Deye / Solark 8K Setup, Some photos an diagrams!

[mrdavvv]

So this beast finally arrived:





(8K LP1 US model) -- At a first glance i really like the quality of materials and the weight of it!.

• Planning on putting 10.5Kw of panels (26 x 405W), in a east - west configuration. Panels would be around 20 - 30° to E / W.
• Each MPPT would have 2 strings in parallel, with 6 or 7 panels each.

Also bought a little pylontech battery (2.2Kwh@48V), it should be able to communicate with the inverter trough CAN / RS485, this is against the DIY spirit .... but didn't have the 2 months to wait for the lifepo4 cells to arrive:



This would be the diagram of the full installation

https://imgur.com/KFFNwDe

Full Image Link

• What do you think?, any mistakes in the drawings?, any recommendations?.

Ill keep posting here with updates during installation.

 

[FilterGuy]

Also bought a little pylontech battery (2.2Kwh@48V), it should be able to communicate with the inverter trough CAN / RS485,

Is the battery advertised as being able to communicate to the Growatt? Just because they both have Can and/or RS485 does not mean they use the same protocol. Having said that, if the Pylontech has an independently operating BMS, you don't really need communications to the Growatt.

Also, have you figured out if the Gro-watt does anything around a Neutral/Ground bond? I sent Growatt this message:

I am assisting a customer that is using the Growatt SPF 3000TL LVM-24P inverter and need information about whether the inverter does internal switching of a ground-to-neutral bond?

Just this morning I received this reply:

The output neutral and ground are not bonded together, thanks!

Best Regards!
Darren Song (chn:宋伟) | America Customer Service Manager
Growatt New Energy Technology Co.,Ltd

It surprises me that they don't manage it internally so I am going to try to have the owner do some tests to figure out exactly what it does.
In the manual it suggests that an external relay needs to be set up but did not explain it well. (It would be unfortunate if an external relay must be set up)

BTW: If you want to learn more about the purpose for the Neutral-ground bond and how inverters are involved, read these two resources:

Grounding Made simpler - Part 1: AC/houshold grounding

To get the paper, click on the orange button at the top of the screen. The subject of grounding is a complex, multifaceted subject, that is often treated as an after-thought but needs to be considered from the beginning of the design and build...

diysolarforum.com

Grounding Made simpler - Part 2: Stationary Systems

To get the paper, click on the orange button at the top of the screen. The subject of grounding is a complex, multifaceted subject, that is often treated as an after-thought but needs to be considered from the beginning of the design and build...

diysolarforum.com

 

[mrdavvv]

Hello FilterGuy.

Yeah the Pylontech its listed in the compatible models of Deye's documentation, it should work.

Btw maybe you have me mixed with someone else, but this project contemplates the 8K Deye (Simillar to Sol-Ark), not sure why you mention the Growatt. (I use MPP-Solars in my other setup and talked a bit about growatts in other peoples threads, maybe thats why).

But im glad you attached that documentation, im still not clear in the grounding scheme, the Deye /solark only ask a neutral / ground bond in some part of the installation, doesn't mention anything about the inverter switching this. Will check. Thanks!

 

[FilterGuy]

Btw maybe you have me mixed with someone else, but this project contemplates the 8K Deye (Simillar to Sol-Ark), not sure why you mention the Growatt. (I use MPP-Solars in my other setup and talked a bit about growatts in other peoples threads, maybe thats why).

Oh C@*&%! Your correct.... I was mixing conversations...... Sorry for the confusion.

 

[FilterGuy]

I downloaded the 8K Deye manual and saw this:


I hope they are only talking about grounding the actual current carrying conductors. You should *always* ground the solar panel frame & mounts.


I also found this:


It looks like they want you to install an external relay to manage the Earth-Neutral bond. (Signal Island Mode)

I also found this:

 

[mrdavvv]

Hello @FilterGuy

Took a couple of days to read all your grounding documentation. Now i'm glad you mix me up!.... as you know there is a lot of contradictory information about PV grounding in the internet... and after a lot of reading and even a couple of hours of mike holtz videos, i was still not sure about the subject..... Your writings are very clear, concise and resolved almost all my doubts!. Thanks for that!

