Ground Fault

[jarrod9155]

  Last night I just happen to walk by my inverter and check on thing s and right then it throw a error message at me e018 ground fault  detected witch in return shut the inverter down and the dumpload kicked on . The mill is still making good power but the inverter just shut off once the watts start to climb on the output . I changed the rectifier didn't help check my junctions boxes no moisture everthing is good the only thing changed was it started raining last night and the problem got worse as the ran increased the ground fault isn't bad but enough for the aurora to shut off all my ac phases measure even , the mill isn't stalling from shorting . Has any one seen were maybe rotor rub on the stator could cause this . My air gap was a little on the tight side . My thoughts are it could be the wires up the pole I really doubt that used plastic conduit all the way up ,  or  some how the stator is leaking throw the pole down to ground Rob at solicity sent me a trouble shooter from aurora because they had quite a problem with this in the wind inverter its really sensitive to ground faults even small . Any ideas or trouble shooting techniques  for tonight after work would be great .

[SparWeb]

Ground faults are notoriously hard to track down.  I once found a ground fault where the Al jacket of BX cable became live (due to a stupid assembly error in a junction box, my fault).  Nothing happened until the cable came into contact with a metal building - which grounded out the circuit that crossed it.  Big mystery because the fuse didn't blow...  it set my dump load on fire instead.  Long story, sorry...  back to your problem.

Troubleshooting mysterious inverter faults is always helped by having the following things:

A complete circuit diagram of your system
A sensitive multimeter
A plug-in ground test device
The Aurora installation/user's manual
A complete circuit diagram of your system

Sorry no specific help about the Aurora but I thought it would help to warn you that some faults work against us homebrew types.  Few of us bother to diagram our entires systems (they evolve so fast, that diagrams quickly go out of date too).

[jarrod9155]

    I have a filling this is not gonna be fun .

[rossw]

The only thing I'd add is that if your inverter is tripping on ground fault, I doubt it'll be on the input (turbine/generator/rectifier) side, that just doesn't make sense.

I'd be much more suspicious of something on the output side. If you can, isolate the output entirely (both active and neutral) and see if it runs. If it does, try a local load directly across the output. If that runs too, then you can be reasonably sure it's in your AC side wiring.

Assuming you have multiple circuits, isolate all of them, reconnect the inverter and test again. Assuming it works, bring back on circuits one at a time until you find the offending one.

[TomW]

The only thing I'd add is that if your inverter is tripping on ground fault, I doubt it'll be on the input (turbine/generator/rectifier) side, that just doesn't make sense.

I'd be much more suspicious of something on the output side. If you can, isolate the output entirely (both active and neutral) and see if it runs. If it does, try a local load directly across the output. If that runs too, then you can be reasonably sure it's in your AC side wiring.

Assuming you have multiple circuits, isolate all of them, reconnect the inverter and test again. Assuming it works, bring back on circuits one at a time until you find the offending one.

Excellent advice, Ross.

Actually a Troubleshooting Circuits 101 level approach.

Excellent advice nonetheless.

Divide and conquer.

Tom

[jarrod9155]

I wish it would been simple but now I opened a can of worms lowered the mill and it was what thought the stator was making contact with magnets and creating stray current to ground with the help of the rain . The stator is saveable but the bearings in the trailor hub are totaled some how the bearing failed , its not really that old 6 months .

[Rob Beckers]

The only thing I'd add is that if your inverter is tripping on ground fault, I doubt it'll be on the input (turbine/generator/rectifier) side, that just doesn't make sense.

I'd be much more suspicious of something on the output side. If you can, isolate the output entirely (both active and neutral) and see if it runs. If it does, try a local load directly across the output. If that runs too, then you can be reasonably sure it's in your AC side wiring.

Assuming you have multiple circuits, isolate all of them, reconnect the inverter and test again. Assuming it works, bring back on circuits one at a time until you find the offending one.

Hi Ross,

With an inverter like Jarrod's Power-One Aurora the ground fault can't be on the AC (grid) side. When the inverter shows a ground fault, the cause is always on the DC side (or the AC wiring coming from the wind turbine), unless the inverter itself is broken, which is very rare and has its own set of diagnostics messages. These inverters are transformerless, there is no galvanic separation between input and output. The DC input in fact has roughly 240V AC superimposed on it (which makes it dangerous to touch even if there's no DC voltage coming in from the wind turbine!). When there is a path to ground on the input side it sets up a fault current due to the AC component that is always present.

