Sunny Boy SB3.0-1AV-40 continually rebooting

[Ultima357]

Hi everyone. I have a Sunny Boy SB3.0-1AV-40 inverter, about 7 years old now, having churned out just over 35Mw in its life. It has 2 strings of 10 x 150w panels each, all due south and no shading. (Location in sunny Cyprus, hence the power generation). I'm now trying to fault find an issue where it is constantly restarting, relay clicking in and out, sometimes a few times a minute. It's not consistent which is also a head scratcher but always occurs from around 9am to 11:30 ish but can also happen at other times too. 

 

So far I've found the strings are iffy, panels now some 12yr+ old as previously powered a battery support system. One string is 265v and the other 365v. Plan to sort that out this weekend having sourced some equivalent replacements from a similar system that has been dismantled. £0 so a good find. 

 

But, not withstanding the above, further experimentation of running on just one string, it still exhibits the same problem. I've only got basic voltmeter and current clamp and the inside of the inverter is one of layered circuit boards, so very unfriendly. Nothing obvious fried or smelly though. Curiously, the 8 large capacitors at the top of the unit are 4 at 178v and 4 at 0v. That seems odd. Measured simply from capacitor top to ground. 

 

Unable to find a circuit diagram, so at a loss as to what else to measure, but beginning to think the inverter is the issue. Oh, no great problems reported on the monitor either. Used to get grid over voltage shutdown but not now, it being typically 240 to 245v. Any ideas? 

[SteamyTea]

SMA are a bit iffy about people tinkering with their inverts.

Is it still covered by warranty. Seem to remember they offered replacement once the old one is sent back. Call the local agent and see.

 

1 hour ago, Ultima357 said:

35Mw

That would fry it for sure, especially when written correctly, kW.

But you mean kWh.

[Mr Punter]

35 minutes ago, SteamyTea said:

But you mean kWh.

 

I think they meant MWh.

[Ultima357]

Yep, meant MWh. 

[Ultima357]

46 minutes ago, SteamyTea said:

SMA are a bit iffy about people tinkering with their inverts.

Is it still covered by warranty. Seem to remember they offered replacement once the old one is sent back. Call the local agent and see.

 

That would fry it for sure, especially when written correctly, kW.

But you mean kWh.

Not sure what the warranty is, but support is limited out here. Yes I meant MWh produced since installed. 

 

Just checked, warranty 5 Yr. 10 Yr if installed after October 2021. Dang! 

Edited October 22 by Ultima357
Additional info

[Ultima357]

End solution. Well I managed to borrow a 2nd hand SMA inverter and plumb it in which proved that the issue was in the arrays. They appear to be leaking to earth. After metering each individual panel, 7 were clearly end of life and even getting an array of 10 that seemed OK, you could measure neutral to earth and see the voltage climb to more than 20v just before the inverter tripped out. So plan now is to renew the array and put in another earth rod. A job for the spring time. 

[Beelbeebub]

How old are the panels?

[SteamyTea]

There has been a debate, for at least 2 decades, about Earthing PV panels (and frames).

Some inverter manufacturers say not to, others say you must.  What does the SMA documents say for your inverter.  There is a difference between transformer and transformer-less systems.

 

Bit here about it.

https://www.renewsolar.co.uk/uncategorised/delving-into-bs-7671-section-712-and-solar-panel-systems/

[sgt_woulds]

When I began installing solar panels (twenty odd years ago) we did not earth the panels unless they could be touched from an open window on a roof, or were ground mounted, or there was an LPS within 10m of the array. 

 

There was a (Swiss?) study at the time that said earthing panels increased the risk of lightning strikes.  All inverters back then had big iron core transformers so there was no ground path to 'attract' the heavens wrath.

 

The first time I ever earthed a panel was when IKEA introduced their PV systems using (god awful) CIGs panels.  We questioned the manufacturer repeatedly if this was a good idea.  I remember those panels gave us lots of static shocks during install which we'd never experienced before.  Dodgy stuff, as were most of the IKEA installs at that time, but I digress...

 

 Around this time, transformerless inverters were introduced, and without the robust physical seperation array earthing became more common, (but we always had to check manufacturers recommendations).   With the introduction of micro inverters that extended the EPZ onto the roof, SPD's became common and array earthing became standard. 

