Furnace on generator throwing error code

[Dave Millman]

I've read some previous threads on generator grounding:
https://terrylove.com/forums/index.php?threads/grounding-2000w-inverter-generator.42656/
https://terrylove.com/forums/index.php?threads/grounding-a-honda-eu3000is-generator.78435/

My situation is a bit different: We use a portable Honda inverter generator to power necessary appliances during outages:

• Refrigerator
• Furnace
• UPS
• Some LED lamps

There is no connection to the house wiring. Total load if everything is running at the same time, measured on a Fluke clamp-on ammeter, hovers around 60% of generator extended run capacity, so no overload issues.

This worked fine at our previous home through many outages. During the first outage here at our new home, the refrigerator, UPS and lamps worked fine on generator power. However, the York furnace shut down and threw an error code:

CODE: 9 RED FLASHES
RESULT: Indicates reversed line voltage polarity or grounding problem
FIX: Check polarity at furnace and branch. Check furnace grounding. Check that flame probe is not shorted to chassis.
​

I'm assuming that the furnace control board is detecting inadequate ground through the generator? The furnace is about 10 years old, or about 25 years newer than the furnace at the previous house. When the power was restored, I plugged the furnace back into the house wiring and it worked fine again, no error codes.

So in this specific case, can anybody recommend a way to run the furnace on the generator? Some of the possibilities I've thought of:

• Connect the generator ground lug to a ground stake
• Connect the generator ground lug to the house ground
• Ground the frame of the furnace to a cold water pipe

I have no desire to create a dangerous situation, so I'm asking the experts. Thank you!

 

[WorthFlorida]

Some generators have a ground connection on the chassis of the generator. You can try a few things and grounding is rarely dangerous. If there is a way to attach a ground wire to the chassis of the generator, connect the wire to it and the other end clamp it to a rod or large screwdriver. Insert the rod or screwdriver into the ground. Preferably the ground is damp. If the house ground rod is nearby, clamp it to it. Do not clamp it to a cold water pipe unless you have 100% copper pipe and the copper is bonded to the ground rod and you see the ground wire clamped on the pipe.

At my brothers home in VT, when the whole house built in Kohler generator transferred power, his Belarus Furnace would error out but no clue as why. After some trouble shooting and using a UPS battery backup, I found that the line voltage went to 142 volts instead of 125v. Replacing the voltage regulator fixed the problem. Since the furnace is probably hardwired to ground anyway, as you suggest, the voltage reference is off the mark and the furnace controller has a problem with it. Since ground wire hardly carries current, any wire will work to see if it takes care of the problem. If it does then you can set it up with a permanent fix.

 

[Sponsor]

 

[Fitter30]

Separately derive system needs a ground check with the generator manual.
https://generatoradvisor.com/ground-portable-generator/

 

[WorthFlorida]

Separately derive system needs a ground check with the generator manual.
https://generatoradvisor.com/ground-portable-generator/

You're right fitter30 but without a model number I didn't mention it. Here is page 18 of the Honda Inverter model EU2200i. https://images.thdstatic.com/catalog/pdfImages/06/06da532e-e2a5-429d-83f7-626a1d100d19.pdf

I'm having a hard time with the last sentence?/ Update; I read it the next day and now I understand what it is saying. If the device needs a ground to function, it wouldn't work without the Inverter generator chassis grounded.

 

[Jadnashua]

It may not be ground related at all!

An inverter generator normally does NOT produce a sinewave output. Worst case, it ends up a square wave, but most of them produce an approximation of a sinewave by doing a stepping function. A square wave produces HUGE amounts of noise on the power line. A stepped approximation of a sinewave still does, but not to the degree of a square wave. To get those nearly vertical changes of voltage, the frequency is essentially close to infinite, and can cause ringing in some circuits. IT can literally cause some devices to fail over time, if they work at all. Power inverter - Wikipedia Different generators may have more or less steps from the simple square wave to one with very small steps which gets it closer to the sinewave, which has NO steps, and thus, no ringing, other than at the primary frequency (50/60Hz for most things across the world).

 

[Fitter30]

Buy a receptacle tester you will find out generator has a floating neutral bad ground. Buy plug either 20 or 30 amp which ever is spare. Jumper wire from silver terminal to ground plug into spare outlet. Receptacle tester should read ok.
https://www.harborfreight.com/elect...MIrfyWkenm8wIVHXFvBB03nwrrEAQYASABEgIN6vD_BwE

 

[Taylorjm]

It may not be ground related at all!

An inverter generator normally does NOT produce a sinewave output. Worst case, it ends up a square wave, but most of them produce an approximation of a sinewave by doing a stepping function. A square wave produces HUGE amounts of noise on the power line. A stepped approximation of a sinewave still does, but not to the degree of a square wave. To get those nearly vertical changes of voltage, the frequency is essentially close to infinite, and can cause ringing in some circuits. IT can literally cause some devices to fail over time, if they work at all. Power inverter - Wikipedia Different generators may have more or less steps from the simple square wave to one with very small steps which gets it closer to the sinewave, which has NO steps, and thus, no ringing, other than at the primary frequency (50/60Hz for most things across the world).

