PLEASE NOTE: If you had an account with the previous forum, it has been ported to the new Genetry website!
You will need to reset the password to access the new forum. Click Log In → Forgot Password → enter your username or forum email address → click Email Reset Link.
First, I'll say that I am NOT one of those "gotta follow the rules because that's the rules!" kinda person. But I do respect something that is right and true. The National Electrical Code is a profit driven machine. It is only written and published to make money for the National Fire Protection Association, a private company. There are many sections of the modern NEC that are expressly written to funnel money to manufacturers. But not the grounding section, Article 250. The grounding requirements of 600 volt or less systems in the NEC make sense once you come to grips with a key problem that is so common, electricians and engineers themselves are confused sometimes:
There's more than one kind of "ground"!
In electronics, ground refers to the point of reference considered as "0 volts". Often the negative of the DC supply, or the center point of dual power supplies. Ground can be a current-carrying conductor.
In a car, the negative battery terminal is connected to the frame and considered "ground", which is consistent with how an electronics tech might use the term. But for power wiring in buildings, IT'S DIFFERENT! It's even more confusing because colloquially, we have muddied the waters by using the term "ground" to mean two related but different functions.
That bare wire connected to a rod or water pipe
That wire is what the NEC calls a "Grounding Electrode Conductor". It's purpose is to connect the neutral of your electrical system to the actual planet Earth. This serves to direct high voltage events, such as lightning or power line faults on the primary, away from your internal wiring. Other than that, that Earth-connected wire does NOT have a use inside your building. The electrical system will function just fine without it. But that wire also happens to be connected, at your main service, to
That bare or green wire on all the branch circuits from your breaker panel
The NEC calls that wire the "Equipment Grounding Conductor" and that is where the problem starts. Yes, it is connected to the ground, but its use is not to direct any current to the Earth. Its use is to connect all the non-current-carry metal parts of electric appliances together to equalize the potential across them, and to provide a path BACK TO THE SERVICE NEUTRAL to complete a circuit in case of a line-to-case fault somewhere inside the building.
SHORT CIRCUIT CURRENT DOES NOT RETURN TO THE EARTH
Electricity flows in complete circuits. The Earth itself has too much resistance to allow a current large enough to trip a breaker most of the time. The bare "ground" wire is there to do that. So, at your service, the neutral is connected to one wire that goes to the dirt, and a second wire that goes to the round holes on all your outlets. I wish we would call the green or bare wire the "Equipment Bonding Conductor" instead of a ground wire. I think a lot of confusion would clear up.
So, one conductor of your system is connected to the Grounding Electrode Conductor and the Equipment Grounding Conductor. In a 120/240 volt 3-wire system, it is the neutral. 3 wires come in from the utility, and 4 wires come out. That new wire is not meant to carry a current under normal circumstances. So while it is true that the ground wire and the neutral are connected, never use the ground as a current-carrying conductor. If it ever becomes open, all the grounded metal cases and frames attached to the same ground will be energized.
The NEC generally doesn't care about how circuits under 50 volts are grounded, and grounding solar panels is another subject altogether, so let's assume everything up to the batteries is all A-OK code wise. Let's also assume this is an off grid setup. You have a beautiful Genetry Solar inverter ready to wire up to your breaker panel. The NEC would treat this as what it calls a "Separately Derived System", and it is basically the same as a utility supplied system. You have three wires coming from your inverter. How do you ground this to code?
Two Ways to Ground your Inverter
You have to connect that inverter neutral to ground, both the Earth kind and the equipment bonding kind. The code allows this connection at either the inverter, OR the first means of disconnect. But not both. It is called a "System Bonding Jumper" in the code. Without this connection, a short circuit from line to ground will NOT trip a breaker or blow a fuse like it should. There is no complete fault current path without that connection. Here is a pic of both acceptable ways. Pretend the transformer in the pic is the inverter output:
<a href="/monthly_2021_02/SeparatelyDerivedSystemBondingJumper250.30(A)Z(1).jpg.333a50c94e26a5ab11df72ab8b85786e.jpg" class="ipsAttachLink ipsAttachLink_image"><img data-fileid="114" src="//forums.genetrysolar.com/applications/core/interface/js/spacer.png" data-src="/monthly_2021_02/SeparatelyDerivedSystemBondingJumper250.30(A)Z(1).thumb.jpg.65e36ebb6d7cc025d53b1ea208c741e1.jpg" data-ratio="99.87" width="751" class="ipsImage ipsImage_thumbnailed" alt="SeparatelyDerivedSystemBondingJumper250.30(A)Z(1).jpg">
I'm good with everything you stated. Good info!
My question is about the earth-ground connection on a powerjack control board. They connect it to the receptacle ground and to the inverter case. Fine. If that's all you are going to do.
