I already communicated with Sol-Ark I won't be updating firmware until they give me something.

 

[Ampster]

I just noticed something strange on my SolArk battery voltage. I am using closed communication with my BMS so the BMS setting is 54.2 volts. The SolArk voltage is also set to 54.2 volts. The issue is that my battery voltage is spiking to as much as 55 volts for a couple of hours. The only thing that has changed is that with the clearer days and less overnight consumption my pack is generally charged to 100 percent by Noon and this situation occurs from Noon until about 4PM when coincidentally charging from the grid is disabled. Selling to the grid is enabled from 9AM until 6PM

 

[GXMnow]

I wonder if the MPPT charge controllers in the Sol-Ark are not able to fully drop to zero output current??

My stand alone BougeRV charge controller drops to under 1 amp when the batteries are full and it goes to "Float" mode. But I never see it actually hit zero until the sun goes down.

What is your battery bank that you want it down at just 54.2 volts? If you know you have some creep, you could set the charge voltage a tick lowr to compensate. In my case, it is the other way. I know the voltage reads high on the BougeRV controller, so I have it set at least 0.2 volts higher than I want, and it will still never get there. I am still debating if I want to shell out the cash for a Victron charge controller.

 

[Ampster]

I wonder if the MPPT charge controllers in the Sol-Ark are not able to fully drop to zero output current??

I will check that out with Solar Assistance. Only 2 kW of my batteries are DC coupled and the other 6.6kW are AC coupled which should be passing through to the grid.

My stand alone BougeRV charge controller drops to under 1 amp when the batteries are full and it goes to "Float" mode. But I never see it actually hit zero until the sun goes down.

My charge controller is internal so it may be harder to adjust.

What is your battery bank that you want it down at just 54.2 volts? If you know you have some creep, you could set the charge voltage a tick lowr to compensate. In my case, it is the other way. I know the voltage reads high on the BougeRV controller, so I have it set at least 0.2 volts higher than I want, and it will still never get there. I am still debating if I want to shell out the cash for a Victron charge controller.

I previously had it at 54.4 which is 3.4 volts per cell. I dropped it to 54.2 to see if it made a difference.
I have adjusted my TOU so that at 1PM it is allowed to discharge to 95 percent and will see if that makes a difference. In my mind that is just a workaround and I would like to understand the fundemental reason that the voltage setting is exceeded.

 

[Ampster]

The battery started going up to 54.3 just before Noon versus the setting of 54.3. I am curious what the change in TOU settings will have at 1 PM which is 12 minutes from now. I did start a ticket with SolArk.

EDIT: According to Solar Assistant at 1:10 PM the batter was at 54.4 but before that it had peaks at 54.5.
The inverter tab on SA shows it will charge at 95% but it stll shows some current going into the battery. None of this is critcal at these lower voltages but my preference was to take my cells to an average of 3.45 volts per cell which is a pack voltage of 55.2. Anything above that is starting to cause my active balance to work harder because the cells are getting into the steeper part of the charge curve.

 

[Ampster]

I did hear back quickly from SolArk and they dismissed it as something that they do not support in terms of communication with the Orion BMS.
It was the same guy who I had communicated with when I was considerint purchasing a SolArk and had sent and inquiry about what they had described as a partnership with Orion. It is not an urgent problem an I am also mindful of what @robby posted earlier about their move to larger headquarters.
Just to isolate the issue I disconnected my BMS from the SolArk and did not see any immediate result. Tomorrow I will see if the issue resurfaces with the SolArk setting at 54.2 volts without the Orion BMS in the mix. I know there is a calibration issue with the SolArk and my Orion correlates with my VOM but that difference is trivial in terms of this issue. More later.

 

[robby]

Anyone else weary of doing an upgrade to such a critical piece of your house with not even a few sentences for a change log? I expect a detailed list with changes, resolved bugs, and still unresolved bugs but would settle for at least a few sentences.
Since I am not experiencing any issues, I already communicated with Sol-Ark I won't be updating firmware until they give me something.

I am running an older version but that’s because I am Lazy and got a lot on my plate right now.
I have no fear of doing an update because their beta testers seem really good at finding bugs and their programmers are very good at eliminating them.
The main thing is that any update can be rolled back if your not happy with it.

 

[Ampster]

Now a day later and I have unchecked the Li Batt box so there is no closed loop communication from the BMS. I saw the same spike around Noon and the voltage goes over the set point of 54.2 volts up to 54.5 volts even though the current has tapered to 4 Amps. Therefore I can eliminate any issue with BMS communications. This seems to correlate to days where I have had a lot of solar and did not happen on days where the system only produced half of its potential. I reported this to my contact on the trouble ticket and told him this was not high priority since I understood they were in the process of moving to their new digs.
I can't tell when this first showed up but I updated to firmware 6.2.2.2 in mid march and it was after that when i first notice it.

 

[robby]

Now a day later and I have unchecked the Li Batt box so there is no closed loop communication from the BMS. I saw the same spike around Noon and the voltage goes over the set point of 54.2 volts up to 54.5 volts even though the current has tapered to 4 Amps. Therefore I can eliminate any issue with BMS communications. This seems to correlate to days where I have had a lot of solar and did not happen on days where the system only produced half of its potential. I reported this to my contact on the trouble ticket and told him this was not high priority since I understood they were in the process of moving to their new digs.
I can't tell when this first showed up but I updated to firmware 6.2.2.2 in mid march and it was after that when i first notice it.

