EG4 6500EX drop outs

[nkhorman]

I'm hoping that someone can help...

I have a few problems;

• RGB led bar just glows barely visible low intensity red on both units. It did work at one point, and I disabled it, but now won't come back on when enabled.
• I have very prominent light flickering when running on Solar-Bat-Utility, wether PV charging or not, but not when running in Utility-Solar-Bat / Solar-Utility-Bat.
  • The light flicker is Incandesent bulb, CFL bulb, LED bulb, and LED shop dual T4 form-factor lights... ie.. the type and quality doesn't seem make a difference.
  • I don't think I have the bonding screw installed in either inverter, based on lack of continuity from neutral to ground when only powered by battery.... no AC In or AC Out wired connections, and the original as shipped firmware, pre-79.71 ( I've forgotten that version number ).

  • I had decided to basically just ignore the flickering as it seemed that no loads stopped working or worked badly other that lights.
• 12/29/2023, I had two spontaneous complete three to five second Load center drop-outs while in Utility-Solar-Bat mode, which I just would not ever expect to happen since the battery pack was 85+% SOC, and PV was nearly non-existent.
  • I had been running in USB for the past couple of very cloudy days... I was conserving my battery charge, since I live in Maine and want it for possible Utility outage.
• 12/29/2023, after switching to Solar-Utility-Bat from USB, I commanded mode change to Solar-Bat-Utilty, and the inverters refused to change.
  • I tried to make the change by both front panel, and WatchPower iPhone app.
  • I was able to sniff the clear-text network traffic and see the WatchPower app request the change, and the inverter "Ack" the request.
  • I had to completely power down the system, battery, PV and Utility feed, to re-boot the inverters. Only then, did it switch back to SBU.
• 1/9/2024 - Load center Drop-outs, 3 or more today on SUB.
  • There where no high current demands when this happened.... < 1000 watts... just coasting

I had it... I've Load-Transfer-Switch'ed from Inverter System power feed to Utility power feed, and put the Inverters in stand-by.
I'm SO upset right now!

I was just starting to get to the point where I really use, rely-upon, and enjoy using the system.

But, now I feel like I've been ripped-off with these inverters, given the numerous product problem complaints in this forum and on youtube, and now that there is no new firmware development... that is tantamount to product EOL... now I basically have expensive bricks.

What can I do to resolve these problems, short of replacing the inverters with something all together different and chalking this up to ... I'm just a sucker... don't by budget friendly inverter(s) ?

I don't want an experimental system... I want something that just works. I have enough other things to worry about without adding this to my life-problems list.

System;

Purchased - 12/2022​

Inverters, panels, batteries​

Initial installation of Inverters and batteries - 6/2023​

Installation of panels - 10/2023​

​

​

Load transition from original house load panel center to new inverter load panel center - 11/2023​

​

Added - 12/2023​

Eg4 Communications Hub​

Chargeverter​

• 2- 6500ex
  • Manufactured 10/2020, according to the stickers on the side of the units.
    • Don't these fall into the unapproved component swap by the assembly service ? that the 79.71 firmware fixes ?
  • 240v / split phase config (2P1 & 2P2)
  • 79.71 firmware on both units.
  • Unit 1 (2P1)
    • PV2 - 1 string of 7 panels - roof mount - Tigo TS4-A-O per panel
  • Unit 2 (2P2)
    • PV1 - 1 string of 6 panels - ground mount - Tigo TS4-A-O per panel
    • PV2 - 1 string of 7 panels - roof mount - Tigo TS4-A-O per panel
• 1- EG4 Communication Hub
  • 1.10 firmware
• 6 - EG4 LiFePower4 batteries
  • V3.32 firmware
• 36 - BlueSun Bi-Facial Mono Perc 440w-460w panels
  • 20 - deployed
  • 14 - planned for summer deployment
• 2 - Tigo TAPs
• 1 - Tigo CCA

 

[Dinobot248]

Seems like you already have the solution based on your comment "What can I do to resolve these problems, short of replacing the inverters with something all together different and chalking this up to ... I'm just a sucker... don't by budget friendly inverter(s) ?"

Right now, Victron Multiplus II inverters just had another price drop. You can get a Victron Multiplus II 48 / 3000 / 120 for about $1200 on Amazon or any other distributor. Since you're running split phase, get 2 inverters. You will need some solar charger controllers. My preference is the 250/100 ($600) or the 150/45 ($220-ish). Each has their own benefit and use case. It will take some "elbow grease" but by the looks of your pictures, you can easily swap them in.

They can be easily setup to automatically draw on the grid when overloaded e.g. if for some reason you pull 8000 watts, the inverters can be easily setup to run on battery with assistance from the grid for as long as you need.

If this is something you are interested in, please tell us and we can guide you through. There are many nuances like your distributor is your tech support so buy from a good distributor.

I went cheap at first but now I have a good working Victron system. It is pricier but when you consider your time, frustration, ease of use, etc, it actually comes out cheaper to go with better equipment.



Also, concurrently start a support ticket with Signature Solar. Wait times are in the 2+ week time frame and any meaningful assistance can take several weeks. You will need to take several pictures and jump through many hoops.

 

[Dinobot248]

The Victron Multiplus II 48 / 5000 / 120v are down to $17xx

 

[fmeili1]

I'm hoping that someone can help...