1.- Updated the diagram with a grounding scheme, i think it mostly good now (attached).

2.-

I hope they are only talking about grounding the actual current carrying conductors. You should *always* ground the solar panel frame & mounts.

Yep not a good explanation from the manufacturer, but im also sure they mean to not ground the current carrying ones.

3.-

BTW: If you want to learn more about the purpose for the Neutral-ground bond and how inverters are involved, read these two resources:

So... this is were im still a little lost.

First of all, as usual the chinese inverter's documentation its not that good, so i prefer the SolArk manuals, as far as i know they don't mention anything about active management of the system bonding jumper (GND to N) from the inverter, or any rele's or any special configurations (I will re-read just in case).

In their main diagram, they show a proper bonding between GND & N in the main panel, wich is the same as my diagram:




So do i need to do something else?, maybe the testing that you mention for the Growatt to be sure??.

• In my system 90% of the loads are "Pre-inverter". The inverter "injects" energy to this circuit trough the "GRID" input / output.
• Only 10% would be critical loads, that will be conected from the "LOAD OUT" output to an special circuit (Critical circuit), that its "separated" from the main panel, and only connected to the secondary panel "Critical loads panel". The only relation from this critical load system to the main panel, its the ground, that its conected directly to said main panel. If needed, i can reroute and connect the grounding of this circuit to the critical loads panel... but i think its the same?.
• The roof of the building its steel, that its connected directly to the metal structure, that goes to ground. Not sure if this counts as a ground loop, since from the structure of the PV panels we get two paths to ground:

1. Path trough metallic structure
2. Path trough GND cable that goes directly to the secundary and main panel.

• Main panel its grounded with a couple of copper rods and also a cable attached to the base of one of the columns (Steel beam from metallic structure)

------------

PD: Reading your manual i realize my other MPP Solar installation its wrong, the installers put an additional grounding rod for the PV panels, wich isn't connected to the main building GND. I had my doubts about that but they are the professionals!. I will need to fix that.

 

[FilterGuy]

Reading your manual i realize my other MPP Solar installation its wrong, the installers put an additional grounding rod for the PV panels, wich isn't connected to the main building GND. I had my doubts about that but they are the professionals!. I will need to fix that.

I am glad my documents helped you understand. My objective in the documents is to help people understand the physics behind things so they can better deal with the real world situations. Sadly, the NEC allows that extra ground rod (But does not require it). There are a lot of sources both in the solar industry and on the internet that put out bad information about grounding. Last I heard Mike Holt was still trying to get the NEC to publish an emergency warning about this but he is fighting the momentom of a large population of people that don't fully understand.

Main panel its grounded with a couple of copper rods and also a cable attached to the base of one of the columns (Steel beam from metallic structure)

That sounds like a good connection to earth. If possible, that should be the only connection to earth.

• The roof of the building its steel, that its connected directly to the metal structure, that goes to ground. Not sure if this counts as a ground loop, since from the structure of the PV panels we get two paths to ground:

1. Path trough metallic structure
2. Path trough GND cable that goes directly to the secondary and main panel.

Yes..... that is a ground loop. Unfortunately in situations like yours it can be difficult to meet code and avoid ground loops.
The NEC is difficult to sort out in situations like this and I am certainly not a NEC expert, but I don't see a provision that allows the metal building roof to be the ground path. Even if there is a clause someplace that allows it, I could easily believe inspectors would not.
The best you can do is to make sure the ground cable connects to the rest of the grounding system as close to the earth connection as possible.

So do i need to do something else?, maybe the testing that you mention for the Growatt to be sure??.

I really don't know.? Without a better understanding of what the inverter does, it is difficult to make a definitive statement.

Warning: PURE SPECULATION FOLLOWS:
Based on the documentation I am guessing that their design center is a stationary set-up where the AC-in always has a N-G bond and that bond is carried forward to the output. If that is the case then nothing more is needed for a stationary system.

However, if it is a mobile set-up, then a lot of times you won't be plugged into shore power and therefor there is no N-G bond on the AC-in. In that case, the relay mentioned in the documentation is needed to re-establish the bond. (Since mobile set-ups are outside their target market, they did not spend the extra couple bucks to build the relay in.

