Adding storage to my Enphase system

[GXMnow]

Just got my Mar. to Apr. So Cal Edison bill. Last year, the same billing cycle, I exported 150 KWHs. This year, thanks to the DC panels and my son being away at college, I exported 326 KWHs. That is an increase of 5.87 KWHs every day. Most of that is from the extra production with the additional 2,000 watts of DC panels.

Production should still increase even more as the days get longer. Mid May was my highest production numbers on the Enphase panels. I am hopeful that the extra production and my stored up energy credits will now cover most of the energy needs to run the A/C this summer. This morning, watching the sun come up, the DC panels are covering the load from the inverter for over a half hour before the Enphase system starts making enough to start charging the batteries from "extra" power. At 8:30 am, the XW is still taking 60 watts from the batteries to cover the loads, but the DC system is already pushing 200 watts into the batteries.

I know it is not linear, but I did a bit of math to try and get a rough idea here. From midnight to 8 am the battery voltage dropped from 5.89 to 53.17. A drop of 2.72 volts. There was basically no solar power during that time, so the only current was from the XW drawing power to run the house. The XW drew 5.6 KWHs in that time. 5,600 / 2.72 = 2,058 watt hours per volt at this middle state of charge. If I allow the battery to run down to 51 volts, that is another 53.17 - 51 = 2.17 volts. Rounding down a bit... 2 x 2,000 = 4,000 more watt hours can be taken before the inverter hit's shut down. That is running the A/C for 75 minutes after the sun is down. But I am exporting over 15 KWHs a day. If the battery was fully run down, It would need 5 KWH's more to fully charge. That still leave another 10 KWHs to run the A/C while the sun is up, that is 2.5 hours (150 minutes) of run time. so in theory, I can get over 225 minutes of A/C run time and still have a zero electricity usage from So Cal Edison. Of course, the reality will happen in a couple months. Let's see how it compares to theory.

 

[faccnator]

It’s been almost two weeks since and update. I’ve grown quite fond of these

 

[GXMnow]

It’s been almost two weeks since and update. I’ve grown quite fond of these

I was out of town for work for a full week, and then trying to catch up since I got back. Not too much to report. The system has been running great with no human intervention. Here is a longer post to make up for the lost time.

The whole time I was on my trip, the system had the battery topped up before 11 am, exported over 15 KWHs each day, and only pulled the battery down to a tick under 55 volts keeping my house alive. Since my girlfriend went with me, it was only running the refrigerator, the furnace on cool nights, my network system, 2 PC's doing some background stuff with the monitors blanked, and Dish Network Hopper recording some TV shows for us. The base load of the house was sitting at about 10 amps from the battery, or about 550 watts. That is still 13 KWHs over a 24 hour day. The Enphase panels were producing about 28 KWHs each day, and the DC panels were pushing the battery up to 57.5 volts each day before 3 pm and going into float mode, and that still works out to just about 5 KWHs each day. April 29th, the next day when we got back, the DC system production jumped up to 6.5 KWHs because we consumed power and it stayed in MPPT mode longer to get the batteries pulled up to float. If we used enough power from the batteries, the DC system should be able to produce over 10 KWHs, but until we start running the A/C in the summer, we just have no place to push all the power. While it's tempting to dial it back to export more to the grid for credit, I do not want to throw up any more of a red flag. It's quite obvious that I am making use of energy storage, but it is less obvious that I added 40% more solar panel power. The DC production is only being used over night, and not being exported.

Here is a sample of the data

That is the battery summery of a day while we were gone. They all look very similar. But on the 24th, we had a grid instability that triggered the XW-Pro to disconnect from the grid for a few seconds at 3 AM. Obviously, there was no solar running at the time, but this is where it gets odd. When it went back on grid, it went into charge mode. My PLC is not programmed to look for solar input at 3 am, so it didn't adjust any settings. The current was at 25% from when the sun was shining the day before. The charge current came on at nearly 35 amps. And it ran for a bit over and hour and pushed the battery bank up to the 56.7 volt absorb setting in the XW-Pro by 4:15, long before the sun came up. Here is that trace.