I have a few problems;

• RGB led bar just glows barely visible low intensity red on both units. It did work at one point, and I disabled it, but now won't come back on when enabled.View attachment 187915
  

I've switched them off because they make a very annoying high frequency noise which I definitively don't want to have. Never tried to turn it on again. I will check and will tell you if it works. But I think this is the least important problem...

• I have very prominent light flickering when running on Solar-Bat-Utility, wether PV charging or not, but not when running in Utility-Solar-Bat / Solar-Utility-Bat.
  • The light flicker is Incandesent bulb, CFL bulb, LED bulb, and LED shop dual T4 form-factor lights... ie.. the type and quality doesn't seem make a difference.
  • I don't think I have the bonding screw installed in either inverter, based on lack of continuity from neutral to ground when only powered by battery.... no AC In or AC Out wired connections, and the original as shipped firmware, pre-79.71 ( I've forgotten that version number ).

  • I had decided to basically just ignore the flickering as it seemed that no loads stopped working or worked badly other that lights.

Thanks to @Adam De Lay about his hint with flickering lights because of specific loads (in his case a laptop power supply), I've searched the load which causes the flickering in my environment. I have bad flickering when an old Vegas style slot machine was running in parallel. The flickering does only occur on the phase where the slot machine is connected and not all LED's on this phase are flickering.

To find out the load which causes the problem, I've switched off breaker by breaker in the house sub-panel while watching the flickering until the flickering stops. After that I knew the room where the problematic load is. After that I've disconnected/switched-off device by device in this room until the flickering stopped.

Later I've identified an additional load which is my 3D printer which also causes the LED flickering (because it was not running so often compared to the slot machine which is usually always running because the are designed to run forever).

My flickering does nearly disappear (with running the problematic load in parallel) when the inverters have higher load >15%.

Without running these problematic loads, I have no more flickering/pulsing LED's - but this is not a solution. I've tried different type of filters placed before the problem causing loads, but without success. Now I think about two options to solve this problem

1. modding the flickering LED's to improve their power supply to make them more resistant to voltage fluctuations (a lot of work, because all my flickering LED's are not LED bulbs with the possibility to change them, they are integrated in ceiling fans).
2. put small DIY mini UPS (online, double conversion) in front of the problem causing loads (higher costs).

So far I don't have the PV panels installed and all my tests are done by battery only, recharged from grid if they getting empty. For long term testing (13 days now in a row), of the system I'm running the whole house (200Amp service) on just the inverters and have no issues so far. The system goes only in bypass mode depending on settings 12 and 13 for "point back to utility" and "point back to battery" to recharge the batteries.

Here are some details which may help you to find differences with your setup:

• six EG4-6500EX in parallel split phase and 60kWh LFP batteries
  • firmware 79.71
  • all bonding screws are in place on all six inverters
  • option 42 set to "ENA"
  • common neutral wiring
  • only using SBU mode
  • AC-in "GND" connected to common ground in the AC-in breaker panel, AC-out "GND" not connected (the are bridged inside the AIO's and I want to prevent ground loops)
  • All battery cables have the same length
  • All AC-in and AC-out wires have the same length to the AC-in breaker panel and AC-out breaker panel
• installed as a "solar-generator" with a manual transfer switch (manual, 2-pole, open transition type) to drive the whole house by grid or by "solar-generator"
• common neutral wiring
• Independent of the transfer switch position, the AC-in is permanently connected to the grid as a backup scenario in case there will be not enough solar and battery power (but I disconnect the AC-in from the grid via contactors which are only activated just before the AIO program setting 12 will get active to reduce grid energy consumption - if grid is just connected with AC-in, each AIO consumes about 50W, which is about half of the idle consumption, from the grid even if grid is not used! The contactors are just there to get rid of this.)

• 12/29/2023, I had two spontaneous complete three to five second Load center drop-outs while in Utility-Solar-Bat mode, which I just would not ever expect to happen since the battery pack was 85+% SOC, and PV was nearly non-existent.
  • I had been running in USB for the past couple of very cloudy days... I was conserving my battery charge, since I live in Maine and want it for possible Utility outage.

I never had a drop out since my long term test (in SBU mode) even if the batteries are at 6% SOC where I start to recharge the batteries from the grid.

• 12/29/2023, after switching to Solar-Utility-Bat from USB, I commanded mode change to Solar-Bat-Utilty, and the inverters refused to change.
  • I tried to make the change by both front panel, and WatchPower iPhone app.
  • I was able to sniff the clear-text network traffic and see the WatchPower app request the change, and the inverter "Ack" the request.
  • I had to completely power down the system, battery, PV and Utility feed, to re-boot the inverters. Only then, did it switch back to SBU.
• 1/9/2024 - Load center Drop-outs, 3 or more today on SUB.
  • There where no high current demands when this happened.... < 1000 watts... just coasting

I had it... I've Load-Transfer-Switch'ed from Inverter System power feed to Utility power feed, and put the Inverters in stand-by.
I'm SO upset right now!

I was just starting to get to the point where I really use, rely-upon, and enjoy using the system.

But, now I feel like I've been ripped-off with these inverters, given the numerous product problem complaints in this forum and on youtube, and now that there is no new firmware development... that is tantamount to product EOL... now I basically have expensive bricks.

What can I do to resolve these problems, short of replacing the inverters with something all together different and chalking this up to ... I'm just a sucker... don't by budget friendly inverter(s) ?

I don't want an experimental system... I want something that just works. I have enough other things to worry about without adding this to my life-problems list.