I think he hit the nail on the head. I had the same issue when testing a backup generator hookup at our vacation home for the first time. Everything worked fine but I decided to try the central a/c since it's a small unit and should have worked. The air handler worked, but it would not switch the outside unit on. Everything had power, but the control board wouldn't let the outdoor unit switch on. From what I researched, newer furnace control boards don't like "dirty" power and won't run unless it's cleaned up first.

 

[WorthFlorida]

.........An inverter generator normally does NOT produce a sine wave output. .......

Dave Millman, I suggest based on Jadnashus explanation is borrow or rent a good old generator and give it a try.
I googled "honda output sine wave inverter generators" and all kinds of confusing statements are made based on inverters.

This site is good on explaining the inverter. I have a EU2200i but no osciliscope to look at the sine wave.
https://www.rvtravel.com/rv-electricity-contractor-generator-clean-ac/

 

[Jadnashua]

Some inverter generators can make lots of steps to make the output function closer to a sinewave, but the cheapest ones just make a simple square wave which will work for some things, but is terrible for others. One way to look at a square wave is that the vertical line is the combination of an infinite number of steps, so can generate all sorts of ringing noise on things that are expecting a nice, clean sine wave. A square wave is more like flipping a switch on/off (in the USA) at 60xper second...When does a light bulb fail most often, or anything electronic? When you turn it on, but with a simple inverter generator, you're doing that at 60Hz. rather than once when you turn the device it's powering once to turn it on. Something like an incandescent light bulb will continue to glow until the flow reverses, so you don't see any flicker, but motors and some power supplies don't like it at all, and can produce some nasty spikes when the power abruptly flips voltage from positive to negative. In a power supply, there will be a capacitor (generally) to smooth out the sinewave changes in voltage peaks...the same ones will not work with a square wave. SOme electronics will literally fry with a square wave input, some won't care at all.

 

[Dave Millman]

Thank you everyone for your replies! I've learned a lot more about generators in the last few days, especially from your posts but also YouTube (good and bad). Here's an update:

@WorthFlorida provided a manual for the EU2200i. The manual for my older Honda EU2000i generator is not nearly as informative:

Note that my manual does NOT say "FLOATING NEUTRAL" like the EU2200i manual. However, at the power outlet, the Honda EU2000i generator does have a label saying, "floating neutral."

@fitter30 recommended a receptacle tester. Here's what a common three-lamp test plug (yellow-yellow-red) indicated:

• Generator running: "open neutral"
• Generator not running: "open neutral"
• UPS running on battery during power outage, supply cord plugged in: "correct"
• UPS running on battery during power outage, supply cord NOT plugged in: "open neutral"
  
• All tested outlets in the house (control group, no connection to this problem): "correct"

The specific problem was the Furnace error code claimed "Indicates reversed line voltage polarity or grounding problem. So I got brave and bonded neutral to ground at the generator. The test plug indicated "correct" (two yellow lamps). The furnace error code disappeared, replaced by the "Normal operation with call for heat" code. We had heat for the outage the next day.

Finally a question: Have I created a risk by bonding neutral to ground at the generator? If so, what should I do? Thank you all!

 

[Jadnashua]

Your commercial power has ground and neutral connected together. That should be the only place where that happens in the system.

In reality, your commercial power is really 240. You do not have two separate 120v inputs...you get the output of the transformer, which has 240vac from one side to the other. Neutral is actually connected half way across the coil, so if you measure from the middle (neutral, to either end, it measures 120vac, but when one side is high, the other side is low, so you measure 240 relative to each other, and 120vac from neutral to the ends of the coil.

When measuring things, what the reference point is, is critical to understand. When I used to work on high voltage power supplies...the filament voltage was only 5vdc, but the plate on the thing was -5000vdc..they sat right near each other, with minimal gap, so if that 5vdc was measured to ground, there would have been a 4995v difference to the plate sitting right next to it and would have arced. But, one end of the filament was 5000v, while the other side was 4995, or a difference of 5vdc, and it worked just fine. It all depends on your point of reference...some devices have parts that operate at 240vac, and some (often a power supply or control board) may be built to use 120vac, so the proper reference point is critical. How noisy the power is can also have a big effect on things. Some stuff just doesn't like inverter generator outputs much, and a simple square wave version is the worst.

 

[Taylorjm]

Dave Millman, I suggest based on Jadnashus explanation is borrow or rent a good old generator and give it a try.
I googled "honda output sine wave inverter generators" and all kinds of confusing statements are made based on inverters.

This site is good on explaining the inverter. I have a EU2200i but no osciliscope to look at the sine wave.
https://www.rvtravel.com/rv-electricity-contractor-generator-clean-ac/

I used an old fashion generator but had the same issue with it being too "dirty" for the furnace control board. The other thing I read something about was someone using a ups before the furnace to clean up the power first. Can't confirm.

