Inverter Neutral to Ground Bonding

[FinEng]

I recently purchased a cheap inverter to use as a backup power source for my gas furnace in the event of a power outage. (This is in the US).

The inverter is 1500W continuous (3000 watts peak), which is adequate for my needs. 12V DC input, 110V Ac/60 Hz output. It is supposed to be a pure sine wave inverter, although I have no means to verify that.

The product details can be seen here: https://www.aliexpress.us/item/2255...t_main.21.8d951802ybxCXB&gatewayAdapt=glo2usa

My question is about bonding ground to neutral. I will connect the furnace to the inverter output using an extension cord, as such it will be a “separately derived system” and the neutral needs to be bonded to the ground terminal for safety. Also, the furnace will not operate if it does not see ground (it is a high efficiency furnace and as I understand it they need to see ground, presumably to ensure the flame sensor operates properly).

The inverter, while it appears to work well, unfortunately comes with no documentation. The seller (manufacturer?) is Chinese and of not much help, possibly because of language difficulties. The only thing they can tell me is that both AC output terminals are live. I have verified by using a multimeter that both are live and floating, with no reference to ground. I will not be connecting any part of the inverter to my house AC system.

Measurements were as follows: NCV feature on the multimeter does not detect any live voltage, however the “Live” setting (single probe from the multimeter) indicates both terminals are live. 115V AC measured between the 2 terminals, about 49V AC measured between each terminal and metallic case (“ground”), but no shock if I touch the case. Voltage between ground terminal of the AC outlet and house system ground shows indeterminate (fluctuating 20-30V AC) voltage. The outlet ground terminal (socket) does not appear to be connected to anything, not even to the case. Voltage between the case and the ground terminal is about 5V AC.

However, if I use a Fluke multimeter on the LowZ setting, then I measure 0.6 volts AC between either of the 2 live terminals and the case, and 0V between either of the 2 live terminals and my house electrical system ground.

This leads me to believe that the voltage being measured between live and case in the non-LowZ mode is a phantom capacitive coupling, would that be right?

The website (listing) has the following info:

Both of the inverter outputs are live wires, and there is no grounding problem. Do not use any one as a ground wire. Connect the inverter to the load. Do not connect any air switch or leakage protection switch in the middle. Hard operation will not only fail to protect, but will damage the inverter.

I have connected a GFCI/RCD between the outlet and a test load and it seems to work fine, not sure why they say leakage protection (which I assume is a GFCI/RCD) will damage the inverter. Any thoughts?

But my main question is will it be safe to bond the one of the outputs (any one I suppose) to the case and designate that output as the neutral, and bond the case to the outlet ground socket. I assume this will not electrify the case (cause a shock) since there is no reference to earth but I am concerned that it might damage the inverter, especially since I am currently (unbonded) measuring about 49 V between either of the 2 live outputs and the case. I think I have read somewhere that the possibility of inverter damage exists, but not sure. The seller also said: Do not use any one (output terminal) as a ground wire

On the other hand, if it is just a ghost voltage then I think I should be safe to bond one of the output terminals to the case and designate that as neutral/ground.

Any recommendations on that would be very welcome. If there is anything I can check by opening up the inverter, please do suggest.

If anyone has a typical schematic for such an inverter I would really appreciate that too.

Thanks in advance.

 

[JayArr]

I wouldn't bond it to the case of the inverter but you may be able to bond it to the ground of the furnace. This would be at your risk and if it blows the circuit board in your furnace the "deal" you got on from the Chinese may not look so good.

A couple of questions need answering. Will the charger that maintins the battery be left connected full time? How is it bonded, does it bond the negative battery terminal to the house ground?

It looks like it's for camping or off grid temp use, not for backup.

Is there any UL sticker on this thing?

If not can you return it and buy a better unit that has UL approval and proper grounding?