Once it was done charging, it flipped back to inverting and the PLC went back to zeroing the grid power. Once the sun came up, the battery was back to 56.7 again by 9:30 am. The amount of power used was easily covered by extra export again, so it didn't really cost me anything. Here is my billing month so far from SCE.

We used power on the 23rd, before we left on the trip. Then the 24th had the unexpected charge event. 25-28 were just nice sunny days and an empty house. The Enphase production also shows that little drop in production each day, so the house used about the same every day, as expected. We got home late on the 29th, it was a bit cloudy, with less production, and I nuked up dinner. The 30th is right back in line with the days before the trip. May 1st was cloudy all day. I am actually very happy it managed to make enough energy to cover all the loads. Production fell from 28.4 KWHs on Apr 30 to just 14.5 KWHs on May 1.
Here is the SCE trace from the 24 with the 3 am charge event.

So the net for the day was still almost 18 LWHs of export. Because the power from the grid and Enphase system topped up the batteries, the DC system went into float mode even earlier. Even though it was a nice sunny day, the DC system only produced 3.6 KWHs and just sat most of the day.

I am sure I could make it work "better" with a bit more code optimizing in the PLC, but it really is working well. So I have not been very motivated to work on the code. Once the A/C starts cranking hard in the hot summer, I will have a better idea if I need to make any settings changes. I need to open up the outdoor condenser and clean it out before it needs to run. It is real dirty and a lot of bird poop too. The house is insulated well enough, even with the days we have had over 80F outside, the house has not gotten hot enough inside to fire it up yet, but it's coming soon.

 

[faccnator]

Thanks for the update! Really sounds like you have your system dialed in. There’s something satisfying about hearing that the system is just doing well. We get a lot of build threads but then once it’s up and running no one really reports anymore. Anyway thanks for your posts. I have enjoyed them.

 

[GXMnow]

Today is now the 4th in a row with clouds. And today it is also raining on top of it. The panels were getting pretty dirty, so the rain is not all bad.

It is only 2 pm now, so I will get a bit more charge before it's done, but here is the Enphase energy output for last week, and so far this week.

If this was happening without the DC panels, I would be buying a fair bit of grid power. But thanks to the DC panels, even this low production is covering the energy demand of the house. Here is what SCE saw up to yesterday.

The last 3 days, with the half production still managed to net a bit of export. The Enphase graph does show part of today, but the SCE graph shows zero until it is updated mid day tomorrow. And it does make sense. When I produce 30 KWHs, I have been exporting more than 15 KWHs, so producing half would be about the break even point. The extra from the DC system just fills in the gaps.

Today might end up being a net import. The XW is still charging at 400 watts right now, and the Enphase solar is producing about 900 watts. That leaves 500 watts to help run the house, but my day time base load is closer to 800 watts, so I am pulling 300 watts from the grid. As the clouds move and the loads change, I see it go from import to export, just a couple hundred watts. But I know there is already enough in the battery to make it through the high rate 4 pm to 9 pm block. I'll report the final result tomorrow.

 

[GXMnow]

This was a very bad week for solar production here. We had 4 days that fell to an average of less than 50% production. Here is the plot from Enphase.

Don't look at May 6th, that is today, and I grabbed the image at 11:30 am. The sun is out today, so I think we are on target to get about 28 KWHs unless another cloud bank rolls in. But the previous 5 days tells the story. Since the clouds and rain kept it cool, we still have not run the A/C so w almost made it without using grid power. Here is the SCE plot so far for this billing month.

Oops, we used a net of just over 1 KWH on May 5th. But once again, it was all at "Super Off Peak". We ended up running the dishwasher and the laundry washer and dryer all before 4 pm, so if the solar didn't keep up, it would take the cheapest grid power. Here is the hourly plot for that day, May 5th.