Over the last 12 month while I build the system and reading about more and more problems with the EG4-6500EX, I've feared about the reliability, also. But after this current 13day test I now have the feeling that the system will be reliable (but I still have no idea how good the system will work when I finally make it to install my PV panels with Tigo PVRSS, MPPT etc. - I have only spare time for the project). My load over the test was between 400W (which is may basic house load) and about 25kW peak over the last 2 weeks).
I've ordered the components end of 11/2022 and the 6500's are built in 09/2022 with all bonding screws in place (delivered firmware version was 79.03).

Purchased - 12/2022​

Inverters, panels, batteries​

Initial installation of Inverters and batteries - 6/2023​

View attachment 187912View attachment 187913View attachment 187914

Installation of panels - 10/2023​

View attachment 187921View attachment 187922​

View attachment 187923View attachment 187924​

Load transition from original house load panel center to new inverter load panel center - 11/2023​

View attachment 187917View attachment 187918​

Added - 12/2023​

Eg4 Communications Hub​

Chargeverter​

• 2- 6500ex
  • Manufactured 10/2020, according to the stickers on the side of the units.
    • Don't these fall into the unapproved component swap by the assembly service ? that the 79.71 firmware fixes ?
  • 240v / split phase config (2P1 & 2P2)
  • 79.71 firmware on both units.
  • Unit 1 (2P1)
    • PV2 - 1 string of 7 panels - roof mount - Tigo TS4-A-O per panel
  • Unit 2 (2P2)
    • PV1 - 1 string of 6 panels - ground mount - Tigo TS4-A-O per panel
    • PV2 - 1 string of 7 panels - roof mount - Tigo TS4-A-O per panel
• 1- EG4 Communication Hub
  • 1.10 firmware
• 6 - EG4 LiFePower4 batteries
  • V3.32 firmware
• 36 - BlueSun Bi-Facial Mono Perc 440w-460w panels
  • 20 - deployed
  • 14 - planned for summer deployment
• 2 - Tigo TAPs
• 1 - Tigo CCA

I hope that some of this information is helpful to you

 

[Adam De Lay]

Thanks to @Adam De Lay about his hint with flickering lights because of specific loads (in his case a laptop power supply), I've searched the load which causes the flickering in my environment. I have bad flickering when an old Vegas style slot machine was running in parallel. The flickering does only occur on the phase where the slot machine is connected and not all LED's on this phase are flickering.

Glad you were able to at least identify what was “causing” the problem even though you shouldn’t have had to. I know how big of a pain it is to try and go through every circuit and connected device…

 

[EG4_Jarrett]

But, now I feel like I've been ripped-off with these inverters, given the numerous product problem complaints in this forum and on youtube, and now that there is no new firmware development... that is tantamount to product EOL... now I basically have expensive bricks.

What can I do to resolve these problems, short of replacing the inverters with something all together different and chalking this up to ... I'm just a sucker... don't by budget friendly inverter(s) ?

I don't want an experimental system... I want something that just works. I have enough other things to worry about without adding this to my life-problems list.

There are a couple of things that we can look at to determine the cause. You've already narrowed down a lot of it in your post too, so I appreciate all of the information that you have provided. We know that the units are a little bit older, so they may not have the bonding screw. The lack of continuity also being a telltale sign. The bonding screw helps with the output sine wave quite a bit. I'm not saying that is a guaranteed fix, but its a possible solution.

Where are you bonded at? Are your PV panels grounded to your AC ground/ground rod? Can you take a closer picture of the inverter wiring?

 

[fmeili1]

Seems like you already have the solution based on your comment "What can I do to resolve these problems, short of replacing the inverters with something all together different and chalking this up to ... I'm just a sucker... don't by budget friendly inverter(s) ?"

Right now, Victron Multiplus II inverters just had another price drop. You can get a Victron Multiplus II 48 / 3000 / 120 for about $1200 on Amazon or any other distributor. Since you're running split phase, get 2 inverters. You will need some solar charger controllers. My preference is the 250/100 ($600) or the 150/45 ($220-ish). Each has their own benefit and use case. It will take some "elbow grease" but by the looks of your pictures, you can easily swap them in.

They can be easily setup to automatically draw on the grid when overloaded e.g. if for some reason you pull 8000 watts, the inverters can be easily setup to run on battery with assistance from the grid for as long as you need.

If this is something you are interested in, please tell us and we can guide you through. There are many nuances like your distributor is your tech support so buy from a good distributor.

I went cheap at first but now I have a good working Victron system. It is pricier but when you consider your time, frustration, ease of use, etc, it actually comes out cheaper to go with better equipment.



Also, concurrently start a support ticket with Signature Solar. Wait times are in the 2+ week time frame and any meaningful assistance can take several weeks. You will need to take several pictures and jump through many hoops.

It's a question of budget.

I've invested 6x $1,300 for the AIO's (inverters/chargers) which are now older than 1 year. To just have the replacement for the inverter part it would cost 6x $1,700 (for the 5000W Victron to get as close as possible to the 6500W like the EG4 - but still get 10kW less possible continuous power output). On top of that I need 8 MPPT charger which needs to handle at least 350Voc - this would require 4 Victron 450/100 ($1,230) or 2 Victron 450/200 ($2,160). Everything together would cost a fortune. As nice it would be to run everything with Victron, it would cost too much - the system should have an ROI of <10 years and with the required power it's just not possible with Victron. The inverters should permanently driving the whole house off-grid - I only added the AC-in grid as a "last resort backup option" because I'm not allowed to be disconnected from the grid because of city code - but the system is designed to be able to run forever without grid!
I can easily live with some flickering LED's (but I'm close to a solution to solve that also) which is so far my only remaining problem. Beside this my EG4-6500EX based system works reliable for an unbeatable price.