For that reason both UL1741 and CSA 22.2 require that transformerless inverters have multiple redundant mechanisms build into them to detect these ground faults, and go offline if one occurs. The inverter test the resistance to ground of the DC inputs when it switches on, and if this is under 1 MOhm throws an error while staying offline. If that passes, it connects to the grid and monitors in real-time the ground fault current through a differential detector: By measuring the difference in AC current on both legs of the 240V grid connection it figures out what's disappearing into ground, just like a GFI protected outlet does. The total of AC and DC going to ground has to be less than 200 mA, if it is more, the Aurora will throw a ground fault error and go offline. There are also requirements for the total DC component, and if the ground fault current changes faster than 30 mA per second it will throw the same error (to provide some measure of human protection in case someone touches a live part on the DC side).

In case you wonder why it allows such a high ground fault current (200 mA); these inverters were originally designed for PV duty, and PV modules in essence are large capacitors (to ground). During normal operation the AC component present on the DC side of the inverter can set up a large capacity current. That's normal, and shouldn't trip the inverter. The (usually) long wiring and stator coils of a wind turbine have a similar effect.

Ground faults can be an absolute nightmare to track down. They are often voltage dependent as well, making it hard to measure them with a regular multi-meter. Putting an isolation transformer on the AC (grid) side of the inverter would solve the problem, at least for the first ground fault (and you now can't detect a second ground fault any more, which can set up large currents on the DC side). Isolation transformers come at a cost though: First there's literal cost, a 7kVA transformer goes for around $1,000 plus shipping. Then the efficiency takes a hit of a few percent, and they draw current even when there is no wind (making it a good idea to add circuitry to switch off the transformer/inverter if the voltage from the turbine drops below a set limit).

Hope this explains things a bit.

-RoB-

[jarrod9155]

Thanks Rob for the diagram on tracking down the ground fault it led me right to the stator . After close examination it looks like the seal went in the bearing and the greece after and bearing next and the stator after all becauce a 10.00 dollar seal went out .

[fabricator]

Did you use the original  bearings that came with the stub?

[jarrod9155]

I did, they were new and well greased

[fabricator]

With a lot of the Chinese bearings it's hit or miss as far a quality goes, they have had entire lots that never got heat treated, you would be better off using a good name bearing, they may also made in China but the brand names keep tighter control on quality.

[jarrod9155]

I used some skf made in Mexico, greased with a clear grease . And Today we had 24 hours of straight wind from 10 to 60 mph gust and I hear a slight squeak not to loud sounds like a little bird almost hope this one isn't already starting to go .

[12AX7]

Hi..

As already mentioned, ground faults can be a bear to trace, tempting some to bypass or eliminate them.
Because of this, it might not be a bad idea to start a thread devoted to GFCI's and troubleshooting them.
I wouldn't mind a clear write up on the new "ARC FLASH" breakers.

I've worked on machines with GFCI's and multiple heater circuits.   I recall there being times where isolating the hot wires did not isolate the problem.
One had to isolate the neutrals as well.  I don't miss those times!

ax7

[wooferhound]

Here is a discussion on Arc Fault breakers with some really good links to more information . . .
http://fieldlines.com/board/index.php/topic,131319.0.html

[Madscientist267]

Ground faults/loops ALWAYS suck major a$$...

At least it led you to your bearing failure before anything got more serious on the mechanical end... 

Troubleshooting ground problems almost always involves eliminating the grounds, one at a time, until you find the one that causes a change... They can be anywhere, and provide maximum hair rip-out power even in the 'simplest' of causes.

My hat is off to you, find it before it does more finding you. That's my best advice. In the case where multiple grounds are at work, this can mean very time consuming troubleshooting, and potentially more damage (like setting dumps on fire) if you don't locate the source of the problem. IMHO, it's better to take the system COMPLETELY offline and look for the problem step by step rather than looking for it 'later'. You'd almost certainly be sorry.

I know this isn't much help, but rest assured I feel your pain, and hopefully you'll ultimately find it, and post the results if nothing else to help a newbie.

I used to work on car electronics, and let me tell you, when a ground goes nuts in a car, it can be one of the most nerve racking experiences. I'm sure that a mill issue is several times that, but believe me when I say I know how irritating it can be.

Good luck, be persistent, and don't give up until you find it, lest you enjoy throwing money away on smoked equipment (as you've already seen...) 

Steve