As what little sun we had went down, the current needed from the grid ramped up. But then at 4 pm, it used battery power o zero the grid and ended up exporting a bit. But at that scale, it is just over 300 watt hours total for the whole day. If we had one more day like this, we would have been on grid power for most of it as this 5 day span depleted it. Here is the battery summary for the same May 5th.

As the clouds were going by, there were only a few little gaps where it charged for a few minutes at a time. For the 5th day in a row, the battery voltage ended lower than it started. And today, on the 6th, it fell to just 51.4 volts before the sun came up. I have the inverter shut off at 51.0 volts, so it was almost there. But now the sun is finally back out and we are charging at over 4,000 watts between the extra power from the Enphase and the DC panels cranking out over 1,500 watts. Battery voltage has climbed from 51.4 at 8 am to 55.3 by 11:45 am. We have a few spotty clouds so the charge power has dips in it, but it is miles ahead of the last few days.

 

[GXMnow]

I just did a quick systems test.

The battery reached the XW full charge state before 11 am today, and the DC panels pulled it up another volt to it's float mode and it was sitting like that for over 2 hours, exporting about 2,500 watts. The DC system float current had fallen all the way to under an amp as the batteries were full.

I fired up the A/C even though we really don't need it yet. It pulled the expected 7 amps on on leg to spin up the furnace fun, then clicked in the compressor and I saw the grid current jump up to 14 amps on both legs.

At first, the XW only ramped up to cover the furnace fan, and the compressor load was still coming partially from the grid. The I remembered, I block sell to 4 pm.

I pushed the "Sell Block End" back to 3:55 pm. And sure enough, it went to pulling 900 watts from the battery bank to zero the grid power yet again. But here was my real test. The DC panels did see the load on the battery. But these cells are so stiff, the voltage barely dipped at all. The drop in voltage was not even registered in the XW Battery Summary. It showed the same 57.45 volts before, during, and after the A/C test. That was running the A/C for about 6 minutes. And it was blowing nice cold air. The Enphase panels were still providing the bulk of the power needed. That's why it only needed 900 watts from the battery.

The BougeRV charge controller obviously has a significant resistance from where it is measuring the voltage to the actual connection at the battery. So as the voltage did start to dip, even as little as it did, the DC charge controller did ramp it's current back up. But it didn't quite match the load current. I am sure if it ran longer, or if it came on sooner when the voltage was lower, the current would go higher as it was still showing the PV input voltage at 97. But the current still did manage to ramp up from under 1 amp to over 6 amps. It would have taken up to 16 amps to zero the true battery current. But at the 6 amps it was making, the voltage measured inside the BougeRV charge controlled did reach it's set float voltage. So it lowers the current to stay at that set voltage.

Now that the A/C is back off, the small voltage dip is still there, and after 5 minutes, the charge controller was still pushing 5.8 amps trying to pull it back up to it's float voltage setting. So it should put back all the power the A/C system used.

It has now been 15 minutes since the A/C shut off, and the current has dropped more, but it is still over 2 amps. According to the battery summary, my short test used 920 watts for 6 minutes. Just 92 watt hours. Not really enough of a test to see how it will balance out in the long run, but it looks promising. I saw it top 400 watts, so in 15 minutes, it easily returned the needed 92 watt hours.

 

[faccnator]

Really interesting how the two systems interact with each other. Just communicating based on battery voltage

 

[GXMnow]

It got pretty hot on Sunday, so I fired up the A/C again, but this time I let it run for nearly an hour. Here is the battery summary for the day from the XW-Pro.

I started it just before 4 pm. The Enphase system was still making nearly 3 KW. Here is the Enphase trace for the day.