It is criminal how poor some of the inverter documentation is around the critical function.

 

[FilterGuy]

First of all, as usual the chinese inverter's documentation its not that good, so i prefer the SolArk manuals,

Be careful. 'Very Similar' may not be the same as 'identical'. A lot of these inverters are made by the same company based on the same overall design, but each company that contracts with them will specify a slightly different feature set. When the inverter gets built, the components for the features that the buyer doesn't want don't get installed.

 

[mrdavvv]

I am glad my documents helped you understand. My objective in the documents is to help people understand the physics behind things so they can better deal with the real world situations. Sadly, the NEC allows that extra ground rod (But does not require it). There are a lot of sources both in the solar industry and on the internet that put out bad information about grounding. Last I heard Mike Holt was still trying to get the NEC to publish an emergency warning about this but he is fighting the momentom of a large population of people that don't fully understand.

It really helped!, some documentation like NEC goes to technical and abstract, very hard to go around... your manuals are are distillation of a bunch of reading, and even with pretty diagrams!, much easier to read and get the basics behind our systems.

Seems like Mike Holt has spend his whole life fighting against the common knowledge and NEC ..

That sounds like a good connection to earth. If possible, that should be the only connection to earth.

Got it, new installers also wanted to put a new independent ground rod, with all this i can fight them about it.

Yes..... that is a ground loop. Unfortunately in situations like yours it can be difficult to meet code and avoid ground loops.
The NEC is difficult to sort out in situations like this and I am certainly not a NEC expert, but I don't see a provision that allows the metal building roof to be the ground path. Even if there is a clause someplace that allows it, I could easily believe inspectors would not.
The best you can do is to make sure the ground cable connects to the rest of the grounding system as close to the earth connection as possible.

Ok, suppose you don't care too much about what NEC wants, and you give more weight about what its technically right. What could be the best setup scenario?, don't attach a ground to PV at all?, just measure with a "groundometer" to see if you have a good connection between PV aluminum structure and your ground end? (Should be, as there is like 5000 screws connecting the roof to the beams, and all structure connections are soldered).

Failed to mention that im outside US, but our code its a copy of NEC and we use the same bi-phase 220V setup. Our inspectors dont even go inside the building, so in theory i can do whatever i want from the door. (At leas in an small installation like this, a bigger system would receive a good check).

Warning: PURE SPECULATION FOLLOWS:
Based on the documentation I am guessing that their design center is a stationary set-up where the AC-in always has a N-G bond and that bond is carried forward to the output. If that is the case then nothing more is needed for a stationary system.

However, if it is a mobile set-up, then a lot of times you won't be plugged into shore power and therefor there is no N-G bond on the AC-in. In that case, the relay mentioned in the documentation is needed to re-establish the bond. (Since mobile set-ups are outside their target market, they did not spend the extra couple bucks to build the relay in.

It is criminal how poor some of the inverter documentation is around the critical function.

With the lack of documentation i could only be sure if i do some testing (Not sure how difficult it would be) or asking other owners, unfortunately Deyes are not common in USA. In my experience the manufacturer / seller doesn't help too much with this kind of questions (As you experience with Growatt). indeed very poor documentation, but that's the cost you ptake for an inverter 1/3 of the price.

So im left with testing and checking better documentation, i can see that sol-ark has some differences, but i expect that most of the fundamentals are carried over. I take them only as a reference, as even between versions of Sol-ark there is some differences in the documentation, like the last model, that added a bigger passtrough (60A) and NEMA enclosure.

I think the differences between DEYE and Solark would be mainly the firmware and bottom connections... i plan on uploading some photos of the internals so we can start to get a better understanding of it .

-------------------------------------------------------

Againt thanks for all the info!

 

[Hedges]

I generally wouldn't worry about "ground loop".

"Ground" can be used to mean the current return path of a circuit. If it is the reference for TTL level logic and also carries amps from equipment, might cause shift in voltage so "1" gets confused for "0".

If conductors form a grid with small enough loops, it is a mesh ground, not a ground loop. "Small enough" means a fraction of a wavelength, 1/20th of wavelength or less. At 60 Hz, wavelength is 5000 km. 1/20th wavelength is 250 km. If a inverter has edge rates equivalent to 60 kHz, then 250m.

