question

Phillip Gutowski avatar image
Phillip Gutowski asked

Unexpected output voltages from Autotransformer

I have a 230VAC source supplied by an Isolation Transformer, power then passes through a Quattro and then an Autotransformer to create 115/230v split phase. The ground relay in the AT is being controlled by the Quattro. While inverting, with the ground relay closed, all is well. The output voltages of the AT are correct as follows:

  • L1 to Neutral: 115V ✅
  • L2 to Neutral: 115V ✅
  • L1 to L2: 230V ✅
  • Neutral to Ground: 0V ✅


However, on shore power (or anytime the AT ground relay is opened) I am getting very different readings:

  • L1 to Neutral: 115V ✅
  • L2 to Neutral: 115V ✅
  • Neutral to Ground: 115V ❌
  • L1 to Ground: 230V ❌
  • L2 to Ground: 0V ❌

In this installation, shore power flows in as 115V is stepped up to 230V in an Isolation Transformer (Neutral and Ground are bonded on the output side), power flows through the Quattro and into the Autotransformer. I have tested continuity between grounds and neutral through the system up until the AT. Only after the AT do I get the incorrect voltages. I have triple-checked all wiring schematics and compared with my wiring.


[UPDATED 7/19/24]


Further tests were performed with the following results:


  1. Source: 115VAC Shore Power stepped up to 230VAC with Isolation Transformer.
    • Isolation Transformer Output: un-bonded
    • Autotransformer Output: un-bonded
      • L1 to Neutral: 115V ✅
      • L2 to Neutral: 115V ✅
      • L1 to L2: 230V ✅
      • L1 to Ground: 155V ❌
      • L2 to Ground: 75V ❌
      • Neutral to Ground: 47V ❌
  2. Source: 115VAC Shore Power stepped up to 230VAC with Isolation Transformer.
    • Isolation Transformer Output: un-bonded
    • Autotransformer Output: bonded
    • (Is this compliant with ABYC and other conventions?) I have left the system this way and am looking for confirmation that it is correct and safe.
      • L1 to Neutral: 115V ✅
      • L2 to Neutral: 115V ✅
      • L1 to L2: 230V ✅
      • L1 to Ground: 115V ✅
      • L2 to Ground: 115V ✅
      • Neutral to Ground: 0V ✅
  3. Source: 230VAC Single Phase Generator to AC Input 2 on the Quattro
    • Generator output: bonded
    • Autotransformer output: un-bonded
      • L1 to Neutral: 115V ✅
      • L2 to Neutral: 115V ✅
      • L1 to L2: 230V ✅
      • L1 to Ground: 230V ❌
      • L2 to Ground: 0V ❌
      • Neutral to Ground: 115V ❌
  4. Source: 230VAC Single Phase Generator to AC Input 2 on the Quattro
    • Generator output: un-bonded
    • Autotransformer output: un-bonded
      • L1 to Neutral: 115V ✅
      • L2 to Neutral: 115V ✅
      • L1 to L2: 230V ✅
      • L1 to Ground: 140V ❌
      • L2 to Ground: 87V ❌
      • Neutral to Ground: 34V ❌
  5. Source: 230VAC Single Phase Generator to AC Input 2 on the Quattro
    • Generator output: bonded
    • Autotransformer output: bonded
    • (Is this compliant with ABYC and other conventions?) I have left the system this way and am looking for confirmation that it is correct and safe.
      • L1 to Neutral: 115V ✅
      • L2 to Neutral: 115V ✅
      • L1 to L2: 230V ✅
      • L1 to Ground: 115V ✅
      • L2 to Ground: 115V ✅
      • Neutral to Ground: 0V ✅

Thanks to everyone, especially @KW4NP for the support. I'm learning a lot here and we have a solution figured out, I just want to make sure it's the correct one. In my head, all N-G bonds should be at each power source, including the iso transformer and the generator, but this produces the incorrect voltage across neutral and ground on the output side of the AT. Ultimately it makes sense to un-bond those and bond N-G on the output of the AT, but have we created any dangerous situation in the wiring before the AT?

AutotransformerIsolation TranformerQuattro-II
eclipse-single-line-230-115v.pdf (1.2 MiB)
at-1-split-phase-240v-to-120-240v-with-quattro-240.pdf (529.3 KiB)
5 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

kw4np avatar image kw4np commented ·

@Phillip Gutowski It would be helpful to have some additional details about your system, like the type (mobile, boat, fixed), where you are located in the world, and the system layout.

1 Like 1 ·
Phillip Gutowski avatar image Phillip Gutowski kw4np commented ·

Roger that. I'll work on sharing my schematics. In the meantime, this is a system in a 42' sailboat located in Annapolis, MD. Thanks for the support.

0 Likes 0 ·
Phillip Gutowski avatar image Phillip Gutowski commented ·

This post says it is awaiting moderation. @Guy Stewart (Victron Community Manager)

0 Likes 0 ·
kw4np avatar image kw4np commented ·

Neutral and ground are bonded on the output side of the isolation transformer?