With all of the other loads in the house, The 3 KW from the Enphase panels does not cover the entire load, so the XW-Pro starts inverting from battery to cover it. As the sun is falling, the amount of power needed from the batteries ramps up. It started at just about 15 amps at 57.5 volts (860 watts) from the battery at 4 pm. By 4:45 pm it ramps up to 33 amps at 57.3 volts (1,890 watts) from the battery. Looking at the energy usage from So Cal Edison, it goes like this.

The 4 pm hour sill shows a small energy export. Between the Enphase production, and the XW inverting from battery, it covered the load in the house and left it just exporting a tiny bit. Once I shut off the A/C it exported a bit more from the Enphase panels for 10 minutes, so it looks like I over produced a bit more than reality.

Looking back a the XW battery plot, you can see the battery voltage come back up again. This is from the DC charge controller still pushing 10 amps back into the system to try and keep the battery voltage at 57.5 volts in float mode. But as the sun angle is going away, it didn't keep up as the A/C was running, but over the next hour, it did still put most of it back. I still had some power coming in from the solar out past 7 pm.

I am still a bit torn on the heading angle for the DC panels. Looking at this test, it makes me want to turn them west to be able to help more after 4 pm. But by 6 pm, it is going to have shadows no matter what I do. I will get more total energy by turning the array a bit more towards true south (East of where they are now) but total energy is not a problem at this point as the system is going into float early. I will know more as we get to hotter days.

I still have 1 more day in the So Cal Edison billing month, so I don't have the real bill yet, but here is the daily production for the month so far.

The site is estimating a -$80 bill credit. May 1 to 5 was heavy clouds and even rain. The battery depleted and on the 5th day, I had to use a little grid power. Not too bad at all. So it looks like I can use at least 12 KWHs a day to run the A/C and still zero the bill.

 

[GXMnow]

I had an odd one today.
Here is the battery summary so far.

Of course, I was not home. My girlfriend was here alone. I got back around 5 pm and I just took a peak at the system to check in on it and I see the battery voltage is up well above my normal full charge. Hmmm. It jumped up at 3:46 pm. My girlfriend said she heard a big thump, like something fell over. Right away I was nervous like I had lost a part of the battery bank. At 5 pm there was still enough sun that the house was running on Enphase, and the battery current was very low, and even the DC system was down to virtually no float current. All my check did not show a problem, other than the voltage was high. I did not find anything that fell or tipped in the garage. Without turning off disconnects, the fuses all appear okay. So what happened??

I checked the event logs in the XW and I did find something.
2023/05/18 15:46:03 -700 error 25 AI Over Frequency.

The grid frequency went too high so the XW disconnected from the grid. The contactor in th XW is pretty loud, is that the thump she heard? I think it was, as it fits the time frame, and it lines up with the voltage jump.

The battery was in float at my set voltage. The Enphase AC coupled panels were cranking out over 800 watts, but the house is only using 450 watts. So the XW had to deal with 350 watts of extra power. I think it sent that to the battery bank for a moment which caused the voltage climb. It then did a fast frequency shift to shut down the Enphase panels. But the grid frequency returned to normal in under a minute, so it reconnected to the grid with the battery graph never showing the current spike. Realistically, the current spike should only have been 6 amps at the battery terminal, but I have learned that the XW reported voltage reading lie, a lot. The DC system had already raised the voltage a fair bit, but since it did not know about that extra charge current, it does not report the climbing voltage right and shows it lagging about 0.4 volts. But when it did produce the charge current spike, it then showed the true raised battery voltage, and it jumped the 0.4 volts that it was not reporting. It actually went from 57.45 to 57.83 so very close to the voltage error I see when the DC charging is maxing out.