 

[nkhorman]

There are a couple of things that we can look at to determine the cause. You've already narrowed down a lot of it in your post too, so I appreciate all of the information that you have provided. We know that the units are a little bit older, so they may not have the bonding screw. The lack of continuity also being a telltale sign. The bonding screw helps with the output sine wave quite a bit. I'm not saying that is a guaranteed fix, but its a possible solution.

I had thought about adding the bonding screw, but elected to not open the units and void the warranty.

Where are you bonded at?

The bonding appears to be in the exterior meter main load center breaker combo. I'm unable to visually verify this assertion since it requires breaking the utility company seal to open the box to examine the ground rod bonding.

The house sub-panel load center that I show in the picture, does not bond the neutral to ground, and is directly wired to that up-line outside meter main load center breaker combo.

I did ohm out / verify continuity between Ground and Neutral exists at that panel, as provided by the SE cable from the exterior of the house.

The inverter system and new load center transfer switch are fed by a 50 amp circuit pair from this original house sub-panel (L1,L2,N,G)

Are your PV panels grounded to your AC ground/ground rod?

The 14 roof panels on the other end of the house are frame grounded with a rod separately.
The 6 ground panels are not presently grounded.

Can you take a closer picture of the inverter wiring?

I'll follow-up with one and wiring details.

Like fmeili1's system above, I took care to keep all DC cable lengths the same, and I believe I did so with the AC cables as well.

 

[Dinobot248]

It's a question of budget.

I've invested 6x $1,300 for the AIO's (inverters/chargers) which are now older than 1 year. To just have the replacement for the inverter part it would cost 6x $1,700 (for the 5000W Victron to get as close as possible to the 6500W like the EG4 - but still get 10kW less possible continuous power output). On top of that I need 8 MPPT charger which needs to handle at least 350Voc - this would require 4 Victron 450/100 ($1,230) or 2 Victron 450/200 ($2,160). Everything together would cost a fortune. As nice it would be to run everything with Victron, it would cost too much - the system should have an ROI of <10 years and with the required power it's just not possible with Victron. The inverters should permanently driving the whole house off-grid - I only added the AC-in grid as a "last resort backup option" because I'm not allowed to be disconnected from the grid because of city code - but the system is designed to be able to run forever without grid!
I can easily live with some flickering LED's (but I'm close to a solution to solve that also) which is so far my only remaining problem. Beside this my EG4-6500EX based system works reliable for an unbeatable price.

I understand money does not grow on trees. My guess is that like many others in this forum, we/they tend to over build, myself included.

I threw ROI out the window for many reasons. Mainly difficult to compute intangibles (eg joy of diy project, hobby, independence), unfair to exclude my labor in ROI calculation, and many others I can’t think of right now.


Back to the tendency to overbuild. In my case, if I were trying to maximize ROI, I would have built a smaller system and leveraged the grid for any overage. The marginal cost of needing to cover a 4000w peak draw in the morning and at dinner time is better solved with simply using the grid. Normal house loads are 100 - 1000w. Daily usage is about 10 kWh. That peak load probably cost me say 0.25 per day or less than $10/month. No need to double the system size and cost for a measly $10 / month. Another way to say it, the smaller system (say 3kw sized) gets more utilization and sits less idle.

Obviously, I chose to overbuild and threw out ROI along with not having the knowledge I know now before I started (hindsight is great but tends to be late). If ROI was the main driver and I could do it all over again, I would build smaller and upgrade only when the financial numbers could prove it. Or in this case, replace some equipment as needed then replace more equipment when the budget allows.


Good luck. Every system will have compromises.

 

[Dinobot248]

I remember one more reason why I had to throw out ROI.

Wife and kids don’t understand or appreciate the literal pennies that I am trying to save each day with this system. I am collecting those pennies back over 10 or 20 years along with the joy of power independence and redundancy.

They don’t want to change their behavior and continue as normal. I can’t put an ROI on them.

 

[Adam De Lay]

Sorry @nkhorman missed that this was a new thread I was tagged in.

First off, great job on the install! Looks great! Sorry you're having all these issues though.

RGB led bar just glows barely visible low intensity red on both units. It did work at one point, and I disabled it, but now won't come back on when enabled.

So this has come up a few times. One fix was to break the split phase connection and then the RGB bar started working again. Here's the thread of the discussion.

Eg4 light bar showing faint red

Those LED strips have power and serial data sent to them and a tiny micro controller does the magic. Very easy to swap the LED bar if you want to.

diysolarforum.com

I have very prominent light flickering when running on Solar-Bat-Utility, wether PV charging or not, but not when running in Utility-Solar-Bat / Solar-Utility-Bat.

• The light flicker is Incandesent bulb, CFL bulb, LED bulb, and LED shop dual T4 form-factor lights... ie.. the type and quality doesn't seem make a difference.
• I don't think I have the bonding screw installed in either inverter, based on lack of continuity from neutral to ground when only powered by battery.... no AC In or AC Out wired connections, and the original as shipped firmware, pre-79.71 ( I've forgotten that version number ).

• I had decided to basically just ignore the flickering as it seemed that no loads stopped working or worked badly other that lights.