So I think multiple paths from frame of panels back to equipment chassis and ground rods is just fine.

I have read instructions for a Snap-On inverter welder that said racks of metal stock should be grounded, something about passive resonators that would increase EMI. If 10' or 20' long, could be be 10's of MHz, harmonics that can come from some switching power supplies. But in PV inverters I expect relatively good attenuation of harmonics.

 

[FilterGuy]

Ok, suppose you don't care too much about what NEC wants, and you give more weight about what its technically right. What could be the best setup scenario?, don't attach a ground to PV at all?, just measure with a "groundometer" to see if you have a good connection between PV aluminum structure and your ground end? (Should be, as there is like 5000 screws connecting the roof to the beams, and all structure connections are soldered).

I don't generally advocate ignoring the code, but since you asked..... From a pure engineering point of view, I would ensure a solid connection the the metal building structure and call it good. However, just because the metal is all bolted together does not mean it is a solid connection. Paint and other metal coatings can be an effective insulator and oxidation can reduce the connection over time.

I assume the building frame itself is well grounded. You mention it is welded and you mention it is tied with the ground rods. I would not be at all surprised if it is also bonded with the metal rebar inside the concrete where it is bolted to the foundation (or it goes into the concrete foundation).

I assume you have read my grounding paper on solar panels (part 3). If there are WEEB washers between the mounting frame components and the panels as well as between the mount frame and roof, you probably have a good connection to the metal roof. Without knowing anything about the roof construction, I really can't say if it is a good connection to the metal building frame. (I assume the '5000 screws' are sheet sheet metal screws.... and yes, that sounds like a good connection. The only concern is if the screws are exposed to the elements, they could rust and loose connection over time)

The problem of ensuring a good connection is why the NEC often does not allow for things like this. Rather than try to specify the thousands of parameters that would ensure a good connection in all the different metal roof situations, they specify a cable and how it gets installed..... that way they know it is a good connection to earth.

I think the differences between DEYE and Solark would be mainly the firmware and bottom connections... i plan on uploading some photos of the internals so we can start to get a better understanding of it

Make sure you don't tear any stickers that say it would void your warranty if torn or removed.

If you open it up, the most important thing is to trace what happens with the neutral. In particular:

1. is the neutral for AC-in directly connected to the Neutral for AC-out. (I suspect it is.... but that gets back to my speculation from the earlier post)
2. Does the neutral go to a relay that would connect to ground. (This may be hard to trace even if the inverter is opened up).

 

[FilterGuy]

If conductors form a grid with small enough loops, it is a mesh ground, not a ground loop. "Small enough" means a fraction of a wavelength, 1/20th of wavelength or less. At 60 Hz, wavelength is 5000 km. 1/20th wavelength is 250 km. If a inverter has edge rates equivalent to 60 kHz, then 250m.

It is not the 60hz that would concern me. It is the switching frequencies of the MPPT and the Inverter circuits (and their harmonics). In this particular case, the switching noise generated by the MPPT on the PV cable would be the most likely source of a resonance on the ground loop. The problem is that it is very hard to figure out before hand if there is going to be any resonance between a loop and the frequencies used in the system. Furthermore, if the PV sytem is generating a lot of RFI, (and a lot of systems do) it can be a real PITA figuring out why. Therefore it is good to do what you can up-front to prevent a problem.

A lot of people blame the inverter for the noise..... and in some ways they are correct because it is the originating source of the frequency. However, how the system is wired can be just as big of a factor. That is why a good Ham Radio operator can often quite down an existing system without changing the primary components. Things like twisting the DC cables and eliminating ground loops can do wonders.

The other potential problem is a magnetic pulse from a nearby lightning strike. Think of the loop as the coil of a generator. A big emp from the lightning strike will generate a pulse current on the loop. That can possibly arc to other circuits, but more likely cause a potentially damaging pulse on other nearby circuits due to inductive coupling. (It could also impose a voltage on other circuits due to capacitive coupling but that is probably less likely). This too is hard to predict. If all the wiring is inside a metal clad building, the building might be enough of a Faraday cage to suppress the pulse from the lightning.....or maybe not.

