question

jbpohle avatar image
jbpohle asked

Autotransformer 120/240 100A marine use USA w/Quattro 24/5000/120-100/100 230vac. Ground vs Neutral

Victron Autotransformer 120/240 100A used in a marine application. This is on a USA based boat in conjunction with a Quattro 24/5000/120-100/100 230vac. The neutral from all sources is abandoned, so we input to the Quattro the L1 and L2 from either the marina or our generator. The auto transformer ground relay is connected to the Quattro and grounds the created neutral when the inverter is providing the power. While connected to shore power the other day I inadvertently Grounded the autotransformer created neutral and got a small spark and the ELCI Equipment Leakage Circuit Interrupter on the shore power inlet tripped. So I measured voltages in the boat while on shore power at the same marina today as follows:

L1-G 123.6v

L2-G 124.2v

L1-L2 214v

L1-N (created by AT) 109.5v

L2-N (created by AT) 103.9v

G-N 62.5v

I see why I got a spark and tripped the ELCI. The system has been working well for about three years now, but I would never connect a neutral to ground normally on the boat. Is this normal?

MultiPlus Quattro Inverter ChargerAutotransformer
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.

jbpohle avatar image jbpohle commented ·

I was asked to measure voltages with other sources. Here are some additional measurements with the generator providing power and with no external power (Inverter) operation. Note that the Quattro does all of the switching between sources. I've included the original shore power voltages as well.

Voltage Source Shore Power Generator Inverter
L1-G 123.6 120 121
L2-G 124.2 120 114.6
L1-L2 214 240 235
L1-N 109.5 123.3 121
L2-N 103.9 117 114.5
N-G 62,5 3.4 0


Generator is a Northern Lights 12kw gen set. Load for Gen and Inv measurements was approx 500watts primarily 120v refrigerator.

Note also the ground relay on the Auto Transformer has grounded the neutral when the power source is the inverter. The Neutral on the Generator (it has one) is also grounded even though it is not connected.

0 Likes 0 ·
4 Answers
Justin Cook avatar image
Justin Cook answered ·

@JBPohle, where are the neutrals abandoned? If the neutrals are abandoned at the ELCI, then nothing inboard of the ELCI should be able to trip it, so that would be a clue that the issue is somewhere outboard of the ELCI -wherever that's installed. Are you able to provide even a rough pencil sketch of how your shore power comes in to the boat, where the ELCI is, where the neutrals have been abandoned, and how your AT is connected to the Quattro? Again, rough is fine, I'm just trying to get an image of the whole problem.

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.

jbpohle avatar image jbpohle commented ·

Thanks for the reply Justin. Here is a sketch of the system that includes the ELCI. My question here isn't really about the ELCI. The ELCI just confirmed that I had created a leakage to ground when I got the spark. The real question is do the voltages I measured make sense... especially the 62.5v measured between the Autotransformer Neutral and Ground.

0 Likes 0 ·
win-20200215-07-51-17-pro-2.jpg (338.4 KiB)
Justin Cook avatar image
Justin Cook answered ·

@JBPohle sorry for the delay; I'm working on getting someone more familiar with how the Autotransformer does what it does, particularly when in this particular configuration, to weigh in on this. My current working theory is that since the Autotransformer is using the created neutral to balance the output legs -meaning that at any given time there can be a significant amount of current on the created neutral- then it makes sense that you're getting a voltage reading between the created neutral and ground.

Again, however, that's my working theory but I'm inquiring with minds greater than mine -and more familiar with both the 230v Quattro and how the Autotransformer works with it- before giving that as a definitive answer. I'll report back here later today and either Alan or I will respond to you via email as well!

2 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.

jbpohle avatar image jbpohle commented ·

In general we don't have any heavy loads on the 120V circuits. Biggest usual load would be the refrigerator and some lights. Occasionally the fresh water pump 120v pump will cycle at about 800 watts for 10 to 20 seconds. All heavy loads are 240V. AC, Dryer, cook top, hot water etc.

0 Likes 0 ·
Justin Cook avatar image Justin Cook ♦♦ jbpohle commented ·

Understood, however there isn't a low-current threshold listed for the Autotransformer's balancing act, only a maximum capacity; as such it's quite conceivable that it will still be carrying current on the neutral under low load if the load on one leg is different than the load on the other leg - even if the total load is low.

Again, we'll be in touch with a definitive answer from someone with more experience in this particular use-case later, because since we don't regularly deal with 230 units, we're not entirely certain how the Quattro/Autotransformer is dealing with the neutral/ground connection in this case to begin with, so we're not entirely sure what to expect in voltage readings between the two.