 

[rodrick]

before putting to much time into it have you tested it to see if it will run the furnace

 

[FinEng]

I wouldn't bond it to the case of the inverter but you may be able to bond it to the ground of the furnace. This would be at your risk and if it blows the circuit board in your furnace the "deal" you got on from the Chinese may not look so good.

A couple of questions need answering. Will the charger that maintins the battery be left connected full time? How is it bonded, does it bond the negative battery terminal to the house ground?

It looks like it's for camping or off grid temp use, not for backup.

Is there any UL sticker on this thing?

If not can you return it and buy a better unit that has UL approval and proper grounding?

Thanks for the super fast response!. To answer your questions:
The charger is connected full time, it does not bond the battery negative to house ground.
No UL, CSA, or CE markings
Too late to return.

You say you wouldn't bond the "neutral" to the case of the inverter, can you please explain your reasoning? On the other hand if I bond the neutral to the furnace ground, then won't that also bond it to the inverter case, since the ground wire in the furnace power cord will effectively connect the furnace ground to the inverter ground (case).

Also, can you please explain why you thin there is some danger of blowing up the furnace control board?

Thanks!

 

[FinEng]

before putting to much time into it have you tested it to see if it will run the furnace

I do plan to do that, but before I did I wanted to do some diligence to ensure I don't blow up something -- the inverter, my furnace, myself!

 

[dcg9381]

This is the "main" for all intents and purposes and it's not grid tied, so you need to treat it like a main. IMHO, it gets a neutral/ground bond. Not UL, there is some risk here. Is there currently continuity between the N/G output legs?

 

[FinEng]

This is the "main" for all intents and purposes and it's not grid tied, so you need to treat it like a main. IMHO, it gets a neutral/ground bond. Not UL, there is some risk here. Is there currently continuity between the N/G output legs?

Yes, I agree the neutral ground bond is needed, and no currently there is no bond/continuity between N and G. My question is not whether I should bond, but rather will it create some problems based on what the seller says, and the voltages I have measured.

 

[dcg9381]

Yes, I agree the neutral ground bond is needed, and no currently there is no bond/continuity between N and G. My question is not whether I should bond, but rather will it create some problems based on what the seller says, and the voltages I have measured.

I think you mean this statement:

Both of the inverter outputs are live wires, and there is no grounding problem. Do not use any one as a ground wire. Connect the inverter to the load. Do not connect any air switch or leakage protection switch in the middle. Hard operation will not only fail to protect, but will damage the inverter.

It's unclear to me what this means. I interpret it as you do not ground the inverter case to earth ground. But that could be wrong.

I'm not qualified to guess at your measurements.

What I do know is that if you do not have a neutral/ground bond, some "sensitive" equipment may not work.

 

[Texas-Mark]

What I do know is that if you do not have a neutral/ground bond, some "sensitive" equipment may not work.

The OPer mentioned a furnace. I recall at least one thread where the lack of bonding would not let the furnace run.

 

[dcg9381]

The OPer mentioned a furnace. I recall at least one thread where the lack of bonding would not let the furnace run.

I've run into the same thing on an EVSE (EV charger).

 

[JayArr]

Good UL listed inverter/chargers switch the ground when they change from utility power to inverter power. There is an extra set of contacts in the relay to do this. When on utility power the ground for the load is connected to the ground wire of the utility feed and not internally to the chassis. When they switch to inverter mode the load ground wire is disconnected from the utility wire ground and is bonded to the neutral wire inside the inverter. You are missing this circuit from your setup. The fact that the manual warns you not to bond either of the output lines to ground/chassis is supicious, like they cut some corner to make the inverter less costly to manufacture. The fact that is has no UL sticker furthers my suspiciaon that you've bought sub-standard gear.

The safe solution is to sell the Chinese inverter on eBay to someone who needs it for camping or other two pronged/no ground applications and buy a proper inverter/charger with a built in transfer switch.

Here's an idea that may help you decide... go get the part number of the control board for your furnace and look up how much it costs for a replacement. That's going to be the price of getting this wrong.