0 Likes 0 ·
Phillip Gutowski avatar image Phillip Gutowski kw4np commented ·

Yes

0 Likes 0 ·
6 Answers
kw4np avatar image
kw4np answered ·

It seems to me that the ground to neutral bond on the output of your isolation transformer is being carried through the Quattro as would be expected. Then the autotransformer can't bring one leg up to 115 because it's already grounded at the source. If there are no safety concerns, try removing the ground to neutral bond on the output of the isolation transformer and see if that fixes your issue.

4 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Phillip Gutowski avatar image Phillip Gutowski commented ·

@KW4NP thanks. I'll remove the N-G bonding wire in the IT today to see if it provides any more data. I thought about doing this and permanently bonding N-G on the output of the AT, but that's not correct from a safety point of view. As it stands now, my GFCIs trip if I plug something in that has a grounding pin.

0 Likes 0 ·
Show more comments
echidna avatar image
echidna answered ·

It would be helpful to post a schematic for a situation like this.

It looks to me that you have not grounded the neutral of the shore power isolation transformer (I assume it has one). This would be required by ABYC. There may be other issues. Likewise ABYC requires the generator neutral to be grounded. The ground can be either at the transformer/generator or at the main panel. In general, the present version of E-11 requires that the neutral of a split-phase source of power is grounded (e.g., shore power isolation transformer output, generator output, and by implication, inverter output).

ABYC does not cover the use of an autotransformer to provide a neutral when a single-phase 230/240V inverter is employed. There is an ABYC subcommittee that has been working on this (I've been involved), but there is controversy about the various possible options. That said, various manufacturers/owners have successfully implemented installations including an autotransformer, but there are technical issues involved, such as circulating currents if both the autotransformer and source neutrals are grounded, or ending up with an ungrounded neutral at the sutotransformer output, both for the case where the inverter is passing through shore or generator power. I suspect that most ABYC-qualified techs. will not be cognizant of the issues involved.

One important issue is that the autotransformer needs to have symmetrical windings (the Victron unit does not) if both the the AT neutral is grounded with the inverter in pass-thru (also, the source needs to have accurately balanced output about neutral).

1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Phillip Gutowski avatar image Phillip Gutowski commented ·

Thank you for your comments. I believe you answered the real question by stating,

"ABYC does not cover the use of an autotransformer to provide a neutral when a single-phase 230/240V inverter is employed. There is an ABYC subcommittee that has been working on this (I've been involved), but there is controversy about the various possible options. That said, various manufacturers/owners have successfully implemented installations including an autotransformer, but there are technical issues involved, such as circulating currents if both the autotransformer and source neutrals are grounded, or ending up with an ungrounded neutral at the autotransformer output, both for the case where the inverter is passing through shore or generator power. I suspect that most ABYC-qualified techs. will not be cognizant of the issues involved."

This is the case for me. I have a single-phase 230V input from all sources: shore iso transformer, generator, and inverter. I have now made certain that all those sources have unbonded neutrals, and I am bonding the split-phase neutral after the autotransformer. I tested all of the other configurations I could come up with prior to accepting this solution.

Thanks again @echidna and @KW4NP for the assistance.


I am a practicing ABYC certified electrician, and this one got me. I'll be looking for future guidance from the ABYC on this issue.

0 Likes 0 ·
Phillip Gutowski avatar image
Phillip Gutowski answered ·

I believe I've found that ABYC E-11 covers the use of isolation transformers and the requirement for neutral-ground bonding primarily in sections E-11.18. There is a ton of information in the standards document, and it's difficult for me to discern the answer as it applies here. Are there any ABYC techs that can assist?

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

echidna avatar image
echidna answered ·

Sorry, my last sentence should read "...if both the the AT neutral and source neutral are grounded with the inverter in pass-thru"

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

echidna avatar image
echidna answered ·

I'll try that again: "...if both the AT neutral and source neutral are grounded with the inverter in pass-thru" to get rid of the extra "the" I thought I had excised!

Now just to add a bit: I confess I did not read through everything thoroughly on a first pass, and still have not in full detail (too much to read, too little time), but let me clarify a little: a one-line diagram does not a full schematic make. I suspect there is something wrong here that would be clarified by a full schematic showing the details of the split-phase wiring, with both active lines (L1, L2) and neutral (N) being drawn out in full.

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

echidna avatar image
echidna answered ·

To add some further comments on your latest solution (neutral grounded only at the autotransformer center tap, and I assume always so [not via the ground relay]), this is probably OK for normal conditions of your system.

I think there is a mistake in your original setup. Case three implies that one side of the 230V generator was grounded. This is incorrect for a split phase system. The mid-point of the generator output is what needs to be grounded. If the generator does not provide a split phase output (3 wires: line 1, neutral and line 2, with 120V between lines and neutral, and 240V between lines), then you simply cannot use it in such a way.

OTOH, case 5 is not consistent with case 3. Something is wrong here and that is why I suggested a full schematic.

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Related Resources

Victron Autotransformer product page

Victron Autotransformer data sheet

Victron Autotransformer manual

Additional resources still need to be added for this topic

Quattro II Product page

Quattro II Online Manual

Pre-RMA bench test instructions.

Additional resources still need to be added for this topic