On 120v units, obviously anything over ~2v between neutral and ground is indicative of a problem, but when there isn't a real neutral -and a device is using a created neutral to balance legs- obviously we're working under a slightly different set of rules.

We'll be in touch soon!



0 Likes 0 ·
ben avatar image
ben answered ·

Hi @JBPohle,

I run a similar setup to you in USA RV parks.

The first thing I notice is that you are not on a traditional 240V split-phase service. What you have is two 120v legs coming from a 120/208V three-phase service. This is code-legal since 2008 at RV parks and marinas in USA.

Unfortunately, it mean the L1-L2 voltage is low, and since your AT makes a new "derived neutral" at the halfway point, you get low voltage on your house side, too. You got lucky at this marina: the starting voltage was higher than usual. Somewhere, you'll find one that starts out nominal (208V) and then sags to 200V when everyone loads up the marina on a hot day, and then you'll be sending 100V to your appliances! Not so great. This is an unavoidable issue with this design, unless you are willing to add other components to your setup. (If you are interested in that, please start a new question about it and I will explain some of the options.)

Anyway, the next thing to note is that you definitely can't connect your derived neutral to your shore supply ground. The ground wire is a safety conductor, not a carrying one. The ELCI will trip the moment 10 or 15mA flow down that wire. Even if the measured voltage between your derived neutral and the shoreside ground were 0.5V, you'd trip it.

I think 60VAC measured is too high, though, and it makes me wonder if something else is also going on.

The first thing I would do is check to see if you have an inadvertent neutral-safety ground bond on your house load side. Do this when you are not on shore or generator, with the inverter's internal N-G bond relay open, and with the inverter's control signal to the AT not triggering the N-G bond relay inside the AT. You may have an appliance that is wired incorrectly.

If you see that all okay, the next thing you might do is put a low-Z meter on that pair of wires. Sometimes stray capacitance can build a "phantom voltage" on unused conductors. It may be fooling you, and the real potential there may be very low (but likely still enough to trip the ELCI; see above). The hassle here is that you may have to borrow a meter with a low-Z measuring mode. Or you could use it as an excuse to upgrade!

The last thing I'll mention is that the voltage split that your AT produces will indeed shift as the loads imbalance the two sides of the coil. That you observed that is normal and is not part of any problem you may have.

Caveat: I am not an electrician. However, most electricians also have no clue when you ask them for help with a system like this.

2 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.

jbpohle avatar image jbpohle commented ·

Ben thanks for the well thought out response. I think you are right about the 208v service. We run into it off and on at various marinas. The air conditioners don't cool quite as well and interestingly we think the microwave takes longer. Otherwise not too much of a problem. More and more marinas are getting their dock power squared away and we haven't felt the need for an iso-boost transformer yet.

I installed the ELCI within 10' of the shore power inlet per ABYC requirements. When I first installed it, 4 years ago, we had to track down a miswired battery charger. Once that was solved it rarely trips. Just for your information the whole boat ELCIs trip at 30ma here's a good link to Blue Seas information https://www.bluesea.com/support/articles/1381/AC_Ground_Faults_the_Boater_and_ABYC-Understanding_Equipment_Leakage_Circuit_Interrupters_%5BELCIs%5D_and_Ground_Fault_Circuit_Interrupters_%5BGFCIs%5D_to_make_your_boat_safer.

Because the ELCI never trips unless I connect Ground to Neutral while on Shore Power, I am certain that there is no N-G bond either on purpose or through a piece of equipment.

The low-Z meter sounds interesting and I'll take a look and see what that is. I am not afraid to buy a new tool or meter.

Thanks again.


John

0 Likes 0 ·
ben avatar image ben ♦ jbpohle commented ·

Hey John,

An inadvertent N-G bond downstream of your autotransformer may not trip an upstream, shoreside ELCI. (It depends on how you have your safety grounds wired, and you didn't specify... yet.)

Because it's relatively easy to test whether you have downstream N-G bond, I would start there if at all possible. Unless you don't have good access to your autotransformer's output wires or you can't readily power down because you have critical loads 24x7.

The behavior you are observing somewhat mimics what we have seen in a number of other installations (including my own, long ago).

0 Likes 0 ·
ben avatar image
ben answered ·

John,

I was waiting for someone else to verify my math before I came back again. It checks out: your neutral-to-ground voltage of ~60V is what you should expect when you are on a 208/120V shore supply. The phase angle of the two legs (120deg) creates this difference between your derived neutral and the shore ground (and the shore neutral).

I wouldn't worry about any of the readings you're seeing.


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.