At 6:44 pm it got warm enough in the house that the Central A/C kicked on. I got out my amp probe and took a bunch of measurements under load. The XW reported it was inverting about 3,400 watts. It showed about 60 amps DC from the battery at 57.2 volts. The current ramped up a little as the voltage dipped. I measured the cable at the XW input and it agreed at 60.0 amps. I then measured the 2 separate battery banks and they were extremely well balanced. The old bank was showing 30.0 amps and the newer bank with longer cables was right at 29.9 amps. And the DC charge controller was adding about 1 amp into the old battery bank, which explains it being a little higher, it normally reads a little lower at night with no solar coming in. The bus bar combining the 2 batteries and the DC charge controller is currently inside the old battery bank. I am building a new DC distribution panel, so I will have individual DC breakers for each battery and the charge controller, but it is not ready yet. I am also thinking about adding a Victron Smart Shunt or something similar in there. If I go with the Schneider Battery Monitor, it will integrate into the same monitoring, but will it be confused by the charge current it does not know about? My JK-BMS can only give me the current and status of the old battery string. I have to think about this one.

 

[svetz]

... is that the thump she heard? ...

Could have been the contactor. If your systems are all working fine my money is on a neighbor smacking into a power pole with a truck. When texting you never see those things swerve out to get you (e.g., she didn't hear the squeal of brakes).

 

[GXMnow]

In this case it was not a telephone pole. We have all underground service here, but there is plenty more to run into. We have had 3 separate incidents of someone driving into a parked car.

There is a street light pole in front of the neighbors house. Hmmm

It's odd that it was a grid frequency high event. Every time I check the frequency here it is within 0.1 Hz 59.9 to 60.1 and it does not seem to move at all. But at 3:45 here, there is still a ton of solar production.

The neighbors softball has hit the house a few times, and it makes a good thump. But what are the odds it happened within minutes of the XW disconnecting from the grid? It's just odd.

The A/C ran 6:42 to 7:12 yesterday, that's 30 minutes. That pulled the battery down a bit as it was later when there was less coming in from the Enphase system. The battery current went over 65 amps. It dropped the battery about 1 volt. Power was right about 3,700 watts. So that should have been about 1,850 watt hours. and for the rest of the 2 hour period, the system was drawing the typical 12 amps from the battery at 56.5 volts, or about 670 watts for another 46 minutes. That would be about 500 more watt hours for 2,350 WH used for the 2 hours. The Energy Comparison graph shows a little less at 965 + 1,300 = 2265 watt hours used for the 6 pm and 7 pm hours. The XW always seems to report less DC power than I am seeing on the AC power side. The voltage readings are close, so I am pretty sure the DC current is just showing a little low. Or the XW-Pro inverter is 104% efficient ;-) This is another reason I want to add a battery monitor.

 

[svetz]

I'm still thinking an outside power event that can cause the issues you saw, transformer blowing up? That's primarily based on the idea that if everything inside is okay it must have been outside.

 

[GXMnow]

I agree, the grid input to my system must have glitched. I am seeing some "Anti Islanding" event between 2 and 4 times a month. Most have been at night when there is no power coming in from the AC coupled solar. I think that makes it a bigger event for my system as it needs to deal with this extra surge of current that was going out to the grid.

 

[GXMnow]

My Enphase Envoy is acting up again, but this time on the internet side.

It stopped reporting to Enlighten yesterday at 5:25 pm local time. But the system was obviously still producing power. And when I log into the Envoy on my local network it comes up, asks for the token, the token works and I see the system is producing power. But it says "Nt Connected to Enlighten". It is obviously on my network, as I can see it local, and even using this

https://envoy.local/api/v1/production/inverters

is working and showing the power from each inverter. So the PLC to the micros is working. But the signal strength is only 2 out of 5 bars now. It was usually 4 or 5 all the time before. I am pretty sure my Envoy is dying a slow death, but Enphase is trying to drag it out until the warranty runs out.

 

[svetz]

What did they say when you called?