 

[Fitter30]

Call honda 888-946-6329

 

[Reach4]

Finally a question: Have I created a risk by bonding neutral to ground at the generator? If so, what should I do? Thank you all!

Have you created either a real or theoretical risk? Let me ask some questions:

1. How does the generator output get to the furnace?
2. Are you feeding a transfer switch? I think no.
3. Are you driving the house wiring through the breaker box with some other method? Or are you plugging the furnace and other generator-driven loads directly into an extension cord or cable from the generator?

 

[WorthFlorida]

I used an old fashioned generator but had the same issue with it being too "dirty" for the furnace control board. The other thing I read something about was someone using a ups before the furnace to clean up the power first. Can't confirm.

I used an APC UPS at my brothers house. I was assuming on the transfer the switch over was too slow for the furnace. Both my brother and I assumed the same since it would only error out during a power failure. With the furnace plugged into the UPS, the UPS was tripping out and seeing this failure before, I realized that it was a over voltage problem from the generator.

Dave, You may need to call York. I just hit me that while I was trying to figure out the problem with the Buderus Furnace, There were a few posts from other sites with this furnace not running on generators. I cannot find it now but Buderus issued a SB on this very problem. One of the controller boards had to be redesign and for some reason, being manufactured and designed in Europe, it did not like generators for the USA. This happened around around 2009-10 time frame. My brother's Buderus did have the improved controller. My brother even bought a new controller and it was the same version as what was in the furnace. It was around $500, because he bought it through an online source and he was able to return it.

 

[wwhitney]

Finally a question: Have I created a risk by bonding neutral to ground at the generator? If so, what should I do? Thank you all!

Looks like your generator has just (2) NEMA 5-20 receptacles for AC output, and the manual specifies that the receptacle EGC is connected to the generator chassis but not to the receptacle "neutral". I assume your furnace is cord and plug connected, and you are just unplugging the furnace from the building wiring and plugging it into the generator. I.e. none of the fixed building wiring is part of the circuit when running on generator.

If so, equipment that requires a neutral-EGC bond will not function on the generator without your adding an explicit neutral-EGC bond, as you have discovered. Just adding that bond will still leave the generator ungrounded (unearthed), and will connect the chassis of the generator and the furnace to one side of the 120V output of the generator. This is not a configuration I have contemplated much, but I think the risk of shock goes up somewhat. I.e. without the neutral-EGC bond, getting shocked would require you to get across generator hot and neutral. With the bond, it would be enough to get across generator hot and the chassis of either the generator or the furnace.

The generator is a power limited source, so I assume it has no overcurrent protection? And it doesn't look like the receptacle has GFCI protection, although for a 2-wire circuit absent other return paths that wouldn't do anything.

I don't think I'm in a position to say that the somewhat elevated risk of shock from the neutral-EGC bond is or isn't acceptable, but it's not crazy. If your gas furnace will run on a GFCI receptacle without tripping it, then the following would be somewhat safer: plug a neutral/EGC bonding plug into one receptacle, and plug a portable GFCI device into the other, and connect the furnace to that portable GFCI device.

That way if the EGC becomes part of a return path that exceeds 6ma, the GFCI should trip. So if you get across the hot downstream of the GFCI and the chassis of either the furnace or the generator, the shock should be limited by the GFCI to a safer level/duration.

Cheers, Wayne

 

[Dave Millman]

Have you created either a real or theoretical risk? Let me ask some questions:

1. How does the generator output get to the furnace?
2. Are you feeding a transfer switch? I think no.
3. Are you driving the house wiring through the breaker box with some other method? Or are you plugging the furnace and other generator-driven loads directly into an extension cord or cable from the generator?

The furnace is plugged into the generator using a 12-ga extension cord. No transfer switch, no connection to home electrical panel. When the power goes out, we play "extension cord musical chairs."

 

[Reach4]

The furnace is plugged into the generator using a 12-ga extension cord. No transfer switch, no connection to home electrical panel. It's "extension cord musical chairs" when the power goes out.

So in that case, connecting ground and neutral together at the generator is right I think. That extra plug to do that temporarily seems like a great idea, but you give up an outlet.

 

[Dave Millman]

If your gas furnace will run on a GFCI receptacle without tripping it, then the following would be somewhat safer: plug a neutral/EGC bonding plug into one receptacle, and plug a portable GFCI device into the other, and connect the furnace to that portable GFCI device.

That way if the EGC becomes part of a return path that exceeds 6ma, the GFCI should trip. So if you get across the hot downstream of the GFCI and the chassis of either the furnace or the generator, the shock should be limited by the GFCI to a safer level/duration.

That's a great idea, and a very educational post. Thanks Wayne!

 

[wwhitney]

So in that case, connecting ground and neutral together at the generator is right I think. That extra plug to do that temporarily seems like a great idea, but you give up an outlet.

The only place the "neutral" from the generator should be exposed is at the receptacle pins, so you have to use a device that plugs into it. If necessary, you could use a device that has a receptacle on it, too.

Cheers, Wayne