Hybrid EPS whole-home wiring topology + PV string split into PV1/PV2 — sanity check

[armo]

Hi all,

I’m looking for some technical confirmation (or correction) from the community on two installation topics related to a Fox ESS hybrid system. I’m not trying to criticize an installer — just trying to understand the correct and intended way these systems are meant to be wired and operated.

System overview

• Fox ESS 3-phase hybrid inverter
• HV battery (11 kWh)
• Whole-home EPS (no dedicated EPS sub-panel)
• Manual 1-0-2 transfer switch (3P+N) at the main panel
• Loads are manually managed during outages (intentional)

Question 1 — Whole-home EPS wiring topology

Design intent:
A manual whole-home EPS solution where:

• The inverter automatically enters EPS on grid loss
• The 1-0-2 switch only selects the house supply source (Grid / Off / EPS)
• Circuit breakers are used for protection/service only, not as operational switches
• No manual switching is required in remote locations (e.g. garage) during an outage

Current situation (image > current setup):
In the installed setup, both the inverter AC-IN (grid) and EPS output were routed through a garage sub-panel. This resulted in:

• Additional breakers in the garage being used as operational isolation switches
• A required manual sequence (turn breaker A off, breaker B on, then flip 1-0-2) to avoid unintended backfeed
• Operational complexity and dependency on garage access during outages

Proposed topology (image > proposed setup):

• AC-IN (grid) permanently connected to the inverter (grid reference)
• EPS output routed directly from the inverter to the 1-0-2 switch and then to the main panel
• No inverter or EPS paths routed through the garage sub-panel
• Garage feed remains independent of inverter operation

Question:
From a Fox ESS best-practice point of view, is this proposed topology the correct way to implement a manual whole-home EPS system?

In other words:

• Should the inverter + transfer switch alone be sufficient for safe EPS operation when wired correctly?
• Are additional operational isolation breakers normally required, or are they a workaround for an incorrect topology?

Question 2 — Single PV string split to PV1 & PV2 using a Y-connector

The PV array with 10 panels is currently wired as one series string, which is then split with a Y-connector into PV1 and PV2 on the inverter.

This raises two concerns:

• Both MPPT inputs are effectively fed by the same string
• The inverter appears to “see” half the current per input, raising questions about:
  • current limiting
  • MPPT behavior
  • whether any real benefit is gained vs using a single MPPT input

Question:
Is splitting a single PV string into PV1 & PV2 using a Y-connector:

• supported by Fox ESS, and
• electrically neutral in terms of total power/current?

Or does this configuration risk:

• reduced current per MPPT,
• sub-optimal tracking, or
• unnecessary complexity compared to using a single PV input?

I’m happy to be corrected if I’m misunderstanding any of this; I’m mainly looking to align the installation with Fox ESS design intent and avoid procedural workarounds if a cleaner topology exists.

Thanks in advance for any insight or experience you can share.

[Dave Foster]

This isn’t a quick answer and really needs a much more detailed analysis of your system showing all phases, neutral and earthing before it can be answered safely.

I’m not seeing an earth rod in any of those diagrams, when the grid fails you have to assume that you have completely lost the incoming feed. Usually Neutral and Earth are bonded with the incoming cable and so when they are lost unless you have a local earth connection your EPS L and N will be floating, and there will be no Neutral>Earth bond made - this means that any safety device (RCD) that relies on earth leakage will not work and any equipment that has a metal cabinet may now have dangerous voltages present.

I would not recommend the proposed topology, It’s illegal (in the UK) to connect an off-grid system to any of the incoming supply connections to avoid the potential for electric shock when engineers are working on the incomer - so in effect when you disconnect from the grid you must isolate all of the connections from the incoming grid.
At a minimum you would need to move the split below the changeover switch so the AC input is fed from the L1/2/3 from position 1 and isolated in position 2 but you need to resolve the earthing and neutral bonding safely.

If you haven’t already looked at products that can do this safely, i’d recommend taking a look at this company’s ATS https://www.eco-ess.co.uk/automatic-ch ... r-systems/

On the second part of your question relating to the 10 panels being serial wired and split into 2 inputs, i’m honestly amazed it has been done like this, having 2 independent MPPT’s trying to manage the same single string in parallel will not work effectively at all.

If the voltage and power rating of the panels is within scope of one MPPT, they should be wired into just PV1. If however the voltage or power rating is above a single string, then the string should be split into 2 distinct strings with 1 connected to PV1 and the second to PV2.

[armo]

Thanks Dave,

That makes sense. I left the ground wiring out of the drawing but the house is grounded via ground spikes.

I need to understand how a whole home manual EPS could work since the current wiring creates an open loop where; as soon as EPS energizes the whole home thinks it has ‘grid’ power again unless two addition breakers are installed to isolate
this back feeding. (Not sure I’m using the right terminology.)

I’m in Portugal: wondering if anyone knows any installers that have experience with Fox and EPS setups?

Thanks,