Yeah, when everything is powered by utility, you shouldn't see the flickering because it's just passing the power through the inverter. It might be worth looking at adding the bonding screws back in. I believe I saw later in the original message that all inverters are now updated to 79.71. There's been some reports that the bonding screw is used for more than just the G/N bond, something related to grounding the MPPT or something like that. I can't remember all the specifics.

12/29/2023, I had two spontaneous complete three to five second Load center drop-outs while in Utility-Solar-Bat mode, which I just would not ever expect to happen since the battery pack was 85+% SOC, and PV was nearly non-existent.

That's odd that you would have drop-outs when your in Utility mode.

12/29/2023, after switching to Solar-Utility-Bat from USB, I commanded mode change to Solar-Bat-Utilty, and the inverters refused to change.

• I tried to make the change by both front panel, and WatchPower iPhone app.
• I was able to sniff the clear-text network traffic and see the WatchPower app request the change, and the inverter "Ack" the request.
• I had to completely power down the system, battery, PV and Utility feed, to re-boot the inverters. Only then, did it switch back to SBU.

I experienced this several times during my testing. It seems like sometimes the inverter "remembers" that it's in a "utility" mode and won't switch out of it till the criteria of option 13 is hit. Regardless of how many times you signal for it to change, it just stays there. I've even had it happen after a reboot.

1/9/2024 - Load center Drop-outs, 3 or more today on SUB.

• There where no high current demands when this happened.... < 1000 watts... just coasting

Again, the drop-out when in utility mode is puzzling. I know after I updated to 79.71 I started experiencing overloads which would drop me from battery to grid when certain loads would turn on. It never happened before 79.71 with the exact same loads. I'm almost wondering if that's what's happening while you're on Utility mode, but I would think it would failover to battery.

2- 6500ex

• Manufactured 10/2020, according to the stickers on the side of the units.
  • Don't these fall into the unapproved component swap by the assembly service ? that the 79.71 firmware fixes ?

I wana say the component swap happened after the 10/22 batch. I had some from 10/22.
The 79.71 fix was to get rid of the need for having two different firmware versions based on if you were using the inverter in a stationary/residential install vs a mobile install. It also removed the need for opening up the unit and removing the bonding screw.

It added a new menu option (42) to control the ground/neutral relay inside the inverter.
- ENA (stationary environments) will disable the G/N bond relay inside the inverter and create a common neutral where the AC In and AC Out neutrals are connected together. In this instance you have to add your own G/N bond (whether that comes from your main grid panel through the AC IN connection or you add it to a sub-panel just after the inverters).

- DIS (mobile environments) will enable the G/N bond relay. If you have AC IN power and your inverter is in a "utility/grid" mode, the G/N bond relay will dynamically be disabled. When your inverter switches to battery mode, the G/N bond relay will be enabled, creating the bond inside the inverter.

Keep in mind you only want 1 G/N bond in an environment.

 

[fmeili1]

I understand money does not grow on trees. My guess is that like many others in this forum, we/they tend to over build, myself included.

I threw ROI out the window for many reasons. Mainly difficult to compute intangibles (eg joy of diy project, hobby, independence), unfair to exclude my labor in ROI calculation, and many others I can’t think of right now.


Back to the tendency to overbuild. In my case, if I were trying to maximize ROI, I would have built a smaller system and leveraged the grid for any overage. The marginal cost of needing to cover a 4000w peak draw in the morning and at dinner time is better solved with simply using the grid. Normal house loads are 100 - 1000w. Daily usage is about 10 kWh. That peak load probably cost me say 0.25 per day or less than $10/month. No need to double the system size and cost for a measly $10 / month. Another way to say it, the smaller system (say 3kw sized) gets more utilization and sits less idle.

Obviously, I chose to overbuild and threw out ROI along with not having the knowledge I know now before I started (hindsight is great but tends to be late). If ROI was the main driver and I could do it all over again, I would build smaller and upgrade only when the financial numbers could prove it. Or in this case, replace some equipment as needed then replace more equipment when the budget allows.


Good luck. Every system will have compromises.

Thanks!

Long story about my motivation of over sizing (or just read the summary at the end because it may be boring to read).

I would like to tell you something about my project and how it has changed over time. Large projects sometimes develop their own momentum and the goals can change over time.

When I started working with solar about 2 years ago, I had relatively little knowledge of the topic (I know a lot about electrical engineering, but not specifically in the solar field). So I started a smaller DIY project to equip our motorhome with a solar system. That worked very well and was a lot of fun.

After that, I started thinking about a solar system for our house. At the beginning I assumed that I wouldn't be able to implement a home solar system myself as a DIY due to time constraints. So I went to a solar provider in our city and asked about the prices for a solar system with energy storage. The storage should be of a size that makes me independent of the power grid for around 1 day. The solar supplier quoted me $120,000 for a system with 40kWh of storage - that blew my mind and I began to question how that could be.

The solar provider explained to me that he only uses grid-connected micro inverters and no longer offers string inverters due to the installation effort and durability. Therefore, only special hybrid batteries that work with the micro inverters are possible and they are incredibly expensive.

Since I still knew the prices for LFP batteries from my motorhome solar conversion, I started designing a system myself. I then became aware of these budget Voltronic based HF AIO clones (e.g. EG4-6500EX) through this forum and saw some positive Y**tube videos about them. In the course of my calculations, I realized that even a system with high power and a large battery is possible for less than 1/3 of the price (but only as DIY and not counting hours). It would even be enough to completely disconnect from the grid and run the house completely off-grid. We have several power outages here every year (some lasting several hours) and most of our energy comes from the Colorado River dams - which unfortunately are losing more and more water. That's why I came up with the idea of calculating a system without a grid,

I realize this only works because I'm fortunate to live in an area that has >320 full days of sunshine per year. Therefore, this idea cannot be implemented by everyone everywhere.

