Lightning Took Out Sungold 6kW LF Sine Inverter--Problems with Replacement Parts

[MWeiss]

On June 30th, we had a direct lightning strike on our east facing PV array. Ironically, the charge controller connected to it was not damaged and once the panels were replaced, the array made power again.

Not so fortunate was my Sungold 6kW LF sinewave inverter. It was charging the batteries at the time, running off the diesel generator. It went up in smoke.

I replaced the power board (heatsink) and the controller board ($490). Unit would not start up. Then I discovered that the display board has a microcontroller that also exploded. I replaced that for $39.

After replacing the controller, the unit would start up, so I thought that was it and installed it in the system. However, I found that it was unable to make the usual 55-60 amps of charge current. It was capped at 9.9 amps. And I suspected that the fans weren't operating as they normally operate when in charge mode. So I took it apart and tested both fans. Neither fan worked.

So I go online and order new fans, the closest I could find had lower amp rating, but the CFM was still pretty high. One fan is a tachometer fan (3 pin) the other is just 2 pin. I installed those today and figured that takes care of all electronics replaced, sans the power transformer and metal chassis.

So I hook it up and she starts making 52 amps into the battery, great. Then I hear the fans kick on, followed by BEEEEEEEEP! and it shuts down. It does this no matter where I set the charge current. It consistently shuts down as soon as the fans spin up.

My hunch is their controller may be looking for a specific fan RPM and if it is outside that range, it faults.

Does that make sense? I'm going to try to order the original fans from Sungold, but it takes weeks to get here from China. I've been ordering parts for this thing since June and waiting. Didn't want to wait for the fans, so bought these had them overnighted, but apparently there is some problem using other than OEM fans. Do fan controllers have a target speed they are looking for?

 

[AntronX]

Can you post pics of damaged panels?

 

[robby]

I wish you had made a post before starting the repairs.
I would have told you to dump the Inverter and buy a new one.
Over the years I have had plenty of experience repairing devices that have been hit by lightning and it typically ends up with just about every part with semiconductors being dead or half dead.
Even when I got the devices to work they would crap out days later.
I would love some more info on exactly what happened to your panels.

 

[Norwasian]

I would love some more info on exactly what happened to your panels.

And I would love to hear how and where the system components were grounded and/or bonded. Was this a grid-tied setup, or off-grid?

 

[Kenny_]

I wish you had made a post before starting the repairs.
I would have told you to dump the Inverter and buy a new one.
Over the years I have had plenty of experience repairing devices that have been hit by lightning and it typically ends up with just about every part with semiconductors being dead or half dead.
Even when I got the devices to work they would crap out days later.
I would love some more info on exactly what happened to your panels.

Yes, this. My first instinct would be to just throw it out and replace the unit...

 

[50ShadesOfDirt]

What we don't know, as others have posted above, is what types and levels of protection might have been in place in this particular incident. It would be excellent learning information if OP can post all that he knows about what was in place for his systems (lps, spd's, grounding, pic's, etc.)!

1. how is the wiring system is grounded (count every ground of any kind, on every panel array, load panel, generator, etc.)?
2. what SPD's are in place, and where (any SPD's on inverter, in load panel, or in front of expensive consumer devices)?
3. is any LPS in place, or now being considered?

A Lightning Protection System (LPS) could possibly protect something from a direct strike ... very few of us have or are even considering retrofitting this kind of system in.

Surge Protection Devices (SPD's) would protect from nearby lightning strike surges (lightning a few hundred feet away, hitting a tree for example), and they could protect from most grid surges (grid events, or lightning hitting grid components nearby), if you've layered these correctly. But, SPD's would probably not protect from a direct strike on the house or its systems.

 

[Luk88]

Most likely the fans you bought have nowhere need the airflow needed at the obstruction presented. It sounds like the inverter switches off with overtemperature. There might be an error displayed you can look up.

Often fan manufacturers will give huge cfm , but say nothing about the pressure which it is achieved. A fan that has 1000cfm at no pressure at all is very different than one than can power through lots of obstruction and still deliver same cfm.

Fan's installed in these Chinese inverters are usually 6k rpm fans about 10x more powerful than your random PC fan. They are extremely hard to find as replacement parts these days.