 

[GXMnow]

I have not called them yet, I discovered it when I had to leave for work, and I just got back now at 10:30 PM. If it is still down tomorrow, I will call them. The online "help" was of course no help. It says "Your WiFi settings may have changed. Use AP mode to disconnect and reconnect the WiFi with the new settings. Well, I can talk to it over my network just fine, it is at the correct IP address and reporting properly, just not to Enlighten.

 

[GXMnow]

I went through all the steps with the online help and the Envoy lit up the 3 green LEDs, connected to cloud etc. But after about 2 minutes, it went back red. I am currently in a text chat with Yesudas T

Of course, they are saying the same thing.. Go through the steps to reconnect the Envoy to the network. Already did it 3 times. No change. I even ran a network cable out to the Envoy and told it to use that to eliminate a WiFi issue. While configuring the LAN hard wired connection, I did run into a small glitch in my ASUS router. It was not letting me get to the page to assign fixed ip addresses. Every other device on my network is working perfectly. DHCP and DNS calls are all working fine. But I figured I would reboot my network just to eliminate any potential issue there. It takes 10 minutes for this system to come back up as it has to reconnect so many devices. I do have about 50 things on this router. My network is big, but it is well organized and fast. Mostly wired gigabit.

Once the network was back up, I once again, told the Enphase Envoy to switch back from the 5 gig WiFi to 2.4 gig WiFi. That forces it to re run the DHCP call and restart the stack. Same crap, it would not connect to Enlighten. I let it sit over 10 minutes, and no change. Enphase wants the AC line voltage checked at the connector to every iQ7 microinverter up on my roof. To diagnose an issue with internet connection? REALLY?

At this point, I shut off the breaker to the Envoy PCB inside the combiner box, and left it off a solid 10 minutes. Not that it is doing shit now anyways. I then powered it back up and it too another 15 minutes or so with the cloud connection LED going from red, to green, to flashing green, back to red. Still reporting "not connected to Enlighten". But then the Enlighten portal showed an hour of data from earlier today. Then added another bar. 20 minutes later, the Envoy Local home page lit up, "Connected to Enlighten". It went back to not connected again, for another 10 minutes, but then t connected and has been online for about 45 minutes now. The PLC signal strength is once again bouncing around from 3 green bars out of 5, then drops to 2 amber bars, and I lose data from a few panels. Right now it is saying only one panel is producing power, but I see over 3,000 watts coming from the system. So this Envoy is still acting bad. And 3,620 watts, with only one panel reporting power, yeah. It has dropped off of Enlighten 2 more times, but is reconnecting in 5-10 minutes. So of course, now Enphase says it's just fine again. NO IT'S NOT!

I an trying to escalate this to have a supervisor just send me the Envoy board and I'll swap it out. If it does not cure the issue, I will send them back the new one. If it does fix the issue, we know what it is, and they sent the right warranty covered repair part.

At one point I asked the guy in the Enphase live chat if he was a stupid poorly programmed AI chat bot. How many times can you keep asking the same question that has nothing to do with the current problem?

 

[GXMnow]

While I still feel my Enphase Envoy is failing, it might not be the cause of the cloud app reporting problem.

I am still seeing issues with the Envoy losing the PLC data from the iQ7 micros, and I am very confident that is an Envoy issue. I am looking at a filter I can add to isolate the PLC data to just the lines between the Envoy and the iQ's. I found one rated at 25 amps, so it should be fine at my 17 amps peak from the Enphase system. It is an open PC board, so I can remove the parallel capacitors on the Enphase side so it is not shorting out the 900 MHz data signal on that side. It then has L1 and L2 inductors and more capacitors on the grid side. The output side of the Schneider XW-Pro does have a fair bit of parallel capacitance as well, and that could be attenuating the PLC data a bit. Having the inductors of the line filter should allow the Enphase side to see the RF signal and isolate it from the 60 Hz filtering capacitors. I ordered one board for testing and I will let you all know how that goes.