But, I connect my powerjack to my house breaker panel. Since neutral is the center tap of the inverter transformer, I've always been leary of connecting the inverter's earth connection to, well, anything. So I never have. That transformer center tap ends up being connected to the breaker panel neutral, which, as you so clearly pointed out, is connected to earth at one point or another. I don't know what reaction I would get also connecting the earth-ground connection on the pj control board to what ends up being that same neutral/center tap. On 110v control boards, the N connection would be effectively shorted to the earth connection. Seemed like a bad idea.
So I leave the earth connection connected to the inverter receptacle and case, but never carry it out of the inverter case. Maybe I'm worrying about nothing. Maybe I'm doing somehting 'bad'. If so, I'm ready for enlightenment.
I don't have a PJ, so I can't look myself, but is there continuity from that board ground to anything else like the high voltage output or battery input? On the inverter I own, the ground of the receptacle is connected to the metal case, but that is it. There is a lug to connect to an earth ground. So, in the case where my main panel is where neutral and ground bond together, the way to ground my inverter would be to bring a wire from the neutral/ground bar to the metal case of the inverter.
2 hours ago, InPhase said:I don't have a PJ, so I can't look myself, but is there continuity from that board ground to anything else like the high voltage output or battery input? On the inverter I own, the ground of the receptacle is connected to the metal case, but that is it. There is a lug to connect to an earth ground. So, in the case where my main panel is where neutral and ground bond together, the way to ground my inverter would be to bring a wire from the neutral/ground bar to the metal case of the inverter.
Inside a PJ inverter, there are 4 10nF (0.01uF) high voltage decoupling capacitors. One each from battery positive, and battery negative to chassis. Then another set from L1 and N to chassis. That's it.
I understand that UL1741 requires some form of ground-neutral bonding in inverters; on the GS inverters, we have a 50A relay that shorts ground-neutral together if needed (i.e. inverter mode, but not charger mode, etc.) This is pretty common in off-grid inverters; some squeak past the switching requirement with some big warnings and a spade terminal that the user can connect/disconnect for the necessary function.
<a contenteditable="false" data-ipshover="" data-ipshover-target="/profile/28-inphase/?do=hovercard" data-mentionid="28" href="/profile/28-inphase/" rel="">@InPhaseThoughts, ideas? We plan to pass UL1741, so we might as well get it all right before paying for a test!
All pj control boards have an 'earth-ground' connection point on the control board that is apparently only connected to output potentials through 4 mov's. No direct connection. Thru 2 mov's to L1 and thru 2 more mov's to L2 (or to neutral if its a 110v control board). Seems to me it probably wouldn't be an issue, but it probably also provides no benefit or protections in connecting it to house ground, so why take chances?
Might want to look a smidge closer. Most of those "blue devices" are really high voltage capacitors...look for a capacitor size code such as "103", a PCB reference of "Cxx", and a schematic drawing of a capacitor on the silkscreen. The ones I have are marked "CD103M". Kinda funny that the manufacturer (matching logo) doesn't acknowledge they make this size... https://www.songtian-ste.com/?post_type=products&page_id=16395/
I believe these caps were originally intended to sorta comply with UL1741, which requires that a grounded chassis basically has to be centered in the AC waveform. Works OK if the board is running 240v...but if it's running 120v, now the chassis is centered between L1 & N.
PJ DOES use MOVs, however...but the board reference ID is not "Cxx", it's "MOVxx". With the appropriate schematic symbol on the silkscreen, and it's directly across the AC input terminals. The one I have is marked "14D561K". Web search turns up it clamps at 925v...some protection, huh?? https://www.mouser.com/ProductDetail/Bourns/MOV-14D561K/?qs=CQ3B1E%2bbPs3fNJrywqtkng%3D%3D
21 minutes ago, Sid Genetry Solar said:Inside a PJ inverter, there are 4 10nF (0.01uF) high voltage decoupling capacitors. One each from battery positive, and battery negative to chassis. Then another set from L1 and N to chassis. That's it.
I understand that UL1741 requires some form of ground-neutral bonding in inverters; on the GS inverters, we have a 50A relay that shorts ground-neutral together if needed (i.e. inverter mode, but not charger mode, etc.) This is pretty common in off-grid inverters; some squeak past the switching requirement with some big warnings and a spade terminal that the user can connect/disconnect for the necessary function.
<a contenteditable="false" data-ipshover="" data-ipshover-target="/profile/28-inphase/?do=hovercard" data-mentionid="28" href="/profile/28-inphase/" rel="">@InPhaseThoughts, ideas? We plan to pass UL1741, so we might as well get it all right before paying for a test!
I haven't read all of UL1741 because I can't afford it🤣 But from what I know of it, I agree with the principle. Let the inverter handle all the grounding and bonding internally and keep the grounds and neutrals separate externally. So 4 wires come from the inverter. That would be best for a new system built from the ground up. But on existing systems with a neutral ground bond in place, it would be nice to be able to remove the bonding jumper from the inverter.
15 minutes ago, InPhase said:But on existing systems with a neutral ground bond in place, it would be nice to be able to remove the bonding jumper from the inverter.
