Lux Power Hybrid - EPS signal

[Tinbum]

Page 11 of this gives a good idea of how SMA do it for their Sunny Island systems. You can by a box off the shelf but its very expensive!!

https://files.sma.de/downloads/SI44M-80 ... -en-11.pdf

[Oldgreybeard]

Could always just use the low value resistor link between N and E as a method that works OK but that doesn't need any relays, sensing or switching. Fitting something like a 50 ohm power resistor between N and E does nothing when the system is connected to the grid. Only added constraint is that the TT earth electrode resistance mustn't be higher than 150 ohms, rather than the 200 ohms that is recommended as the highest value.

If the system switches to backup mode then the N is referenced to E via the 50 ohm resistor and this is plenty low enough to stop the N from floating relative to earth and will still allow all the protective devices to work OK. This method was used (may be still is used) to tie IT installations (mostly comms related in the UK) to a reference so that the line and neutral stay the "right" way around relative to earth potential.

[sharpener]

Could always just use the low value resistor link between N and E as a method that works OK but that doesn't need any relays, sensing or switching. Fitting something like a 50 ohm power resistor between N and E does nothing when the system is connected to the grid. Only added constraint is that the TT earth electrode resistance mustn't be higher than 150 ohms, rather than the 200 ohms that is recommended as the highest value.

If the system switches to backup mode then the N is referenced to E via the 50 ohm resistor and this is plenty low enough to stop the N from floating relative to earth and will still allow all the protective devices to work OK. This method was used (may be still is used) to tie IT installations (mostly comms related in the UK) to a reference so that the line and neutral stay the "right" way around relative to earth potential.

Or according to Wikipedia, as used in mining installations in India...

As you pointed out earlier the 200ohm limit is to ensure stability of the earth electrode, the maximum for a 30mA RCD to trip with the N-E voltage not to exceed 50V is actually 1667ohms. You could achieve the same with a 2.2 uF capacitor from an old fluorescent fitting, this is perhaps more elegant as there is no real power consumption possible, only a (usually helpful) capacitive VAr requirement.

I had previously thought about doing this when I suspected the neutral of the incoming mains was being pulled off centre but the problem has since disappeared. I might still do it after the LA inspector has been and gone.

[Oldgreybeard]

Could always just use the low value resistor link between N and E as a method that works OK but that doesn't need any relays, sensing or switching. Fitting something like a 50 ohm power resistor between N and E does nothing when the system is connected to the grid. Only added constraint is that the TT earth electrode resistance mustn't be higher than 150 ohms, rather than the 200 ohms that is recommended as the highest value.

If the system switches to backup mode then the N is referenced to E via the 50 ohm resistor and this is plenty low enough to stop the N from floating relative to earth and will still allow all the protective devices to work OK. This method was used (may be still is used) to tie IT installations (mostly comms related in the UK) to a reference so that the line and neutral stay the "right" way around relative to earth potential.

Or according to Wikipedia, as used in mining installations in India...

As you pointed out earlier the 200ohm limit is to ensure stability of the earth electrode, the maximum for a 30mA RCD to trip with the N-E voltage not to exceed 50V is actually 1667ohms. You could achieve the same with a 2.2 uF capacitor from an old fluorescent fitting, this is perhaps more elegant as there is no real power consumption possible, only a (usually helpful) capacitive VAr requirement.

I had previously thought about doing this when I suspected the neutral of the incoming mains was being pulled off centre but the problem has since disappeared. I might still do it after the LA inspector has been and gone.

Used to be very common in mining installations to use an inductor rather than resistors for the neutral to earth links, just to reduce the loss in the neutral earth resistors that would otherwise be used. I remember seeing gensets that used resistors, often with a link bar, for the frame to neutral connection, the idea being that the genset could be used in IT mode (as the vast majority of small generators are very often used) or could be used with an earth electrode if that was thought necessary. Our local market probably has around a dozen small generators running as IT supplies every week, mostly powering metal trailers. As the market is held on a paved area, there are no local earth connections anywhere.

[Swwils]

[dupe]

[Swwils]

I don't have an earth rod yet. But will get one once I've sorted what to actually do for circuit switchover. The inverter does have UK ratings. But I can't find any info about internal relays. It does click clack when it loses grid power and click clack a few minutes after grid AC is returned.

Sounds promising, that will almost certainly be the anti-islanding relay(s), the delay is while the inverter is matching its voltage, frequency and phase to the grid.

Might be worth checking if you get the same result to OGB's test whether it is in Mode A or Mode B.

My inverter is on a rcbo so tested by flipping the breaker:

With grid on to inverter:

• N-E on the EPS output was 0.3 ohm.
  
  E on the EPS output to install earth was 0.3 ohm.
  
  N on the eps output to install earth was 0.3 ohm.

On flipping the breaker off:

• N-E on the EPS output goes to 1.8k ohm
  
  N on the EPS to install earth goes to 1.9 M ohm
  
  E on the EPS to install earth goes to 0.3 ohm.

N-E changes coincided with relay clacking, i.e. almost immediately.

With EPS mode turn off in software:

Grid on:

• N-E EPS output was 0.3 ohm
  EPS output E to install E was 0.3 ohm.

Grid off:

• N-E EPS output was 0.L which is out of range high on my meter
  EPS output E to install E was 0.3 ohm.

During my testing after a few switchovers the inverter went into a standby connection mode and began to delay the relay switchovers by 30s.

[Swwils]

I've had a reply from lux power.

No units currently have a N-E bond relay inside.

They then also sent some questionable diagrams to help solve this.

One lists a parallel from grid neutral to EPS neutral. The other earth the EPS neutral line.

As far as I can see what these diagrams suggest is not allowed under the regulations.

[Oldgreybeard]

The reply from Lux doesn't surprise me, although I'm surprised there isn't a neutral earth link. Getting key details about some of these inverters is like pulling teeth, the manufacturers of some of them are exceptionally poor at providing easy to understand and complete manuals.