Our house is completely electrically powered (200A service). During the hottest summer months here in the desert, we require up to 114kWh of energy per 24h, mostly for air conditioning in the house (we average about $280 in electricity bills per month). At the beginning I planned with 4 AIO's because 26kW inverter power is usually enough for our house.

In the end, I realized that I could basically build a DIY system that could run completely independently. I then oversized the system even further (6 AIOs) so that there would be no loss of comfort when using high-load consumers in parallel (that was the point that my wife feared most ). 2 additional AIOs are no a big addon in relation to the total costs of the system and it also increases error redundancy.

At the end of the considerations, I had a precise plan of a system with which I would be completely self-sufficient and have no restrictions on comfort when using high power loads in parallel and with affordable component prices.

In my naivety, I then tried to find a solar company that would implement my plan with some parts DIY. No solar company was willing to implement the project according to my plans - of course not, because it was naive of me to try...

So I decided to do everything myself. Getting permits from the city and the energy supplier (even i'ts an off-grid system, the cities building department requires always a permit from the energy provider if solar is involved) was a bureaucratic gauntlet. But in the end I had the OK from both of them and the building permit in my hands.

Now, about a year later, the system is installed (except for the solar modules, which I hope to have installed on the roof in the next 6 weeks). So far and after successful preliminary system tests overs some weeks with batteries, my plan seems to be working.

In addition to the immunity against power outages (which can be life-threatening here in the desert if they last longer - the lakes which produces our energy getting lower levels from year to year...) and the prospect of never receiving an electricity bill again (other than the basic fee, because city code does not allow to be disconnected from the grid), I had a lot of fun with the project. I also make a small contribution to reducing CO2 (I trust the 99.8% of scientists and believe in man-made climate change). Furthermore, I didn't have to enter into a feed-in contract with poor conditions and long terms with the energy supplier. Since the system is no grid-interactive, the energy supplier did not dictate any components that they had specifically approved (beside a manual transfer switch because they treat the system as a generator solution), which would usually require significantly more expensive tier-1 equipment.

Summary:
You have to significantly oversize a system if you want to be completely independent of the grid WITHOUT sacrificing comfort.
Normally this doesn't make sense, but if you live in a place with a lot of sun, use inexpensive components, and do everything yourself, it's possible to achieve all of these goals at an affordable price.
Due to the use of budget components, additional research and finding ways to work around some problems must definitely be taken into account. You may also have to live with some minor problems (e.g. LED flickering).
It's like always in the life, you can't get everything for less - you need to do compromises!
That's the reason why I don't complain too much about the EG4-6500EX for it's price, it's a really good AIO.

The other option, using Tier 1 equipment (e.g. Victron - which I really like and used in my RV project) would have resulted in less effort for additional research and finding workarounds, etc., which would have been nice - but without achieving my goal, to be completely independent without comfort loss.

I chose the first path. Over time I'll see if it was a good decision...

 

[nkhorman]

Sorry @nkhorman missed that this was a new thread I was tagged in.

No worries.

First off, great job on the install! Looks great! Sorry you're having all these issues though.

Thanks !, I wish that only looks mattered.
My install was inspired by your second install with the gutter underneath.

So this has come up a few times.
One fix was to break the split phase connection and then the RGB bar started working again. Here's the thread of the discussion.

Eg4 light bar showing faint red

Those LED strips have power and serial data sent to them and a tiny micro controller does the magic. Very easy to swap the LED bar if you want to.

diysolarforum.com

Thanks for the link.
Yeah, I cruzed by that one once TLDR and because didn't apply at the time.

I looked at it again. It doesn't look like it actually gets solved.... just .... coerced back to function... maybe... sometimes.

Without the benefit of that thread, to my eyes, it looked like a hardware malfunction, not a software issue, as is implied by the behavior described.

And to be frank, if the issue doesn't actually affect operations, it's very low priority, but it still should work, and and assuming it's not actually an electronics hardware issue, it indicates that the firmware is not well maintained / coded.... too many unknowledgeable hands in the pot. (I speak from experience... not that I had a hand in thsi firmware, but that I've been coding since the 80's and understand how things like this get........ broken.)

Yeah, when everything is powered by utility, you shouldn't see the flickering because it's just passing the power through the inverter.

Yeah, that is mostly what I percive, no real prominent flickering, although there are small amount sometimes, and I'm not talking about when a large load become active... again... just coasting. I'll try and correlate that against PV charging.

It might be worth looking at adding the bonding screws back in. I believe I saw later in the original message that all inverters are now updated to 79.71. There's been some reports that the bonding screw is used for more than just the G/N bond, something related to grounding the MPPT or something like that. I can't remember all the specifics.

I recall seeing something here to that effect, and I considered opening the unit and adding the bonding screw on the assumption that it wasn't present because of my G/N continuity testing, but didn't want to void warranty.
I'm wating for SS or EG4 to tell me to do it.

That's odd that you would have drop-outs when your in Utility mode.

Yeah... that was my conclusion.
In any mode, USB, SUB, or SBU, I don't expect the inverter to ever drop load under normal operating conditions... ie. If it has sufficient power to draw from U or S or B, it shouldn't drop load, unless there is an over-current / over-temp or other type of self-protection fault. EVER !