I too would live to hear more about the grounding the bonding, surge protection, what worked and what didn't (and if you had pictures of the panels that would be great). It is very interesting to see this kind of damage.

 

[MWeiss]

I suspect the controller is looking for a specific RPM and the substitute fan may be too fast or too slow to hit that.
As I said the unit starts up, runs for 8 seconds, then the fans kick on then it shuts down with the alarm. There is no error code on the simple display. The unit hadn't time to warm up, so it's no overheating. I can cycle the power off/on all day and it will continue to run for 8 seconds, have the fan kick on and then shut down.

The Sungold inverter is the ONLY piece of gear that failed. Now, for total disclosure, we had just gotten an MEP803A military generator hauled in the day before on a trailer. I had just connected it with a longer extension cable to hook into the system. The generator itself was not grounded yet, as it was to be removed from the trailer and placed on a pad soon after arrival. The storm hit just hours later. The generator was supplying power to an EG4 Chargeverter and this Sungold (in charge mode). The lightning hit the east-facing panel array on our west hillside. At first, I'd thought it took out the Growatt SC4880 charge controller, as it had gone dark after that event.

The next day I found out that the array was putting out 11 volts. I determined that the bypass diodes had shorted and so I popped the covers and measured a few, but visual showed a bunch of exploded diodes. After removing all the diodes, the array was making 165V and the charge controller was working. This array had taken a direct hit and the charge controller survived. But the wiring some 90' away, from the generator to the charging equipment induced voltage from the EMP and took out the poorly designed Chinese inverter/charger, while the EG4 product was still functional.

I've contacted Sungold again, and the fans are $70. At first they told me $35, but they invoiced me for $70.

I've replaced all of the electronics except the fans, transformer and chassis for about $550 so far. I could not see scrapping it because the replacement cost is much higher.

This event was bad timing. A newly arrived generator, being used in a condition of rainy weather and no PV generating capacity, the battery bank was exhausted and I had to start up the generator while it was still on trailer and make an extension cable to get power.

I've got a temporary ground rod for that generator now, but am waiting on repairs to my excavator ($3000 for hydraulic hoses) before I can move it off trailer and onto pad.

The array that got hit was the only legacy array of 20 used panels that was part of the pilot test and was not yet grounded.

 

[robbob2112]

You said there were 2 fans and they told you $35. I have to assume that price was each and you got invoiced for $70.

The othe tidbit was one of the fans you had was a 2 wire, which means no sense wire to determine speed by the inverter. So it tries to start fans and can't tell they are running so it turns off.

Fans are rated in amps and CFM. If a fan says 1000cfm but lists 1/2 the current that means it has a lot lower static pressure it can push against.

 

[MWeiss]

I asked her for the cost of the fans (plural) and she said $35. Apparently there was a misunderstanding.
One of the fans has a 3 wire connector. It's the input side fan by the MOSFETs. The 2 pin fan is the exhaust by the transformer.
Good point about the static pressure. But these fans move a lot of air. Maybe too much air and the RPMs are too high, so the controller figures it's out of range?

 

[Luk88]

If it doesn't even work long enough to heat up I agree it is not overheating. As there was a fast 3 wire fan in there it is possible it is not seeing the speed it wants.


But having spent ages tracking one last such fan in stockbin Europe (at a time one Alphacool distributor had it) I know first hand how difficult they are to find.

My fan parameters where 80mm x 25mm 6500rpm 12V 1.2A! 14W. I dont know the cfm, but let me tell you that fan when powered on easily lifts its own weight. The other one (one without the third wire) was a lot less poweful. 4000rpm and 0.6A.

For comparison's sake your typical PC fan at this size is 1~2W and if its speed reaches 3k rpm it's considered high speed The usual speed being 1800rpm (as it is really quiet at it).

 

[robbob2112]

Generally you have to use the part number and manufacturer to get the specs then try to match it.

 

[MWeiss]

Well it's probably a situation where their controller is not only looking for "fan spinning" but fan spinning within RPM range.

I've popped the $70 and will wait another few weeks as they ship from China.

 

[Luk88]

Well it's probably a situation where their controller is not only looking for "fan spinning" but fan spinning within RPM range.