 

I've not seen any recent studies that compare earthed systems to unearthed for lightning risk, but I'd be happier not introducing ground path inside my house.  

 

I never liked transformerless inverters; even though they are (slightly) more efficient and easier to carry, they are not as rugged.  I know of a customer who's SB700 has been going strong for 23 years, but the transformerless inverters in the same building have already required replacement after 10 years.   Of course, this may be due to the cheaper electronics and PCB used today; those old SB's had good old seperate circuit components.  

[-rick-]

1 hour ago, sgt_woulds said:

I never liked transformerless inverters; even though they are (slightly) more efficient and easier to carry, they are not as rugged.  I know of a customer who's SB700 has been going strong for 23 years, but the transformerless inverters in the same building have already required replacement after 10 years.   Of course, this may be due to the cheaper electronics and PCB used today; those old SB's had good old seperate circuit components.  

 

I'm pretty sure you and Steamy know this, but transformerless inverters are not transformerless. The transformers are just smaller, higher frequency ones. The amount of isolation provided is down to the manufacturer and is a separate parameter.

 

High frequency ('transformerless') inverters have got a bad rep, because it is newer technology and plenty of manufacturers made mistakes in there design but I think we should be at mature technology now, and the efficiency and weight/size advantages are not to be sniffed at. Victron who for years made a big fuss about how their traditional low frequency inverters were so robust have now launched high frequency inverters (MultiRS/InverterRS) and have stated that they are now comfortable that the technology is robust enough (though admittedly even their first revision of the MultiRS apparently had issues).

 

Most newer inverters now that promise high efficiency >97% are likely to be high frequency as low frequency has inherently more losses.

[SteamyTea]

24 minutes ago, -rick- said:

and Steamy know this, but transformerless inverters are not transformerless.

I didn't, but it has been over a decade since I was involved in specifying systems.

 

Still no consensus on whether to earth or no.

We never used to, and I am not sure how sensible it is to inject a DC voltage down the neutral side of the incoming supply (TN-C-S, protective multiple earthing).

Probably not a problem at 20V, but 600V, even for a short time, may cause issues.

 

As for lightning strikes, a bit of 4mm cabling won't help much when 1 GJ of energy goes though it.

[-rick-]

34 minutes ago, -rick- said:

but transformerless inverters are not transformerless. 

 

Upon reflection I feel I should clarify this. All the 'transformerless' inverters I've looked at in detail in fact have transformers but I'm not saying that there are zero truely transformerless inverters out there. Electrically it's possible to create something truely transformerless. Transformerless has been used as short hand for 'doesn't have the whacking great 50Hz transformer'.

[sgt_woulds]

 

2 hours ago, -rick- said:

 

Upon reflection I feel I should clarify this. All the 'transformerless' inverters I've looked at in detail in fact have transformers but I'm not saying that there are zero truely transformerless inverters out there. Electrically it's possible to create something truely transformerless. Transformerless has been used as short hand for 'doesn't have the whacking great 50Hz transformer'.

 

This is where my understanding of the magic that happens inside inverters ends 🙂 

 

When TL inverters were introduced, SMA and Mastervolt told us that there was no conventional galvanic seperation, that's all I can remember.  We had endless tripping problems with DC leakage until we changed to type B RCDs.   

 

 

2 hours ago, SteamyTea said:

 

 

As for lightning strikes, a bit of 4mm cabling won't help much when 1 GJ of energy goes though it.

 

 

Agreed, array earth is not designed to be an LPS!  But earthing the panel means that you have now created a path inside the house, where previously a strike on the building would more likely flow external to ground via the Faraday effect.

 

The essance of the lightning study I read back then, was that earthing an array increases the chances of a lightning 'return stroke' forming when a storm is negatively charging the air above the building.  Having an inverter that is not Galvanically isolated can't help things, so you are making it doubly likely to occur, and then you are directing the surge nicely into the consumer unit too.

 

[I remember learning that lightning starts on the ground in GCSE science a billion years ago....]

 

 

 

Edited yesterday at 14:12 by sgt_woulds