 

[FinEng]

@JayArr

I completely agree with what you say, which is why I am not rushing to connect my furnace to this inverter. Rather, I want to proceed cautiously, test it thoroughly, and then decide whether I should use it for this purpose.

To that end, please help me understand what issues might it cause -- for example spikes, over voltage, ground fault, etc? Or something else?

So far I have connected the neutral to ground (to the inverter case, not to earth ground) and tested on a lamp as the load. It did not short out the inverter and the lamp works. I do not get any high voltages on the case. I have also tested for short circuit, the inbuilt protection works well. I will now be testing for ground fault protection. I will also be connecting the case to earth ground (the ground in the house electrical system) and checking for any issues after doing that.

Your input would help in deciding what other tests I should make. Thanks.

 

[dcg9381]

To that end, please help me understand what issues might it cause -- for example spikes, over voltage, ground fault, etc? Or something else?

As I understand it, it can break the functionality of required GFI sensitive circuits. In my case, doing it wrong was detected by our EV charger and it would fault at start up.

 

[FinEng]

As I understand it, it can break the functionality of required GFI sensitive circuits. In my case, doing it wrong was detected by our EV charger and it would fault at start up.

Can you help clarify this -- what specifically is the "it" that breaks the GFCI? And would you know why?

Thanks.

 

[dcg9381]

Can you help clarify this -- what specifically is the "it" that breaks the GFCI? And would you know why?

Thanks.

I believe it works by comparing the current along conductors. Apparently it will still work if there is a "bad ground".
I don't know what why our EVSE was failing - all I can tell you is that their engineers gave me details that the EVSE was detecting "bad ground". It may have been checking for continuity between N/G.. not 100% sure.

 

[FinEng]

Yes, you're right -- the GFCI compares the current between live and neutral, and proper bonding of neutral to ground and proper grounding are essential.

But, I'm still not understanding why I'm being cautioned about using this setup with my furnace if all the bonding and grounding is done properly. If there are other possible issues I do want to understand and test for them.

Thanks.

 

[dcg9381]

But, I'm still not understanding why I'm being cautioned about using this setup with my furnace if all the bonding and grounding is done properly. If there are other possible issues I do want to understand and test for them.

I think you're being cautioned that some "sensitive" controls (perhaps like the furnace) will not be happy if you don't bond properly. No one is indicating a problem if bonded in the correct manner. You're being cautioned because your device is from China, not UL listed and there seems like there are limited ways to bond it.

 

[JayArr]

But, I'm still not understanding why I'm being cautioned about using this setup with my furnace if all the bonding and grounding is done properly.

That's the thing, there is no way to properly do the bonding.

Bonding occurs at the source, not the load. Your house, for example, had the neutral and ground bonded at the main fuse box and then subsequent sub panels are not bonded (by code). Outlet boxes do not bond N and G and loads/appliances do not bond N and G. When you bond neutral to ground at a load or downline from the main bonding point this creates ground loops and is against code.

The inverter you want to use specifically tells you not to bond either of the outputs to ground and as such cannot be "properly" used as a source in a house, it's fine for running stuff while camping or at the beach. That's likely why it was so cheap.

It sounds like you don't have enough money to do this properly so you keep asking new questions hoping for a solution that doesn't require you to spend more but the simple truth is that the source neutral needs to be bonded to ground and the cheap Chinese inverter you bought with no UL approval specifically tells you not to do this.

You've made a mistake and bought the wrong inverter, trying to make it work anyways could just compound the mistake from one to two or more mistakes. It sucks when you make a mistake that costs you money but the smart thing to do now is sell this inverter and get one with a UL approval that alllows N to G bonding.

 

[FinEng]

Got it -- thanks!

 

[garetwo]

Good UL listed inverter/chargers switch the ground when they change from utility power to inverter power.

This is good to know. Alleviates bonding neutral to ground multiple times when the utility is introduced. Have you seen inverters that do not do this? Should we (as an electrician) introduce some means of bonding when no utility and separating when utility is introduced?