Ac Volts bleeding into solar lines?

[tomy2]

It started with a shock from the metal conduit between the roof solar panels and the inverters. The conduit surface reads 170 volts AC., not good. There are two EG4 3000 hybrid inverters connected to a load panel split phase. and another sub-panel 60amp two phase breaker supplying grid power to the inverters. Four 48V 100amp hour 3G4 batteries too. there are 8 400-watt panels feeding each inverter. The multi-meter reads 170 volts AC at the solar cable inputs on both inverters with the solar disconnected, even with the 60amp input breaker off. I double checked all the wires feeding into the inverters, no problems. The grounds to and from the inverters are good . I sent pictures to SS and am waiting for a response. Maybe someone has seen this before, and can help. My limited electrical ability is at a loss.

 

[Quattrohead]

Ground wire 6AWG from panel frames to rack, to conduit, to inverter to house ground.

 

[pvgirl]

Sounds like the conduit isn't bonded to ground properly. Metal conduit, when it terminates in a plastic box, or at a concentric knockout needs a bonding bushing connected to ground.

 

[timselectric]

The conduit surface reads 170 volts AC.

From the conduit to what?
You are definitely missing some grounding, somewhere.

The multi-meter reads 170 volts AC at the solar cable inputs

This is normal. And why you should turn everything off when working on it.

 

[tomy2]

Ground wire 6AWG from panel frames to rack, to conduit, to inverter to house ground.

that definitely is good advice, but it doesn't explain why AC voltage is in conduit carrying DC volts

 

[tomy2]

From the conduit to what?
You are definitely missing some grounding, somewhere.

This is normal. And why you should turn everything off when working on it.

that is normal? AC volts at the DC input.

 

[Quattrohead]

that is normal? AC volts at the DC input.

It is on some inverters. It is not full current but more like a leakage from filter/suppression components.
Bonding/grounding is for several different purposes, all very important.
It is easy to get things running....getting them running safe and reliable is the hard part.

 

[timselectric]

that is normal? AC volts at the DC input.

Yes
As explained above.

 

[tomy2]

Yes
As explained above.

I would consider 170Volts more than leakage.

 

[zanydroid]

The DC solar lines on an non isolated system system are known to ride on the AC signal of the inverter during normal operation. And AC can couple through galvanic barriers (think transformer)

If you ground the conduit that will bleed off any induced voltage and current from the coupled AC.

 

[McKravitts]

I would consider 170Volts more than leakage.

With that much voltage I would recommend that any "new ground connections" attached to the system be done through a resistor first.
Why take a chance on blowing an expensive piece of equipment.
If you verify that the voltage drops signifcantly with a small amount of current through the resistor it's probably safe to ground the connection directly.

 

[timselectric]

I would consider 170Volts more than leakage.

It's more of an induced voltage.
Leakage, is a good enough word for it.

 

[zanydroid]

With that much voltage I would recommend that any "new ground connections" attached to the system be done through a resistor first.
Why take a chance on blowing an expensive piece of equipment.
If you verify that the voltage drops signifcantly with a small amount of current through the resistor it's probably safe to ground the connection directly.

This is not a bad idea but I'm pretty sure it's better (and possibly required by code/sanity sake) to solidly ground (no resistor) in the final config

If the 170VAC at the solar cable input is measured by ground to one/both of the DC conductors that's 100% believable and normal, right?

If the 170VAC is measured from ground to the ungrounded conduit that does feel kind of high for induced voltage.

 

[tomy2]

Ground wire 6AWG from panel frames to rack, to conduit, to inverter to house ground

SS says the panels should not ground to inverter or house, but to a rod

 

[zanydroid]

SS says the panels should not ground to inverter or house, but to a rod

SS first line support is unqualified. timselectric has a lot more knowledge than SS first line support and 99% of the turnkey installers that might turn up to your house.

 

[Adam De Lay]

Unfortunately this leakage is "normal" on the cheaper HF inverters when things aren't properly grounded.

 

[timselectric]

SS says the panels should not ground to inverter or house, but to a rod

Epic fail, on their part.

 

[pvgirl]

SS says the panels should not ground to inverter or house, but to a rod

The electric code disagrees, and all metal parts(panels, mounts, conduit, inverter, etc) need to be bonded together and to the existing building electrical ground.

 

[zanydroid]

SS first line support is unqualified. timselectric has a lot more knowledge than SS first line support and 99% of the turnkey installers that might turn up to your house.

At least from the stories of support woe, a lot of them are also really bad at knowing what they don’t know or admitting to it

 

[McKravitts]

SS says the panels should not ground to inverter or house, but to a rod

Most house grounds are rods.
You may be violating code by adding a "rod" to your electrical system.