I've popped the $70 and will wait another few weeks as they ship from China.

Can you post a link where did you get these fans from?

Edit: I see you got them from the inverter seller. OK.

Am I right to assume the inverter was grounded and only the generator connected to it (and the chargeverter) wasn't?

How long was the generator extension roughly? You said it was 90ft away but not the length. Knowing it is quite interesting.

I assume no SPDs (Surge Protection Devices) installed anywhere as you would probably mention them already.

Very interesting that the chargeverter survived. It would be interesting to know how exactly the generator extension was connected to both. First to the inverter then to chargeverter? Exactly parallel etc?

 

[mrzed001]

voltage from the EMP and took out the poorly designed Chinese inverter/charger, while the EG4 product was still functional.

Was the Sungold inverter grounded?

 

[Norwasian]

Was the Sungold inverter grounded?

Grounding, of course, provides a path for lightning.

 

[Luk88]

Grounding, of course, provides a path for lightning.

As the air itself too

Not grounding (or isolating) in order to prevent lightning surges is very hard to do if you have more than one wire coming from the outside. There is always a risk one wire has broke down to ground and the other has full surge voltage.

In such situation you're much better grounding your equipment and having an SPD on both incoming lines than counting on there not being path to ground.

There is also system capacitance to consider. Before potentials equalise certain charge has to flow even in absence of path to ground. Then if that high potential enters the device there will almost certainly be other parts that are at different potential causing possibility of mosfets blowing etc. It doesn't take a lot of current to blow the isolation layer of a mosfet through.

 

[Norwasian]

As the air itself too

Not grounding (or isolating) in order to prevent lightning surges is very hard to do if you have more than one wire coming from the outside. There is always a risk one wire has broke down to ground and the other has full surge voltage.

In such situation you're much better grounding your equipment and having an SPD on both incoming lines than counting on there not being path to ground.

There is also system capacitance to consider. Before potentials equalise certain charge has to flow even in absence of path to ground. Then if that high potential enters the device there will almost certainly be other parts that are at different potential causing possibility of mosfets blowing etc. It doesn't take a lot of current to blow the isolation layer of a mosfet through.

As you rightly point out, there are hazards both ways. With my particular setup, I'll take my chances with the isolation method. I would like to know of anyone here who has been fully isolated who has been bitten by lightning. As your post suggests, even a slight or unintended ground somewhere in the system would expose the entire system to risk of damage. I would like to know of any definite case of a fully non-grounded system taken out by lightning--and not one that had the exposure of a high prominence, either.

It's a bit like seat belts: In most accidents, they save lives; in some accidents, they take lives (though this gets no press). The common mantra is "always wear your seat belt!" But should you? I know of people saved by not wearing their belts, and killed by wearing them. If I'm driving beside deep water or cliffs, I prefer to not wear the belt. At other times, I usually do. There are pros and cons to anything, and relative risk should be assessed on a case by case basis. That goes for grounding, too. At least, that is my opinion.

Based on my own experience, lightning is attracted to ground rods; and I don't think any ground rod is capable of dissipating 30,000+ volts of electricity before it has a chance to damage anything else that happens to be attached to the rod. In the case of the OP, and for anyone else with long wires, induction from the EMF may be nearly as serious a threat as a direct strike--in which case grounding may be the better option.

DISCLAIMER: The opinions expressed in this post are my own, and not those of an electrician--I am not qualified to give any advice on the subject of grounding versus isolation. I would suggest simply that people consider the options and research the facts carefully, understanding that the typical advice is one-sided and few have considered the alternative.

 

[Steve_S]

Gee. a LESSON LEARNED HERE: The $$ spent on parts, S&H and the time, actually cost more than a replacement of equal grade and a repaired (sorta) unit without warranties etc. It is really easy to be Penny Wise & Dollar Foolish when you dive into repairs with good intentions. Proper mitigation & control for Lightning Strikes etc are a key part of being Successful with Solar Systems.

 

[MWeiss]

As mentioned, the generator arrived hours earlier. I was to mount it on a pad but my excavator is awaiting repairs.

So it was on a trailer and the storm popped up suddenly.

The generator was connected to an extension cable about 30' to the house.