But, as I've discovered the hard way, that is not the case. And... It never throws and error code that I can find.

All that said, I've been ruminating on the drop-load occurances, and and it triggered a recolection of one of the threads here that talked about; the inverters have this kind of problem when the battery pack nears full charge.

Presently, I do believe that the drop-load occurances all coincide with battery pack full charge state.

That was the case yesterday that finally motivated my post. It reached 99% / 100% SOC from PV.
And the 85% case from 12/29 was also when my pack reached 99% / 100% from PV, but the SOC wasn't callibrated (And I think I was running old battery firmware... not entirely sure about that.)

I experienced this several times during my testing. It seems like sometimes the inverter "remembers" that it's in a "utility" mode and won't switch out of it till the criteria of option 13 is hit. Regardless of how many times you signal for it to change, it just stays there. I've even had it happen after a reboot.

I'll keep option 13 in mind. Thanks.

Again, the drop-out when in utility mode is puzzling. I know after I updated to 79.71 I started experiencing overloads which would drop me from battery to grid when certain loads would turn on. It never happened before 79.71 with the exact same loads. I'm almost wondering if that's what's happening while you're on Utility mode, but I would think it would failover to battery.

Ok, that sucks. Well, like I mentioned eariler... was just coasting.. no real load changes.

I wana say the component swap happened after the 10/22 batch. I had some from 10/22.

It' be nice if SS or EG4 would provide unit identifiers with "the problem"

The 79.71 fix was to get rid of the need for having two different firmware versions based on if you were using the inverter in a stationary/residential install vs a mobile install. It also removed the need for opening up the unit and removing the bonding screw.

It added a new menu option (42) to control the ground/neutral relay inside the inverter.
- ENA (stationary environments) will disable the G/N bond relay inside the inverter and create a common neutral where the AC In and AC Out neutrals are connected together. In this instance you have to add your own G/N bond (whether that comes from your main grid panel through the AC IN connection or you add it to a sub-panel just after the inverters).

- DIS (mobile environments) will enable the G/N bond relay. If you have AC IN power and your inverter is in a "utility/grid" mode, the G/N bond relay will dynamically be disabled. When your inverter switches to battery mode, the G/N bond relay will be enabled, creating the bond inside the inverter.

Yeah, I remember seeing the video where you were trying to come to grips with the wording ENA / DIS vs. when the relay was active / inactive vs. when the bonding was active or not... and me too!
I glad you summarized it here. Thank you.
Between that and the "flickering lights fix", that is the reason a updated to 79.71, with no change in flicker behavior... I was disappointed.

Keep in mind you only want 1 G/N bond in an environment.

Yep... that is why i'm a little hesitant to start messing the the bonding screw and the 42 option. I don't want to add to my issues.

 

[nkhorman]

Something else to add to the mix...
Has anyone seen this ?



I think it showed up when I added the EG4 ComHub.
It's on my 2P1, where I have the battery coms connected.

 

[Adam De Lay]

Something else to add to the mix...
Has anyone seen this ?

View attachment 188104

I think it showed up when I added the EG4 ComHub.
It's on my 2P1, where I have the battery coms connected.

I've never seen it, but I've also never had any EG4 batteries before either.

 

[Adam De Lay]

I recall seeing something here to that effect, and I considered opening the unit and adding the bonding screw on the assumption that it wasn't present because of my G/N continuity testing, but didn't want to void warranty.
I'm wating for SS or EG4 to tell me to do it.

I'm not sure if @EG4_Jarrett can document on your account to give you permission to add the screws back or not.

Because I'm always forgetting the context, I went through and found the conversation where Markus from EG4 stated some possible fixes to adding the bonding screw back in. I added that information to my 6500 changelog.

EG4 6500 Firmware Changelog - Updates

diysolarforum.com

Yep... that is why i'm a little hesitant to start messing the the bonding screw and the 42 option. I don't want to add to my issues.

If you already have a bond in your system, which it sounds like you do, then you would want to set 42 to ENA.

 

[nkhorman]

I've switched them off because they make a very annoying high frequency noise which I definitively don't want to have. Never tried to turn it on again. I will check and will tell you if it works. But I think this is the least important problem...

Thanks to @Adam De Lay about his hint with flickering lights because of specific loads (in his case a laptop power supply), I've searched the load which causes the flickering in my environment. I have bad flickering when an old Vegas style slot machine was running in parallel. The flickering does only occur on the phase where the slot machine is connected and not all LED's on this phase are flickering.

To find out the load which causes the problem, I've switched off breaker by breaker in the house sub-panel while watching the flickering until the flickering stops. After that I knew the room where the problematic load is. After that I've disconnected/switched-off device by device in this room until the flickering stopped.

Later I've identified an additional load which is my 3D printer which also causes the LED flickering (because it was not running so often compared to the slot machine which is usually always running because the are designed to run forever).

My flickering does nearly disappear (with running the problematic load in parallel) when the inverters have higher load >15%.

Without running these problematic loads, I have no more flickering/pulsing LED's - but this is not a solution. I've tried different type of filters placed before the problem causing loads, but without success. Now I think about two options to solve this problem

1. modding the flickering LED's to improve their power supply to make them more resistant to voltage fluctuations (a lot of work, because all my flickering LED's are not LED bulbs with the possibility to change them, they are integrated in ceiling fans).
2. put small DIY mini UPS (online, double conversion) in front of the problem causing loads (higher costs).