Have you done some actual measurements both at the Inverter and at the battery?
When I was in open loop mode I had to up the Voltage settings in the Inverter slightly to compensate for losses in the cable.

 

[DougfromdaUP]

I run my 12K in open loop, and the absorption and float voltages are always higher than the setting.
For example, with float set to 54.4, the voltage (as measured by 5 bms and one or two flukes, all in agreement) is .1 to .3 volt higher, at 54.5 to 54.7. This isn't attributable to IR losses, as it is independent of current flow, and I have very low resistance connections and cable and bus.
I called Sol-Ark about it last summer, and they told me it was because of some magic in their control algorithm.
I'm very happy with the performance overall, so I haven't tried to chase it down. I just fudged the voltage set points accordingly.

 

[kennyb]

Could it be a temperature coefficient? setting? Is there one of those on the sol-ark?

My xw used to spike quite badly. I eventually figured out if I set the temp coefficient to 0 it stopped happening.

 

[Ampster]

The problem is not one of calibration. Attached is a chart of Voltage and current which shows everything stable at 54.2 which is the setting. Then at about Noon, Voltage and current start fluctuating for about an hour and then stabilixe at 54.5. This is a pattern that repeats every day, more or less at about the same time. Sometimes at 11:30 and sometimes as late as 12:45 I just sent this PDF to the tech at SolArk because when I looked at it on PVPro the scale of the graph obscured the significance of the deviation. I may be missing a setting that lets me terminate battery charging when current tapers to a voltage setting or after so much time at the CV setting.

That was an option on my Outback Skybox and on many other charge controllers or inverters which others have described. At that point the battery has been at 54.2 for a couple of hours and the charge current is about 20 Amps on a 840 Ahr battery so I am not losing much capacity by terminating the battery charging at that point in time. The SOC is also reading 100 percent so the Coulomb counter has been reset. As you can see on left side of the chart at 4PM, the prior day, my TOU setting kicks in and the battery starts covering the loads even though there is still plenty of solar to do that. That time is when battery charging is turned off. I understand that I could just terminate charging at Noon every day but that would leave me with the possibility on a cloudy day that my batteries might not get to full and that is not acceptable. I would settle for a nice clean taper of current at a constant voltage because that is probably okay for the battery compared to the 0.3 increase of voltage above the setting.

 

[Ampster]

I run my 12K in open loop, and the absorption and float voltages are always higher than the setting.

To test that hypothesis, I might just set the float voltage lower by that 0.3 volts and see if that changes anything. I will wait for the SolArk tech to respond with his interpretation of the chart of voltage and current.
Previously I was in closed loop and it happened the same way. I don't think my BMS communicates a float setting anyway. I took it out of closed loop because the tech told me that they do not support my BMS. I took it out of closed loop to prove to him that it was not a problem with my BMS. By then he found out that closed loop communications was supported but I have just left it because I did not want to be changing too many things in the middle of a diagnostic excercise.

 

[Ampster]

I did set the float voltage and it kept the voltage flat. There was a small blip as voltage and current fluctuated for half an hour but the voltage only went up by 0.1 volt. Clearly there is a bug in the firmware. I pointed it out to the tech and haven't heard yet. I am away from the system for a few days and will let them deal with it.

 

[thewilliam]

... the easiest way for me to add incremental capacity is with micros...

... use the Generator in port for my AC coupled micros ...

So your  original micros are still grid tied and you are adding capacity with PV/micros connected to the generator input of the inverter? No DC PV connected to the Sol-Ark? A simple single line would be a good addition to this thread.

Are you using all IQ7's or are there any IQ8's?

 

[Ampster]

No DC PV connected to the Sol-Ark? A simple single line would be a good addition to this thread.

My original and new micros are grid tied through the generator port..Yes I also have 2kW of DC solar connected. I don't have the ability to produce a single line diagram but this thread gives a good idea how it works. It follows one of the diagrams in the SolArk manual even though at first I had an Outback Skybox. It is a fairly generic configuration. The only difference is with the Skybox the micros were grid tied through the essential loads panel which was fed by the Skybox and is still fed by the SolArk.

Are you using all IQ7's or are there any IQ8's?

No, only IQ7s since the IQ8s do not work on the same network and I would have had to buy another Envoy which was too expensive to just add a few panels

 

[Ampster]

Today I tested AC coupling and got functionality. It is an overcast day and my micros were only putting out 3kW compared to the 6 kW that they put out on a sunny day. I did connect a 4 kW EVSE to one of my EVs to put that load on my SolArk. I used one of the PREPA OF Grid settings on my micros and the SolArk kept the frequency at 61 Hz. I think there was some curtailment of the micros. I also have 2 kW of DC coupled solar connected to the SolArk and that did not look like it was being curtailed. My battery was at 88% SOC and the battery did have to assist to support the load. I may continue the test as the sun is starting to come out now, to see how it behaves when the system is putting out more power than my loads.