 

[mrdavvv]

I don't generally advocate ignoring the code, but since you asked..... From a pure engineering point of view, I would ensure a solid connection the the metal building structure and call it good. However, just because the metal is all bolted together does not mean it is a solid connection. Paint and other metal coatings can be an effective insulator and oxidation can reduce the connection over time.

I assume the building frame itself is well grounded. You mention it is welded and you mention it is tied with the ground rods. I would not be at all surprised if it is also bonded with the metal rebar inside the concrete where it is bolted to the foundation (or it goes into the concrete foundation).

I assume you have read my grounding paper on solar panels (part 3). If there are WEEB washers between the mounting frame components and the panels as well as between the mount frame and roof, you probably have a good connection to the metal roof. Without knowing anything about the roof construction, I really can't say if it is a good connection to the metal building frame. (I assume the '5000 screws' are sheet sheet metal screws.... and yes, that sounds like a good connection. The only concern is if the screws are exposed to the elements, they could rust and loose connection over time)

The problem of ensuring a good connection is why the NEC often does not allow for things like this. Rather than try to specify the thousands of parameters that would ensure a good connection in all the different metal roof situations, they specify a cable and how it gets installed..... that way they know it is a good connection to earth.

All clear. I was thinking about measuring ground at a couple of points with an earth meter (Need to research about this), to be sure if our connection from PV to final ground its good.. as you mention there is a bunch of variables that can affect the conductivity. Or if a ground loop such as this it not something very worrying, we can put the additional ground wire as NEC requires.

Make sure you don't tear any stickers that say it would void your warranty if torn or removed.

If you open it up, the most important thing is to trace what happens with the neutral. In particular:

1. is the neutral for AC-in directly connected to the Neutral for AC-out. (I suspect it is.... but that gets back to my speculation from the earlier post)
2. Does the neutral go to a relay that would connect to ground. (This may be hard to trace even if the inverter is opened up).

1.- Yes checked with a multimeter and both LOAD anc AC lines neutrals are the same.
2.- There is a bunch of rele's in the AC stage... so very hard to trace, would need to disassembly the PCB.

I generally wouldn't worry about "ground loop".

"Ground" can be used to mean the current return path of a circuit. If it is the reference for TTL level logic and also carries amps from equipment, might cause shift in voltage so "1" gets confused for "0".

If conductors form a grid with small enough loops, it is a mesh ground, not a ground loop. "Small enough" means a fraction of a wavelength, 1/20th of wavelength or less. At 60 Hz, wavelength is 5000 km. 1/20th wavelength is 250 km. If a inverter has edge rates equivalent to 60 kHz, then 250m.

So I think multiple paths from frame of panels back to equipment chassis and ground rods is just fine.

I have read instructions for a Snap-On inverter welder that said racks of metal stock should be grounded, something about passive resonators that would increase EMI. If 10' or 20' long, could be be 10's of MHz, harmonics that can come from some switching power supplies. But in PV inverters I expect relatively good attenuation of harmonics.

Dear god comments like this make me realize how little i know about signals and electronics

 

[mrdavvv]

There are a lot of sources both in the solar industry and on the internet that put out bad information about grounding. Last I heard Mike Holt was still trying to get the NEC to publish an emergency warning about this but he is fighting the momentom of a large population of people that don't fully understand.

So today i talked with the installers and they are adamant about placing their new independent ground... supposedly for lighting protection. They say our current ground path its too long and has too many 90° bends, they want a more direct path to ground for the PV.

Our current path its 2mt down from the roof to the secondary panel. From there, back up to the "attic" about 1mt..then 12mt horizontal, then 90° angle and down around 3mt to the main panel and grounding bar. From the grounding bar, around 1mt and 2 horizontal to our originally planned grounding rods...

I understand one of the reasons of not having two independent ground rods, is that for atmospheric discharges can cause a big difference in potential between rod#1 and a nearby hypothetical rod#2... that difference of voltage instead of flowing trough the ground, it would flow trough our electric system as it has a lower impedance... and you don't want that.

And i realize that if you get a direct lighting hit you are probably going to have a lot of damage aniways.... but i can see why they want a more direct path to ground... from the point of view of a direct hit.