So on the new GS Rev. B control boards, you would simply set the "Ground Bond" setting to "Off" to disable the internal ground-neutral bonding. Yeah, I like settings 😁. (Older Rev. A.1 boards don't have a ground bond relay.)
On 2/27/2021 at 10:30 AM, InPhase said:You have to connect that inverter neutral to ground, both the Earth kind and the equipment bonding kind. The code allows this connection at either the inverter, OR the first means of disconnect. But not both. It is called a "System Bonding Jumper" in the code.
On 2/28/2021 at 9:37 AM, Sid Genetry Solar said:So on the new GS Rev. B control boards, you would simply set the "Ground Bond" setting to "Off" to disable the internal ground-neutral bonding.
So if I'm understanding you two correctly if my panel has the neutral to ground bond I would connect the 4 wires (L1, L2, Neutral and Ground) from the inverter to my electric panel and set the inverter's "Ground Bond" setting to off?
If my panel wasn't bonded then I'd still connect the 4 wires and then have the setting on?
1 hour ago, Cali Carlos said:So if I'm understanding you two correctly if my panel has the neutral to ground bond I would connect the 4 wires (L1, L2, Neutral and Ground) from the inverter to my electric panel and set the inverter's "Ground Bond" setting to off?
If my panel wasn't bonded then I'd still connect the 4 wires and then have the setting on?
Ha, I guess that's the general gist. In other cases, like if the ground-neutral bonding was in a connected generator, you'd also have to set the function to "off."
Something about grounding loops...
Please note, the OUTPUT of the inverter must not be connected to a live breaker panel. At best, it'll error and refuse to start...and worst case, it'll blow up the inverter.
In my humble opinion, (OK, maybe not so humble), most of these regulations are designed to make money for government agencies/inspectors and/or contractors. That said, I don't want anyone to do anything unsafe. But, as long as your electrical panel is correctly installed and connected to ground properly, it doesn't really matter whether you connect the ground connection on your inverter control board to house ground.
Your powerjack split phase inverter neutral output wire will be grounded as soon as you connect it to your electrical panel. If one chooses to connect the inverter's ground connection to that same ground connection on your electrical panel, you are effectively connecting the inverter's ground connection directly to the inverter's neutral lead/connection. To me that seems like a poor choice. Maybe it's fine, but I'll leave mine unshorted.
4 minutes ago, Sid Genetry Solar said:Please note, the OUTPUT of the inverter must not be connected to a live breaker panel. At best, it'll error and refuse to start...and worst case, it'll blow up the inverter.
I'll second that! I have to admit I have blown up an inverter by accidentally connecting it to a live grid panel. Expensive mistake and boy did I feel stupid!
Might want to look a smidge closer. Most of those "blue devices" are really high voltage capacitors
Sid, I'm guilty of assuming those blue caps to be mov's. Sorry. My vision is not what it used to be.
In my humble opinion, (OK, maybe not so humble), most of these regulations are designed to make money for government agencies/inspectors and/or contractors. That said, I don't want anyone to do anything unsafe. But, as long as your electrical panel is correctly installed and connected to ground properly, it doesn't really matter whether you connect the ground connection on your inverter control board to house ground.
I obviously messed that whole quote thing up.
Might want to look a smidge closer. Most of those "blue devices" are really high voltage capacitors...look for a capacitor size code such as "103", a PCB reference of "Cxx", and a schematic drawing of a capacitor on the silkscreen. The ones I have are marked "CD103M". Kinda funny that the manufacturer (matching logo) doesn't acknowledge they make this size... https://www.songtian-ste.com/?post_type=products&page_id=16395/
You found that missing too huh? I looked it up when I blew up the Upower and they didn't make them according to their website. As far as the 10.3 boards go, the ones I have have places for 2 between Ground to Neutral and 1 from Line 1 to Ground. They only used 1 in each one though so one of the N to G slots is empty. And if you check with an Ohmmeter, you will not see any connection between the Case Ground and Neutral. But if you put the meter on a capacitance setting instead, it does show up as 34pf if done on the board and 35pf at the case and the output's Neutral terminal.
I did not bond my Ground to the Neutral in my breaker panel in the shed. I did ground the Ground Rod to the inverter case and the Grounds in the outlets. There is a GFCI outlet on the 120 Volt outlets and it is lit up when the power is on and it does trip when tested. Something I could not do when I was using the 120 Volt outputs of the Jupiter MSW 2000 Watt inverter or any of the 120 Volt Reliable PSW Inverters as as soon as you powered one on, the GFCI would trip. As I had read about most of the 120V HF inverters blowing up if wired to a house with a bonded panel, I built the automatic transfer switch for the refrigerator as a completely isolating unit. If in inverter mode, it uses the inverter Neutral/Ground. If in commercial power mode, it uses the Main Panel Neutral/Ground. I should check and see if the GFCI would function on the other leg of the 240 but do not think it would. Which would be another example of PJ not fully understanding split phase. The other being no filtering of Line 2.