SMA 10kW inverter DI converter fault mystery

[Markyrocks69]

Was this ever solved. I'm curious as to what the final outcome was?

[oregon_phil]

It's difficult to find a solution to a problem since SMA didn't give a very good explanation of the fault and corrective measures. Two consecutive firmware updates (knock on wood) solved my nuisance arc fault issues, but could an unknown firmware update be causing the TL issues?

Why am I bringing up firmware? Because by default, SMA sets auto firmware update to YES. Inverter variables might get changed during these firmware updates. I had a Level 3 SMA tech tell to turn Auto Update to NO once my system was stable.

Just throwing it out there.

[Markyrocks69]

I might be misremembering some stuff. Thinking it over further I think there's actually a 'main' panel at the pole under the meter that feeds the shop and the house. So the panel at the pole would really be the 'main'. I thought the ground and neutral were bonded in the house.... I'll have to text him... could this cause intermittent faults?

Eureka. Seems like were getting somewhere. Ok if there is a main shutoff at the pole, like I said the neutral and ground should be bonded there. Now you need to find out if the grounds and neutrals are mixed in the panels inside the other panels on the property. The grounds and neutrals in these panels should be separate. Meaning ground on one bus, neutrals on the other. They cant be mixed. Even adding one neutral to the ground bar can send current down the grounds. This intermittent issue might only be occurring when someone just so happens to be using a specific circuit. When the circuit is idling nothing is happening but if someone is putting a load on the circuit then the amps on the bus change and could be the reason the error is happening.

[nwdiver]

I'm not talking about between the pole and the meter I'm talking about between the meter and the main panel.

I might be misremembering some stuff. Thinking it over further I think there's actually a 'main' panel at the pole under the meter that feeds the shop and the house. So the panel at the pole would really be the 'main'. I thought the ground and neutral were bonded in the house.... I'll have to text him... could this cause intermittent faults?

[Markyrocks69]

How does this have anything to do with answering question at the beginning of this thread?

Read the thread.

[Ampster]

I'm not talking about between the pole and the meter I'm talking about between the meter and the main panel.

..

How does this have anything to do with answering question at the beginning of this thread?

[Markyrocks69]

In California, I've never seen a disconnect or protection between the pole and meter. It's always been IN the main panel after the meter.

I'm not talking about between the pole and the meter I'm talking about between the meter and the main panel.

Apparently the nec doesn't specify a distance but if you Google it it varies slightly from place to place but most people are saying 6 ft. Apparently this is up to the discretion of the local inspector. Basically the code says the main service disconnect needs to be outside and accessible or inside at the closest point of entrance. 2 totally contradictory statements but whatever. Again I'm just trying to get the actual problem. The distance really isn't the issue.

[Mike90250]

Ya but how exactly does that provide over current protection to the run of wire between the meter and the main panel?....

In California, I've never seen a disconnect or protection between the pole and meter. It's always been IN the main panel after the meter.

[Markyrocks69]

Over current protection upstream of the main breaker is rather rare. I don't think I've ever seen that. The next OCP device is usually the fuse feeding the XFMR.

I think the ground is just grounded with a ground rod at the meter...

It's not rare at all. Like I said its code. Now I'm not going to claim I'm some idiot savant when it comes to electric but this is exactly how my property is setup and I wondered why. I looked into it and I'm just repeating what I found. I'm not trying to give a lecture I'm just trying to figure out what the problem is.

Saying the ground is just grounded with a ground rod isn't the answer I'm looking for. What is the point of it? What does the other end go to? If that rod is bonded to the neutral and the main is grounded by another rod and the neutral and that ground are bonded and the only thing connecting the 2 sets of rods is the neutral then theres all kinds of issues there. The whole point of the way the code is setup is so that voltage on the neutral travels from branch circuits to the bus and back up the pole in a straight shot. If you have heavy loads in the main panel and varying resistance in different directions that voltage could be traveling to places that it's not supposed to. And that is precisely what your problem seems to be. When these errors are occurring is this a heavy electrical usage time of day? Dishes being done, showers, coffee, making breakfast ect?

[Ampster]

Ya but how exactly does that provide over current protection to the run of wire between the meter and the main panel?

........

Good question, In my case it is 6 inches from the meter to the main breaker so I guess they figure the utility over current protection will suffice. In the case of overhead drops that wire size is often smaller than what NEC would require on the customer side.

[nwdiver]

Ya but how exactly does that provide over current protection to the run of wire between the meter and the main panel?

Anyways, what is the ground at the meter for? Is it bonded to the neutral in the meter box? If the meter box is considered the main disconnect then it should be. If it is then the neutral and ground should not be bonded in the main panel. The point of all this is to provide a path of least resistance. If you have varying paths of resistance to ground voltage on the neutral could be traveling in different directions including energizing the grounds.

Over current protection upstream of the main breaker is rather rare. I don't think I've ever seen that. The next OCP device is usually the fuse feeding the XFMR.

I think the ground is just grounded with a ground rod at the meter...

[Markyrocks69]

Around these parts the meter is the service disconnect... if they want to disconnect service they pull the meter.

Ya but how exactly does that provide over current protection to the run of wire between the meter and the main panel?

Anyways, what is the ground at the meter for? Is it bonded to the neutral in the meter box? If the meter box is considered the main disconnect then it should be. If it is then the neutral and ground should not be bonded in the main panel. The point of all this is to provide a path of least resistance. If you have varying paths of resistance to ground voltage on the neutral could be traveling in different directions including energizing the grounds.

[nwdiver]

Are you absolutely positive there is no service disconnect at the meter? I only ask bc by code there should be a shutoff within 5 feet of the meter. That's kinda the point I'm getting at is in theory theres should be a shutoff (molded case breaker) and the ground and neutral should be bonded there and nowhere else. I realize this kinda seems off the subject but if the neutral is bonded in multiple places that could be part of the issue.

Around these parts the meter is the service disconnect... if they want to disconnect service they pull the meter.

[Markyrocks69]

Are you absolutely positive there is no service disconnect at the meter? I only ask bc by code there should be a shutoff within 5 feet of the meter. That's kinda the point I'm getting at is in theory theres should be a shutoff (molded case breaker) and the ground and neutral should be bonded there and nowhere else. I realize this kinda seems off the subject but if the neutral is bonded in multiple places that could be part of the issue.

Another question is if the ground and neutral aren't bonded at the meter pole why is there even a ground out there? To ground the meter box? The meter box (I believe) should just be bonded to the neutral. I'm not 100% positive on that fact but fairly certain.

[nwdiver]

That's actually a very good point. I have heard (from the utility company) that the pole if it absorbs water can become energized and send voltage into the ground. If the ground for the ac side goes to a pole its possible that the ground is getting some current on it that way. The trick is to isolate the problem and move towards a solution.

As a side note you said you found a broken ground wire at the pole. How did you fix that and what does the rest of the grounding situation look like on the ac side? are the rods holding up? Is the waterlinew bonded? How many rods are there? I'm wondering if theres other deficiencies in the system.

Also how far away from the house is the meter? Is there a ac shutoff for the main service outside? Where is the ground and neutral bonded?

- I found a broken ground at the pole where the transformer is mounted ~200' away from the house but it's wired in parallel to 3 more transformers so there were other grounds further away. It looked like it had been cut by something hitting it. I took a piece of #6 bare copper wire and reconnected it to the ground rod.

- There are ~3 ground rods on the service side. One at the meter, one outside the home and one aux ground rod at the shop. I think the water line is pex now so no grounding there.

- I think the ground and neutral are bonded at the main panel.


Screen Shot 2019-08-27 at 10.08.44 AM.jpg