Floating ground

[johnag]

I am installing a 7500 wat grid tie system . I am using used panels . I got everything installed and i am getting positive on the ground wire when grounded to earth ground . I took total voltage and counted back panels to find the bad panel . I then ran a jumper to that panel to be sure it was the bad panel . .When checked the ground moves on to a different panel . I just cant wrap my head around what is going on .Why am i getting positive feeding into the ground that floats ? thanks for your help .

 

[Mattb4]

Your description is hard to follow. Panels do not have a ground wire, they have a negative wire and a positive wire. Are you are getting a voltage from your array when measuring between earth and the positive output? Or is this measurement from your microinverters to ground?

Describe in detail how and what you are measuring.

 

[johnag]

Positive I get about 1/2 the volts to earth ground . from the neg I get about 1/2 to earth ground . I dont get a ground fault on the inverter . It is a string inverter . No microinverters . If i take the string down to 1/2 the panels the voltage moves to 1/2 the panels again . 1/2 the voltage goes each way . I will have 2 strings of 15 panels when done . I only got 15 up now one string . panels are canadianSolar 250 Maby it is how they work . I was thinking it would be all the voltage on the pos and none on the neg when grounded to earth ground . It is gonna be bad weather fir a few days so I get a break from it .

 

[BentleyJ]

What brand of inverter are you using? Most have specific instructions to the installer NOT to earth ground either Pos+ or Neg- from the PV array, reread the manual before wasting more time. Any metalic PV frames and/or mounting rails (most are aluminum) still need to be grounded, this is called the Equipment Ground vs. a System Ground which in most cases the DC System is not Earth Grounded and doesn't need to be because the circuits and wiring are insulated. That's why inverters have ground fault sensors in case there is an insulation failure. In normal operation there my be potential between the DC wiring and earth ground.
You may be thinking in terms of an automotive system where the chassis is part of the electrical system and is mistakenly called ground when its simply attached to the negative side of the battery.

 

[OzSolar]

You may not have a problem.

Story time.... So I've wasted some serious hours tracking down problems that ultimately weren't there so now a days I've learned to quickly recognize (sometimes) that when readings don't make sense it's most likely because either my meter or my technique is the problem. Take a moment and really think about how you're testing. EG: Are you truly isolating the array from the inverter and everything else? Do you get the same results with a different DVM? I've lost count of the number of DVM's that were lying to me for one reason or the other.

 

[timselectric]

Reading 50% of nominal voltage to ground on an ungrounded circuit is exactly what it should be.
You are reading the potential, but it would not be able to carry any current.

 

[johnag]

I have the solar string 100 % isolated and disconnected during the test with the multimeter . disconnected at or before the combiner box . As stated by timselectric . I was just thinking that the disconnect needs to be thrown to give the voltage direction . Like a loose ground wire on 220v the live wires can float .Thank you all for your help . When the weather gets better I will hook it up and put it to the test .

 

[Vigo]

Per usual, outing myself in hopes of learning something..

IF the array is not connected to the inverter or anything else, i don't understand how you would read voltage between the PV circuit, and some other point that has no electrical path 'back to source'. I might be misunderstanding the mechanism by which voltage is measured by a voltmeter, but in my mind the only way you'd be getting half the circuit volts, was if the 'return path' from the point you are touching back onto the PV circuit, happens to have similar resistance to the meter itself, something like 10 million ohms. Basically, the meter giving you a 'bs' number right before it stops being able to measure voltage at all?

Somebody help, lol.

 

[timselectric]

Per usual, outing myself in hopes of learning something..

IF the array is not connected to the inverter or anything else, i don't understand how you would read voltage between the PV circuit, and some other point that has no electrical path 'back to source'. I might be misunderstanding the mechanism by which voltage is measured by a voltmeter, but in my mind the only way you'd be getting half the circuit volts, was if the 'return path' from the point you are touching back onto the PV circuit, happens to have similar resistance to the meter itself, something like 10 million ohms. Basically, the meter giving you a 'bs' number right before it stops being able to measure voltage at all?

Somebody help, lol.

It's just a reading of potential difference between two conducive materials. It wouldn't carry current. And is only seen by a digital meter.

 

[johnag]

I powered it up today . The Neg wire had voltage . I counted back and it was panel 10 . I could not change it so i let it run . Tonight the voltage dropped because of weather and it moved to panel 12 . There is no ground fault on the inverter . I guess let it run and see what happens . It may be dif on a sunny day .