Autotransformer balancing - split phase inversion AND generator balancing possible?

Q&A and troubleshooting
, , , ,
1

Hi all, looking for some input for those who have used an Autotransformer to balance split phase legs on a North American split phase 240V setup. It’s an off grid cabin that uses a 240 split phase generator to charge the system and dual 48V 5000W Multiplus’s configured as 240 split phase. Multiplus AC output 1 goes into the Autotransformer to balance the Mulit’s, per installation schematics.

Here’s the challenge: When inverting, the Multiplus’s are perfectly balanced, regardless of load imbalance, at expected with the Autotransformer. However, when the generator kicks on and is charging/passing through power, the generator legs are not balancing. My theory is that when generator power is supplied, the Multi’s go into pass through mode and somehow don’t balance. All Multi output is on ACOUT-1. There are two breaker boxes, one on the input side (generator feeds it, L1, L2, N and G all used in this box. One more piece of information: I have Power Assist turned off. If I turn it on, will the connections be made internal to the Multi to allow the Autotransformer to do it’s thing? The generator does NOT have a N-G bond, at the gen or in the feeder box, the post Multi main box does, the autotransformer also has the ground relay active… yes, I should only have one N-G bond, but still figuring this out…

How about disconnecting the N on the generator side? Would that force the AT to do it’s thing even when passing through?

Unfortunately the installation is a couple of hours away and I’m not able to test these things easily. I want to have a solid plan in my head before I make the drive to modify and test.

My alternative is to add a second Autotransformer to the generator feed side, but if I can make a single AT work, that would, obviously, be ideal.

2

Welcome!

Please confirm/answer:

  • Are the input neutrals of the Multipluses connected together?

  • Are the input neutrals also connected to the generator neutral (should be)?

  • Are the output neutrals of the Multipluses connected together and to nothing else?

  • Are you doubly sure there isn’t a neutral-ground bond in the generator?

  • Is the autotransformer neutral completely isolated from the multiplus input neutrals (even through a common ground path)?

  • Does the main panel neutral connect only to the autotransformer neutral?

I’m speculating that the autotransformer neutral is connected to the generator neutral either directly, or indirectly through ground bond in the main panel. This puts the AT coil in parallel with the generator coil, and current is flowing back.

3

Hi Rick, thanks for the quick response.

To answer directly each of your questions:

  • Are the input neutrals of the Multipluses connected together?. YES, IN THE INPUT BOX.
  • Are the input neutrals connected to the generator neutral (should be)? YES
  • Are the output neutrals of the Multipluses connected together and to nothing else? YES
  • Are you doubly sure there a neutral-ground bond in the generator? THERE IS NOT, I MADE THE N-G BOND IN THE MAIN BOX (I reused two breaker boxes, one has the pre-sytem generator input and the other is the post-system main box)
  • Is the autotransformer neutral completely isolated from the multiplus input neutrals (even through a ground path)? HMMM… THE GROUNDS ARE COMMON.
  • Does the main panel neutral connect only to the autotransformer neutral? YES

Reading through this, is sounds like I should have bonded N-G in the supply box and not in the main box, and de-bond the G between the two boxes. That actually makes a lot of sense now that I better understand how the AT works. Am I on the right track?

4

Unfortunately, this stuff is really hard to diagnose remotely.
And we are both remote!

Bring a multimeter and a current clamp meter.
Look for unexpected current flowing on ground or neutral wires back to the generator.

Be careful when working on live circuits.
Don’t use an ohm meter to test for Neutral Ground bonds when the power is on!
I’ve destroyed a meter doing that. Volt meter setting is better.

You cannot separate the grounds, you must have one grounding system per code.
When inverting, any other inadvertent neutral-ground bonds before the inverter will not matter. But if it turns out there is one, then the input relays in the multiplus will connect to it when the generator is in use and you will see this behavior.

I don’t see a problem with a single N-G bond in the main panel.

I would get nervous about disconnecting neutrals, because a floating neutral will destroy your 120v devices, so beware. That includes the multipluses. Sorry I can’t provide further advice, you’ll have to measure things when you get there.

5

It’s hard enough to diagnose with all the proper test equipment in hand and standing in front of the system. Nearly impossible remotely… thanks for trying!

I’m going to try bonding the G-N in the supply box (feed circuit) and remove it in the main box (output). This will provide a N-G bond on the supply side when running the genny (correct), and the AT will bond the output side when inverting. This will remove any possibility of neutral backfeed through G (between boxes). This respects grounding requirements and is something I’m comfortable with. Not sure if this will then drive balancing of the generator or not, but it’s a safe thing to experiment with.

Thanks again for your feedback!

Cheers

1 Like