Deadly AC current on 12V battery terminal

[mrzed001]

Not a joke

Here is a little video with measurements demonstrating it:

In Hungarian language, but in a browser you can turn on subtitles and set it to Auto translate to English





What you see is a simple setup. From left to right

1. 12V battery
2. 230V EU style inverter (Line - Neutral = 230V), 500W (2,1A). A really small inverter
3. Ground rod
4. RCD 30mA. The blue Neutral is connected to the PE cable (the bond). Not connected to ground rod first.
5. A second ground rod representing a Person
6. Multimeter for current measurements

First test:

• Inverter works, L-N: 202V (small inverter )
• Line to Ground: 0V.
• System is not grounded.

Second test:

• PE (bonded to N) connected to Ground rod.
• N-Ground: 0V
• L-Ground and L-Person: 202V
• RCD test successful, RCD disconnects
• It is a grounded system now

Third test (with the previously grounded system):

• Battery terminal - Person : 104Vac

Fourth test (with the previously grounded system): Current measurement

• Battery terminal - Person : 0,7A = 700mA (from the inverter's max 2,1A)
• If he would touch the battery terminal ... that could mean lethal shock.

Fifth test (with the previously grounded system): Current measurement

• Battery terminal - Ground rod : AC Short Circuit (sparks)
• Inverter stops working

The reason is: the inverter is a NON ISOLATED inverter.
There is no isolation between the AC part and the DC part.
And so the AC part will be present in the DC side.


Andy run into this problem too, but did not investigate it further.

The All-in-one inverters are usually isolated to battery
The big red circle is the isolated DC/DC converter in the MPP Solar like inverters



Also I want to mention that the MPPT (below the circle) is NON ISOLATED.
This is the reason you should not ground the PV cables in these systems.
That would mean immediate AC short circuit.

 

[John Frum]

I must have a fundamental mis-understanding of electricity.
I thought that for current to flow it must return to the source.

I don't see how the dirt circuit gets back to the inverter.
Is the return path through the OSB board and inverter chassis?

 

[mrzed001]

I must have a fundamental mis-understanding of electricity.
I thought that for current to flow it must return to the source.

I don't see how the dirt circuit gets back to the inverter.
Is the return path through the OSB board and inverter chassis?

It does return to the source

The route is:

1. Inverter N terminal - Blue cable
2. RCD N terminal - Blue and PE cable
3. Ground rod - PE cable
4. Ground rod 2 - dirt/earth
5. Battery terminal - Ampere meter
6. Inverter battery terminal - Battery cable
7. Inverter L - non isolated switching inside the inverter

 

[John Frum]

1. Ground rod 2 - dirt/earth

How does it get from earth to the battery terminal?

1. Battery terminal - Ampere meter

 

[mrzed001]

How does it get from earth to the battery terminal?

Through the Ampere meter (only to measure the current flowing).
A simple cable would represent a people touching the battery terminal too

 

[RCinFLA]

The little cheap inverters have a 12v to HV DC converter. This converter has a feedback loop to battery side of transformer to regulate the HV DC. Most newer units use a opto-isolator to get the HV DC sensing back to battery side MOSFET switching controller to do the HV DC regulation. Older inverters, and perhaps some really poor quality cheap newer units do not use the opto-isolation between HV DC and battery side circuitry. These have a common ground between HV DC output and 12v input. This is bad news. The AC output must remain floating and neutral cannot be grounded to chassis.

The HV DC goes to an H-bridge MOSFET or IGBT chopper to create modified sinewave or PWM sinewave output so the actual HV DC is flipped in polarity at the line frequency rate on the two AC output pins. You need the opto-isolation in order to be able to ground neutral.

Here is an old schematic of a cheapo 120vac 300 watt modified sinewave inverter with no isolation between battery side and HV DC side. Common ground connections, all with discrete parts in the days before high-side MOSFET drivers IC's existed with no opto-isolated feedback control.

You cannot ground neutral on this type of inverter.

For most newer inverters there are EMI capacitors on both pins for AC output to chassis ground. This can create a little 'tingle' AC voltage on ground by the low value (0.1 uF) EMI capacitors. On newer inverters, these EMI caps can be enough to trip ground fault breakers.

This also applies to inverter-generators. You can place a high impedance AC voltmeter between either AC prong to its AC outlet ground pin and measure about half the AC output voltage. If optically isolated, most do allow the neutral AC pin to be ground to chassis ground of inverter-generator.

 

[Charleswgibbs]

I swear the more I read on grounding the more confused I get.

So the inverter is not meant to be grounded at all? So when you run the ground wire to the ac negative in the breaker that makes a connection between ac and ground that shouldn't be there?