Im not sure what would happen with a metallic buidling against said direct lighting hit... probably the energy will come down trough the structure itself, as it would have the best conductivity. So maybe the installer's idea of putting a direct copper cable to ground its not going to work anyways, at least with a metallic building (Maybe ill do so with a wood / concrete home). Im guessing your only chance against lighting hits its to put a propper lighting rod to guide all the energy to the ground.... but in the middle of the city i doubt we need something like that!

But after all this now im not so sure about our originall grounding diagram.

 

[FilterGuy]

So today i talked with the installers and they are adamant about placing their new independent ground... supposedly for lighting protection. They say our current ground path its too long and has too many 90° bends, they want a more direct path to ground for the PV.



I understand one of the reasons of not having two independent ground rods, is that for atmospheric discharges can cause a big difference in potential between rod#1 and a nearby hypothetical rod#2... that difference of voltage instead of flowing trough the ground, it would flow trough our electric system as it has a lower impedance... and you don't want that.

And i realize that if you get a direct lighting hit you are probably going to have a lot of damage aniways.... but i can see why they want a more direct path to ground... from the point of view of a direct hit.

Im not sure what would happen with a metallic buidling against said direct lighting hit... probably the energy will come down trough the structure itself, as it would have the best conductivity. So maybe the installer's idea of putting a direct copper cable to ground its not going to work anyways, at least with a metallic building (Maybe ill do so with a wood / concrete home). Im guessing your only chance against lighting hits its to put a propper lighting rod to guide all the energy to the ground.... but in the middle of the city i doubt we need something like that!

But after all this now im not so sure about our originall grounding diagram.

I am not sure what to say. But here are a couple of observations:
1) An auxiliary ground rod is *not* required by NEC. Furthermore, the NEC places no requirements on auxiliary ground rods. You could put a copper nail into the ground and wire it with 22awg wire and it would be sufficient to meet the requirements for an auxiliary ground rod. If an auxillary ground rod was important, wouldn't the NEC put some requirements around it?
3) If the original grounding scheme is so poor..... how did it meet the requirements?

In the end, it is your property. The only one with any authority is the inspector from the city..... not the installers.

 

[Hedges]

So today i talked with the installers and they are adamant about placing their new independent ground... supposedly for lighting protection. They say our current ground path its too long and has too many 90° bends, they want a more direct path to ground for the PV.

They're correct about that. If the path for lightning to reach ground encounters a sharp bend, that will be higher impedance and lightning will probably jump off the bend to the next piece of metal.

Probably this should apply to a lightning rod with straight wire to a ground rod, so lightning doesn't enter the PV frame or ground wires.
With that protection, I could see where NOT having a ground rod tied into system ground would be good, so charge dumped by lightning into the earth doesn't spread through your ground wire. But my ground-mount PV arrays have frames extending to the earth so can't help it.

 

[mrdavvv]

They're correct about that. If the path for lightning to reach ground encounters a sharp bend, that will be higher impedance and lightning will probably jump off the bend to the next piece of metal.

Disclaimer... some speculation ahead:

I think the problem is, the installers are considering using the ground cable for both grounding and lighting protection... and both functions can maybe not work together.

For example, if we are talking about grounding, ideally the grounding conductor cable connects to our main panel (One path to ground). But that requires to run it inside the building and do some turns. And we don't want a lighting follow that cable!

Now, if they are talking about lighting protection, you probably need lighting rod, a more thicker cable, outside the building and straight to the ground. But... if you connect bot systems together by the PV frames, now we have the problem of having 2 separate grounding rods!, that can produce other issues, for example a nearby lighting strike producing a differential between the two separate earth ground points:



In our situation, with a metal roof that connects directly to the ground by an structure with 8" steel beams, im not sure if any lighting system would help. As far as i read, simple anti - lighting systems are not that reliable, as you are working with that much unpredictable amount of energy..... i think we should only focus about a good grounding system (For static and near atmospheric discharges) and forget at the moment about solving direct lighting strikes, as that would require probably a whole separate system!