So far I don't have the PV panels installed and all my tests are done by battery only, recharged from grid if they getting empty. For long term testing (13 days now in a row), of the system I'm running the whole house (200Amp service) on just the inverters and have no issues so far. The system goes only in bypass mode depending on settings 12 and 13 for "point back to utility" and "point back to battery" to recharge the batteries.

Here are some details which may help you to find differences with your setup:

• six EG4-6500EX in parallel split phase and 60kWh LFP batteries
  • firmware 79.71
  • all bonding screws are in place on all six inverters
  • option 42 set to "ENA"
  • common neutral wiring
  • only using SBU mode
  • AC-in "GND" connected to common ground in the AC-in breaker panel, AC-out "GND" not connected (the are bridged inside the AIO's and I want to prevent ground loops)
  • All battery cables have the same length
  • All AC-in and AC-out wires have the same length to the AC-in breaker panel and AC-out breaker panel
• installed as a "solar-generator" with a manual transfer switch (manual, 2-pole, open transition type) to drive the whole house by grid or by "solar-generator"
• common neutral wiring
• Independent of the transfer switch position, the AC-in is permanently connected to the grid as a backup scenario in case there will be not enough solar and battery power (but I disconnect the AC-in from the grid via contactors which are only activated just before the AIO program setting 12 will get active to reduce grid energy consumption - if grid is just connected with AC-in, each AIO consumes about 50W, which is about half of the idle consumption, from the grid even if grid is not used! The contactors are just there to get rid of this.)

I never had a drop out since my long term test (in SBU mode) even if the batteries are at 6% SOC where I start to recharge the batteries from the grid.

Over the last 12 month while I build the system and reading about more and more problems with the EG4-6500EX, I've feared about the reliability, also. But after this current 13day test I now have the feeling that the system will be reliable (but I still have no idea how good the system will work when I finally make it to install my PV panels with Tigo PVRSS, MPPT etc. - I have only spare time for the project). My load over the test was between 400W (which is may basic house load) and about 25kW peak over the last 2 weeks).
I've ordered the components end of 11/2022 and the 6500's are built in 09/2022 with all bonding screws in place (delivered firmware version was 79.03).

I hope that some of this information is helpful to you

All this sounds very similar to how and why I built my system.
I want to have assurance of short-term, 24hr'ish power availability primarly without the use of a generator, but also to try and offload the cost of monthly utility, which is just silly expensive. /rant - 50% of my bill is fee this, fee that, delivery fee... avoid the PUC price regulation etc. and the other 50% is the actual electricity cost/

I started with mounting and wiring the inverters, batteries, and new load center.
I then transfered a couple of low priority circuits from the existing load panel to the new one, and did some limited system behavior testing.
That is when I encountered light flickering, without having any PV installed.
I chased the problem down to one device.... An 27" 2011 iMac when powered on.
I have to say... I did not expect a computer power supply to affect the inverters at all !

At this point, I paused moving the rest of the house to the new load panel, primarily due to the flicker problem, but also because I had no PV.

I later ignored the flicker problem, and pushed forward because of power outages and the up-comming winter season.
This was followed later by the 14 panel 2@7s PV installation, where I connected a 7s string to each inverter.
I followed with some house / system usage patterning to find out that those panels were not going to cut it.

They are not at optimal sun facing southern exposure, and are partially occluded by trees depending on the time of day sun angle.
At this time of year, there generally is just no way that I'll recharge the battery using just that PV set, so against my preference, I scrambled and put up a "temp" ground mount and installed a PV 4s string.
That 4s string out performed the other roof mout set by it's self.
The nature of the temp ground mount had enough space for two more panels, so I did that, and rewired it from 4s to 6s.
This 6s panel set ALWAYS out-performs the roof mount set.

The lesson I learned as a result, is that occlusion and angle is VERY important to performance, which I knew, but just didn't expect it to be this dramatically different.

Now with this configuration I'm able to capture / generate 11kwh to 16kwh on good days, which will recharge my battery from 50% to ~90% depending on the weather / clouds / length of sun-day / time of year, which presently is 9am to 2 or 3pm'ish.

 

[nkhorman]

I'm not sure if @EG4_Jarrett can document on your account to give you permission to add the screws back or not.

Because I'm always forgetting the context, I went through and found the conversation where Markus from EG4 stated some possible fixes to adding the bonding screw back in. I added that information to my 6500 changelog.

EG4 6500 Firmware Changelog - Updates

diysolarforum.com

If you already have a bond in your system, which it sounds like you do, then you would want to set 42 to ENA.

Which inverter ? 2P1, 2P2, or both ?
.... and is there sequence, disconect load, make change, reboot ? or something ?

 

[EG4_Jarrett]

Something else to add to the mix...
Has anyone seen this ?

View attachment 188104

I think it showed up when I added the EG4 ComHub.
It's on my 2P1, where I have the battery coms connected.

Can I get an EG4 technician to give you a call? If so, then please feel free to send me your info and a time period that you would be available to talk via direct message. I think that it would be better to speak over the phone and have someone that has worked on these as much as I have (if not more) take a look at the system and see if we can get everything up and running for you.

 

[Adam De Lay]

Which inverter ? 2P1, 2P2, or both ?
.... and is there sequence, disconect load, make change, reboot ? or something ?

Option 42 can be set on either inverter and the value mirrors to the other one. All you have to do is make the change.

As far as the bonding screw is involved, knowing it affects more than just the relay, I would add back to both. Obviously you would want to kill all power from AC, PV and DC before you open the case.