The Sungold inverter was being used as a charger and was paralleled with the Chargeverter. The Sungold product was grounded at the case ground.

My take on it is the generator, being a large box, was an efficient collector of EMF from the lightning strike. The AC line coming into inverter/charger and Chargeverter probably took on a potential for that instant. The Sungold product was the only casualty, outside of 20 PV panels that were actually hit. I was surprised the connected charge controller was not damaged.

The lesson here is don't buy a generator and hook it up until you're ready to permanently install and then hook up properly. Generator on a trailer is not lightning safe.

 

[AntronX]

Found your video on Facebook. I love the sound it makes (at 8sec mark). Hope you don't mind if I post it here via youtube.

 

[mrzed001]

The generator was connected to an extension cable about 30' to the house.

The Sungold inverter was being used as a charger and was paralleled with the Chargeverter. The Sungold product was grounded at the case ground.

My take on it is the generator, being a large box, was an efficient collector of EMF from the lightning strike. The AC line coming into inverter/charger and Chargeverter probably took on a potential for that instant. The Sungold product was the only casualty, outside of 20 PV panels that were actually hit. I was surprised the connected charge controller was not damaged.

The lesson here is don't buy a generator and hook it up until you're ready to permanently install and then hook up properly. Generator on a trailer is not lightning safe.

Sorry this happened. This is a very good documented event ... so we all can learn from it

The inverter case ground is connected to the AC ground inside.
So I think the LEMP generated an impulse in the long AC cable.
And because only the inverter was grounded it was the only path to earth.
This saved your generator from damage (and standing on an isolated pad).
If it had a ground cable without an SPD then the generator could been damaged too.


The frame or the holder of the solar panels did melt on some part ?

 

[Luk88]

As you rightly point out, there are hazards both ways. With my particular setup, I'll take my chances with the isolation method. I would like to know of anyone here who has been fully isolated who has been bitten by lightning. As your post suggests, even a slight or unintended ground somewhere in the system would expose the entire system to risk of damage. I would like to know of any definite case of a fully non-grounded system taken out by lightning--and not one that had the exposure of a high prominence, either.

But how do you isolate a system that spans more than one structure? For example it includes a ground mounted array?

Or I guess you mean things like an RV install etc?

Found your video on Facebook. I love the sound it makes (at 8sec mark). Hope you don't mind if I post it here via youtube.

Wow. I'd have never guessed with these tall trees in the background it would hit the array.

Also, I'd have totally missed that in a loud storm. Amazing to have it captured on camera.

Can you say more about the storm? Was there lots of nearby strikes that preceded this? Or was it "out of nowhere"?

 

[MWeiss]

Yep that's the video my wife took. Glad you were able to find and post it.
The storm was particularly strong and the unusual nature of it prompted my wife to start recording. Too bad she was moving the phone camera back and forth and it was behind the curtain the moment of the strike, resulting in a blurry image.

I took down all of those panels and believe me, I expected to find at least one shattered completely, but I could find no visible damage other than every bypass diode being shorted.

I think had the generator been grounded, the LEMP would not have been as strong going back to the chargers. The Sungold has two grounds--one from the generator neutral and the chassis ground which went to earth ground. The Chargeverter had only neutral from the generator, no place to ground their chassis.

We're 1400' from a 500' tall broadcast tower, so in general, it acts as a lightning rod for the area, however, I've seen lighting strike the ground within 10' of a radio tower once (I used to work in the industry until retirement), so that obliterated my false notion that a tall metal object is always going to take the strike. Lightning can do strange things and almost demonically seem to "intentionally" hit where it wants, and not the "obvious" path of least resistance. Yes, there are 90' tall oaks right behind the array, on a hill slope that goes up 15' (those trees kill all my solar by 4PM when their shadow covers my entire property) but yet the bolt came straight down and appeared to hit the aluminum frame and then in the next frame of the video, jumped from frame to frame. Here is that single frame from the video:



I've noticed this year we seem to be having a lot of cloud to ground lightning, whereas normally it's cloud to cloud. A dangerous situation for us earth dwellers!

The new replacement array is in and frames are grounded.

 

[Kenny_]

Those trees too... I know lightning is unpredictable but ...