Sunnyboy DC Overvoltage

[ovonrein]

I've got a sneaking suspicion that this could have been caused by leakage current flowing between the 2 strings to create a path for the 2 strings to effectively run in series.

+1 - very good thinking, Gavin.

[cliveb]

when you say he checked the cables, did he actually use a proper tester, or just check them visually?

Just a visual inspection, and only at athe inverter end - he didn't get up on the roof to check the panels.

I ask because I've got a sneaking suspicion that this could have been caused by leakage current flowing between the 2 strings to create a path for the 2 strings to effectively run in series.

This could possibly either be due to cables getting nipped in the frame, or leakage from the panels to frame, or badly fitted connectors between the panels, probably made worse by the dew getting into whatever the issue is (leakage from the panels is definitely worse when wet, as is insulation resistance in the connectors), or just a loose connection / cable that had been nicked in the isolators. If it was this, then it's possible that it would be at such relatively low ampages that the inverter would only pick up on it at start up when the available power was very low anyway, then at higher outputs it'd just ignore that tiny spike on the power curve.

Interesting analysis. It is certainly the case that these overvoltages only happened in the morning when the system was starting up. And on the days it did happen, there had been a heavy frost the night before, so I suppose some ice could have got in somewhere and bridged a gap.

It hasn't happened again, but if it does and Hatch suggest replacing the inverter, I will pass on your thoughts.

Many thanks.

[GavinA]

I just wanted to correct a mistake in my last post - it was 8 micro amps, not nano amps that we measured, I got the terminology mixed up.

[cliveb]

Time for an update. I was away on holiday last week, so I downloaded the event log when I got back, and sure enough there were several DC overvoltages in that time. Yesterday and this morning I noticed further DC overvoltages, and this time I made a note of the voltages reported on the SB front panel.

On both strings, the incoming DC voltage is shown around the 280V mark during the time that the overvoltage error was indicated. Assuming there are 8 panels per string, that's about 35V per panel, which seems right. Presumably if there was some kind of short between the two strings, the SB panel would show a voltage of double that?

This leads me onto a question. I don't really understand the difference between a "string" and an "MPPT". Is the 550V limit on the string or the MPPT? If both strings were wired onto a single MPPT, then at 280V per string that would put 560V into the MPPT - which is above the limit. As I said, I don't really understand the technical side of this, but any further info that I could pass on to the installer would be useful.

[ovonrein]

Excellent question.  I have the same inverter and just read thru all the docs but cannot find a clear answer.  It is rather baffling.  The datasheet writes Imax = 2 x 15A.  It does not write Vmax = 2 x 550V.  It seems likely that the SB4000TL only feature one single DC-AC inversion stage.  Which begs the question how the two strings feed into it?  A series connection might explain your problem but leaves us having to explain how such a circuit would equalize I (which is different on the two strings).  Mind a parallel connection would have similar difficulties with V.  So perhaps the SB4000TL has only one DC-AC stage but an additional DC-DC stage where the Wstring2 is turned into Istring1 * Vout.  Then Vstring1 + Vout make up the input for the DC-AC stage? Bizarrely, though, there would need to be some limit to what the DC-DC stage can handle, so one would expect this to mentioned in the datasheet...

The SB4000TL has two strings so that you can have different electrical characteristics on each string, allowing you to find two different max power points (hence the 2x MPPT).

EDIT:  Page 35 of the installation manual is highly suggestive of Vmax=550V on each string.  It is worth putting your question to SMA direct.

[ovonrein]

Check out this http://files.sma.de/dl/7418/Ableitstrom-TI-UEN114620.pdf - that doesn't read well.  Might be worth asking whether here is anyone on this forum running the 245W Suntechs on the SB4000 w/o transformer.  Or you installer should check with Suntech.

[BruceB]

In answer to the terminology question above, a string is a number of panels connected in series.  Strings can be paralled up outside the inverter or inside the inverter after the external input terminals.  I do not have an SMA diagram to hand which I think was the context of the question, but here is a Power One block diagram that shows a dual mppt inverter.

[marshman]

Check out this http://files.sma.de/dl/7418/Ableitstrom-TI-UEN114620.pdf - that doesn't read well.  Might be worth asking whether here is anyone on this forum running the 245W Suntechs on the SB4000 w/o transformer.  Or you installer should check with Suntech.

Not sure what the leakage current mentioned in the article has to do with the DC Overvoltage problem but does explain why a 30mA RCD is not a good idea with the 4000TL.

My second system consists of a 4000TL and 15 x Suntech 245WD panels. These are arranged as 2 strings - one with 9 panels and one with 6 - as advised by the Sunny Design program. Its never missed a beat except when we have a nearby thunderstorm when it usually takes out the 100mA RCD as soon as we get a nearby strike  (as does my other system with a SB3000 transformer version)

Roger

[jasonkirk]

I had a similar problem which kept cutting out throughout the day.  The cabling was changed/upgraded and more importantly the settings were still configured to European and they had to ensure changed to G83 on the inverter  - that did the trick and the problem as not occurred since.

[asorton]

I had a similar problem which kept cutting out throughout the day.  The cabling was changed/upgraded and more importantly the settings were still configured to European and they had to ensure changed to G83 on the inverter  - that did the trick and the problem as not occurred since.

Thats a very good point worth checking, many of these (several threads on here) have been left on the wrong settings (mine was left on german setting).

OP, have you checked its on UK setting on inverter startup or in settings via sunny explorer?

[cliveb]

OP, have you checked its on UK setting on inverter startup or in settings via sunny explorer?

Yes, the SB is on G83 settings.

Further things I've noticed in the last day:
- There are times when the panel displays DC Overvoltage and the red LED does not come on; other times the LED comes on.
- I've now seen a wide variety of displayed DC voltages when the Overvoltage message appears. The highest was 305V, and the lowest was 176V (!). Even double the 176V is nowhere near the 550V limit, so I suspect this is some kind of glitch/fault in the inverter's monitoring system.
- These overvoltages only ever happen in twilight conditions - usually in the morning as its starting up, but I've also seen it happen at the end of the day.

I'm going to call the installer back.

[micko]

How reliable are these inverters has anyone had one DOA. It sounds like a faulty inverter to me. 

[cliveb]

How reliable are these inverters has anyone had one DOA. It sounds like a faulty inverter to me. 

Yes, I'm beginning to think that myself. Yesterday it gave a DC overvoltage around midday during brilliant sunshine - not twilight conditions. It's been tripping in & out a lot over the last few days. I've seen the displayed DC voltage during the "overvoltage" periods reported as low as 176V (and as high as 307V). It seems to me that the inverter is simply throwing up phantom errors.

I called the installer two days ago, and yesterday, and again this morning. They are being rather less responsive than the first time I called them out. Maybe I should call SMA direct and see if they will talk to me.

[ovonrein]

Maybe I should call SMA direct and see if they will talk to me.

Send them an email.  They are very responsive and very professional.  This is unlikely the first time they hear of something like that.  They more than likely to have a suggestion what might be going on here.  I think it unlikely - but still possible - that the inverter is faulty.

[iantowil]

Cliveb, your system is around 6 months old with a problem that has been there since new - I wouldn't be as patient with your installer as maybe you are being. The profit on these installs is in the thousands, so the installers have already made more than enough profit to cover warranty work - get them out to change the damn thing! While the reason behind the faults are interesting, they shouldn't be occurring - if the same thing happens with another inverter, then they'll have to look elsewhere - but you've paid good money for fault free system, so make sure that's what you end up with.
Good luck!