I am not sure what to say. But here are a couple of observations:
1) An auxiliary ground rod is *not* required by NEC. Furthermore, the NEC places no requirements on auxiliary ground rods. You could put a copper nail into the ground and wire it with 22awg wire and it would be sufficient to meet the requirements for an auxiliary ground rod. If an auxillary ground rod was important, wouldn't the NEC put some requirements around it?
3) If the original grounding scheme is so poor..... how did it meet the requirements?

In the end, it is your property. The only one with any authority is the inspector from the city..... not the installers.

Correct!, and i want a safe and proper installed system for this place, that's why i end up doing some research by myself and asking around. Alot of enginners around here just do things exactly as code requires (Wich aparently sometimes its not the safest / better solution), they are misinformed or just not updated, so i need to double check all the work and check other sources.

Installers will do as we told them, but need to be sure that what we told them its correct, and better if we can explain them why.

Last I heard Mike Holt was still trying to get the NEC to publish an emergency warning about this but he is fighting the momentom of a large population of people that don't fully understand.

Do you have by any chance the info about this?. I would like to talk about this to the installers. In case we need to justify some decisions:

----------

Reference image:

 

[mrdavvv]

Wanted to see the insides of the cooling module:



Looks pretty nice!, the fans are standard and easily replaceable, huge heat sink and everything its completely sealed and separated from the internal electronics.

Just hoping its not too loud!

 

[newbostonconst]

Here are my observations reading through this and seeing some confusion.

1. Grid into your house is normally 2 legs at 120vac each that can be used to get 240vac and a neutral (not a ground). The "neutral" is the reference that keeps the 120vac lines stable when used independently and not as 240vac.
2. The neutrals in what appears to be your critical loads panel needs to be connected to serve the function in item 1 above.
3. The ground rods are very important to provide a path to ground when needed for safely getting rid of excess voltage. They are for safety only and are connected to the neutral in one place only and that is called bonding and happens usually in your main panel.
4. If the ground is connected to the neutral in more then one place, then current that powers things can flow in the ground cable and create what is often called a ground loop.
5. If your service is 200 amps or more you will need 2 ground rods linked together but must be greater then 6 feet apart in the ground at your main panel.
6. If your extra panels are not close to your ground rod from your main panel you will need to drive a ground rod at each panel. This again is giving energy a easy place to go in an unsafe condition.

 

[mrdavvv]

Here are my observations reading through this and seeing some confusion.

1. Grid into your house is normally 2 legs at 120vac each that can be used to get 240vac and a neutral (not a ground). The "neutral" is the reference that keeps the 120vac lines stable when used independently and not as 240vac.
2. The neutrals in what appears to be your critical loads panel needs to be connected to serve the function in item 1 above.
3. The ground rods are very important to provide a path to ground when needed for safely getting rid of excess voltage. They are for safety only and are connected to the neutral in one place only and that is called bonding and happens usually in your main panel.
4. If the ground is connected to the neutral in more then one place, then current that powers things can flow in the ground cable and create what is often called a ground loop.
5. If your service is 200 amps or more you will need 2 ground rods linked together but must be greater then 6 feet apart in the ground at your main panel.
6. If your extra panels are not close to your ground rod from your main panel you will need to drive a ground rod at each panel. This again is giving energy a easy place to go in an unsafe condition.

Hello newbostonconst.

Thanks for your observations.

1.- Thats correct, i dont know if the diagram gives another impresion but thats the plan. 90% of our equipment its 120Vac, so the loads in the main panel are divided and balanced between the 2 phases.
2.- I see in the first diagram i didnt put the neutral in the critical panel, 2nd version is updated.



3.- Yes we are bonding at utility meter or main panel, one of the two.
4.- Got it.
5.- Service is less than 200A, but still considering multiple ground points just in case, as refered in new diagram. Grounding rods are >6ft apart.
6.- Critical loads panel is 12mt horizontal, and 5mt vertical from the main one. There is any rules about the point you mention?. Im not sure we can put a new ground near the critical load panel, as we would need to link together all the grounding scheme as that not possible in the building, we want to avoid the problems that filterguy mentions in here:



As i mention the installers also want to put a new ground rod for the critical loads panel and the PV system, but not sure how we can avoid the situation of the diagram.