Isolating the Neutral

Electrically, it doesn't matter where you bond the neutral as long as you only do it at one place. If the generator neutral is unbounded then bonding the neutral at the main is correct, and vice versa. You just don't want it bonded at both ends because then there's a parallel path for neutral current to flow on the ground wire. 

The frame of the generator will be grounded by the ground wire COMING BACK from the panel.

6 hours ago, CHG_Coin said:

If the generator has a floating neutral, can it share the inverter's Neutral on AC OUT?

If the generator has a grounded neutral, can it share the inverter's Neutral on AC OUT?

NO, and NO again.

At least on the current Rev A.1 and Rev. B boards.  I will have to revise this on the B.1 boards, as it looks like this has potential to be quite an issue for customers.

The issue stems from the fact that split-phase AC has 3 wires, and at least 2 of them must be broken to fully disconnect an INPUT AC signal from the OUTPUT AC signal.  I worked around this issue by only providing 2 wires for AC Input...now only 1 of the 2 must be broken.  However, if you connect an INPUT signal to the OUTPUT leads, this is bypassed...and now the inverter can no longer disconnect an AC Input signal.

If the inverter is wired up with an INPUT split-phase neutral connected to the OUTPUT neutral, there are 2 possible results:

• if INPUT is turned on while the inverter is off...when the inverter is turned on, it will enter a "???" mode (I need to make this a "miswired" error)
• if INPUT is turned on while the inverter is ON...it will probably blow the inverter up, as the input AC signal is basically directly connected to the inverter output unsynchronized and uncontrolled.

Please note that with a single-phase 240v AC input to the inverter, it will still output split-phase 240v AC, thanks to the transformer's center tap.

6 hours ago, CHG_Coin said:

Are GS inverters floating neutral or bonded?

Floating neutral for the first shipment (Rev. A.1 control boards)...second shipment (Rev. B control boards) will have an option for neutral-ground bonding...third and future shipments (Rev. B.1) likely will have floating ground with an addon for a ground bonding relay if desired.  (Seems quite superfluous to me to waste board space and parts on that when most breaker panels tie ground and neutral together anyhow.)

In short: INPUT signals must not be connected to OUTPUT terminals under any circumstances--and that includes "indirect" connections via the neutral-ground bonding present on most systems/panels.

On 4/13/2021 at 1:54 PM, Sid Genetry Solar said:

In short: INPUT signals must not be connected to OUTPUT terminals under any circumstances--and that includes "indirect" connections via the neutral-ground bonding present on most systems/panels.

Does this mean the ground of the generator and the ground of the inverter, also, should be isolated from one another?  

There are three grounding points which come to mind -

Isolated Generator ground to earth
Negative Terminal of Battery as ground
Isolated Inverter ground to earth
 

This applies to the first few GS inverter shipments; I will revise the Rev. B.1 control board to disconnect both L1 & L2...which will resolve the issue in the future.

2 minutes ago, CHG_Coin said:

Does this mean the ground of the generator and the ground of the inverter, also, should be isolated from one another?

Only if ground is bonded to Output Neutral on the inverter via a breaker panel AND the generator's neutral line is also bonded to its ground...making a direct connection from Generator Neutral to inverter Output Neutral.

Sid, for my Version 1 unit, will the attached suffice?

Sorry, this being closer to what I was trying to express...

As long as the Sub-Panel also has a floating neutral.  The issue here is the potential for any ground-neutral bonding to turn the common shared ground line into an inadvertent power line.  (This is what bit Sean in an attempt to make the inverter's ATS work as a whole-house switch...umm...that common grounding and ground-neutral bonding bit us hard.  Frustrating when we don't think about these things when designing!)

I may need to work up some sort of solution for this and the 1st gen inverters (Rev. A.1 control boards).

You can test any hypothesis by wiring up your inverter as desired...and with it POWERED OFF, pop the generator on (i.e. provide power to the AC input of the inverter).  If there's AC voltage on the output terminals of the inverter, the ground-neutral bonding has turned the ground wire into a power line, backfeeding power through the inverter.  (This would be providing power through the autoformer principle of the transformer secondary; with the inverter off, it's not doing anything--nor can it control anything.  If turned on, it should enter a "???" mode and refuse to enter inverter mode.  If the inverter is already on when the generator is fired up...well, it likely will blow the inverter up.)

Here is the crux of the issue: in your 240v desired setup with the A.1 control board (and Rev. B as well), INPUT L2/N is literally wired straight through to OUT L2.  There is no disconnection relay on it.  The INPUT L1/L goes to a relay (which connects it to OUT L1 when AC input is desired.)  I obviously have to fix this on the next control board revision; maybe at some point once we have stock of a good design...then we can go back and provide free upgrade boards to customers affected by this issue.

I'm a bit confused still - with the wiring above in place it shouldn't have this problem?  But if something is wrong with the wiring I will need to make sure the generator is turned on first before turning on the inverter?  Or are you saying that even with the correct wiring in place, I would need to worry about such an event whereby I will always need to make sure to turn on my generator first, then my inverter?

How does this play into the automatic-on function of the generator which I'm working on next?

Sorry, I wasn't clear.  The generator turning on isn't going to be a problem.

What I detailed is just a one-time test to make sure there isn't any unexpected ground-neutral bonding that could cause inverter failure.  If the system passes the test, then you're good to go.  If the system fails the test, at least the inverter will not be damaged.

Is that why the slave gs inverter gets wired up with only two wires l1 and l2 to the input?  while the master gets wired to the output with l1 l2 and n?  so I don't blow up the slave?  

Sorta...largely because the design only disconnects L1.  I'll revise this in the Rev. B.1 board to disconnect both inputs--that way, you could do a full 3-wire split-phase daisy chain setup.

I'm getting a transformer polarity error.  Does polarity of H1 with respect to H2 matter?  Switch the two?

*warranty void*  🤪

Was bound to happen at some point 🤣😇

What did you change to get this error? 

If you disconnected the transformer wires and got 'em mixed up or whatever...that would definitely cause this issue.  Swapping T-L1 and T-L2 should fix it; alternately, you could reverse the 2 tranny wires going to the FET heatsinks.

Yes, the transformer polarity does matter--this is how it can detect AC power direction, as well as sync properly for parallel / 3-phase.

I'd feel a lot better if you'd explain exactly what you did...BEFORE trying it again 😉.  Changing the input voltage setup from 120v to 240v shouldn't affect any of the white transformer wires...so I'd rather get it sorted out before blowing anything up.