Back to the error not reporting to Enlighten.
I started seeing the drop outs in the data near the last two weeks. It would connect for a bit each day, so it was filling in the data, but then on the 23rd, it just didn't connect for 2 days. Everything else on my network was not having any issues, so I was not suspecting a network issue. And I have not made any network changes either. Not even any new devices. I did several network quality tests with data speed, and ping tests and everything was looking great. But, when I logged into the router, I did see an odd problem. I tried running an ethernet cable to the Envoy to bypass any WiFi issue. But to do that, I had to issue the IP address to a different MAC address. So I logged into my ASUS router. It stores a table of my MAC addresses and I set it to always issue each device the same ip address. It is a great feature. The devices stay in DHCP mode, but I know exactly what IP each device gets. Well, the Web User Interface of the ASUS router was really slow. It even logged me off twice. Hmmm. I rebooted the route and it started to respond better, but it was still slow. It was showing the CPU cores hitting over 95% usage, and the memory was at 90% used as well. And the Web UI locked up again, and I had to reboot the router again.

After the second reboot, without trying to log into the router, I restarted the Envoy yet again, and it took 10 minutes, but it connected to Enlighten and started reporting data again. It stayed working from last night to noon today, and then it disconnected again. Checking my network, again, the Web UI was dog slow on the router. Still, no other device on the network is showing any errors. It seems that something the Envoy is doing is conflicting with the ASUS router now. But it has been working fine for 3 years, and I have not changed any settings. But Enphase did update my Envoy firmware. And I still think something in the Envoy is failing. Could it be sending packets that are causing an issue in the router? I did a little looking, and found a few other people complaining about the crazy slow Web UI in this model ASUS router. There is a new firmware update, but it seems to be a very small step. The change log says "minor bug fixes" and a few security vulnerability fixes. No mention about packet losses or the slow UI.

But when I went to the firmware update page.... There is also a signature file in the router. And ASUS pushed a new signature file to my router on May 13th 2023, a few days before I started seeing these odd errors showing up. Since I did not make any changes, I was not looking for one like that. I really don't like "Auto Update". I want to know when something is changed, so I can track if it causes an issue. I do not see any way I can manually change this signature file. Since I have auto firmware update disabled, it seems the new signature file is not compatible with my older firmware version. I upgraded the firmware from 3.0.0.4.386.48262 to the latest 3.0.0.4.386.51665 skipping 2 versions between. It took about 3 minutes to install the new firmware, and another 10 minutes for the system to reboot and for all my devices to reconnect to the network again. It did retain all my settings, so I didn't have to redo my IP address list.

It has been up and running on the new firmware for about 3 hours now. The network seems better. The Web UI on the outer is now fast. It is only using 65% of it's ram, instead of 95%, the 2 core CPU is only at 10% to 20% usage on each core, only peaking up to 50% once in a while. Before the firmware update, it was hanging above 50% peaking to 100%. And the Enphase Envoy has been connected since the update as well.

I will keep a close eye on it for a few days, but it seems it was an odd combination of a signature file, conflicting with the firmware, and the Enphase Envoy trying to route something that was causing it's port to lock up and not forward the packets to Enphase Enlighten. I wonder if the packets were just filling up the ram in the router?

On a side note, my raw internet data speed also improved a chunk with this router update. Will Spectrum internet notice and readjust my speed cap? I am paying for 200 Mbps down and 10 Mbps up. I am getting nearly 250 down now and 12 up. Before I updated the firmware, it was consistently only hitting 170 Mbps down.

We are having cloudy and hazy again, so the production is well below normal for late May, but thanks to the extra DC panels and the battery bank, I am still easily covering the house load all day. I am still averaging an export of over 7 KWHs each day. But when we get heavy clouds near noon, I did import a little power at the "Super Off Peak" rate. I don't let the inverter us battery power on that lowest tier time of use. If the solar can't cover the house load, it reverts to grid.

 

[400bird]

That's an unexpected result. Good find, hope the solution sticks ?