Generator/dc grounding issues

icarusicarus Solar Expert Posts: 5,436 ✭✭✭✭
My neighbor and I have been struggling with a problem. The system consists of the panels, a xantrex sw inverter, iota charger. It powers a year 'round building in Northern ON. Canada. The back up generator is a 120vac Mitsubishi mge2900. This generator is a bonded neutral (rather than floating)

The genny sits in a shed, is bonded to a ground rod at the shed. It feeds the building with piece of UF wire. When it enters the building, it comes to a transfer switch and breaker box. These devices are NOT grounded except through the ground wire in the UF cable. It then feeds the outlets of the house. (The neutral is not bonded at the building)

The pv array comes into the inverter, then to the said same transfer switch and feeds the outlets. The Iota charger, the battery bank and the inverter are bonded together and share a common ground wire and ground rod. (I hope I got that all correct!)

The problem is this. 4 times in 5 years, the generator voltage regulator control board has failed. It is a printed circuit with a couple of transisters and a heat sink on it. The local (300kms away) Mitsu dealer said, "Oh you need a control board!" and put one in. The next time, the owner replaced it himself for a savings, but still real money, ~$350. Time number 3 and 4 same thing.

I got ahold of Mitsubishi's tech rep and here is what he said. "Anytime an AC ground is shared with a DC ground you provide a direct path back to the source..ie the generator. AC current is disrupted by DC current.
The AC-DC grounds must be isolated from one another to ensure no damage to the source will occur.
I would insure all AC grounds are isolated from DC grounds."

My question is two fold. First, is this a reasonable answer? (I am not enough of an electical expert to know. In my mind ground is ground, but maybe not.) The second, for my own self interest, I am building a very similar system, except mine is powered with a Honda Eu 1000 genny. I want to avoid similar problems with my system, as well as help my neighbor out. (We are in a very remote location, two miles apart on seperate islands, 50kms from the nearest neighbors. With -40 temps in the winter, failure is a problematic experiance)!

After a few e-mails back and forth when I complained that the original dealer should have looked harder at the problem, the Mitsu rep reluctantly agreed to send me one more control board gratis. (To his credit!) And that it doesn't state anywhere in the manual that the generator should not share a ground with a DC system. He also suggested that plugging a charger directly into the AC outlets of the generator could cause a similar problem, due to harmonic distortion? (Once again, no warnings in the manual!)

I know that inverters with floating neutrals can (and do) burn up when connected to building wiring that has bonded neutrals. The final question is, when I wire my own place, should I bond the neutral at the panel, to a ground rod, and should I bond all the DC stuff to it's own ground rod. Or should I rely on the generator ground to ground the building?

Once again, thanks for all the experise!

Icarus

Comments

  • BB.BB. Super Moderators, Administrators Posts: 32,611 admin
    Re: Generator/dc grounding issues

    It is a tough call--many times because the actual bonding/requirements are not clear between all of the various pieces of equipment. So--you are left either trying something (and replacing scorched parts), over designing the system, or taking things about and trying to see what is actually happening.

    I am not the expert here--but from what I have read and seen.... Modified square wave inverters normally cannot have a bonded neutral and a grounded (+ or -) DC input because there is no electrical isolation on the AC output stage (would require an extra transformer and more money).

    Generally, a full sine wave inverter can support a grounded neutral because the output Hot/Neutrals are floating with respect to ground.

    Generator with Inverter output--I have read that folks with a eu2000i or a eu3000i can support bonded neutral applications--but it is not clear in the eu2000i manual if this is supported or not.

    Generators with normal alternator outputs should be able to support a grounded neutral output.

    To, possibly, fix your first problem, you could put an isolation transformer on the generator output and support the local earthed neutral, and the remote earthed neutral at the building. Look at something like the PSX-240 isolation transformer from Outback.

    Generally, where one gets into problems between AC and DC grounding systems is that when because of code, or other issues, that you end up with multiple grounds (one at the generator, another at the building, and incidental grounds like the earthed neutral, grounded cases on 12 VDC appliances, etc.).

    The issue, that I see, with multiple DC grounds (non-capacitive, metallic connections, grounding) is the accidental (for example) grounding tied between a DC chassis ground and an AC neutral bonded ground. And what can happen is (for example) 1/2 the DC current flows through the - DC harness, and the other 1/2 of the current flows through the white bonded neutrals and/or green safety grounds.

    And DC circuits, typically are an order of magnitude (or larger) that the AC Hot/Neutral/Ground... A 120 VAC circuit is typically 15-20 amps rated. A typical DC circuit (for an equivalent device) would be 150-200 amps (at 12 VDC)--So if there is an accidental alternative path through the AC safety ground for the DC load currents--one can easily blow the AC safety ground circuits and their related components.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • icarusicarus Solar Expert Posts: 5,436 ✭✭✭✭
    Re: Generator/dc grounding issues

    Bill,

    Thanks for the useful info. I suspect I should contact to see what they say about bonded neutral with the eu1000. The only reference in the "manual" says that any connection to a building should be done "by a qualified person"!

    As to the transformer,,,can you describe how you would wire it please? Also do I eliminate the problem just by using seperate AC and DC ground wires and rods?

    Thank

    Tony
  • BB.BB. Super Moderators, Administrators Posts: 32,611 admin
    Re: Generator/dc grounding issues

    I did a quick read-thru on the Outback PSX-240 transformer manuals (scroll down to find two -240 manuals here)--and this transformer cannot be used as a general purpose 120/240 VAC isolation transformer... It has a lot of neat uses as an auto-transformer--but general isolation is not one of them.

    Here is an example of (what appears to be) a very good isolation transformer that is designed for shore power to boat connections (PDF File) (isolation of shore ground/line/neutral from boat ground/line/neutral).

    I have no idea about this company or their product--but their installation manual does a pretty good job of describing the various methods of connections and the reasons why for each.

    These isolation transformers are not small (this 30/16 amp 120/240 VAC version is roughly a 10" cube and weighs ~60lbs) and probably not cheap. You might find some that are more applicable (and less money) to a home/business at your local electrical contractor supply house.

    Using two ground rods (one for AC the other for DC)--in my mind is a "cheat". You could just end up with two energized ground rods (could still get a shock if both grounding systems were touched/shorted together accidentally or on purpose).

    You are better off to make sure that you have no unintended ground loops and that all ground runs are "home run" back to a single point. What you don't want is an AC/DC return connection (- battery or white/neutral) that have two (or more ways) back to the power source (the second way being through the safety ground or alternate ground connection... i.e.; some DC - current flowing through AC white or AC green wire, or AC neutral going through the DC - or AC green/safety ground wire)... The only time you would have "cross circuit DC or AC" current flow is if there is an accidental short (and the AC safety ground it to trip the breaker--there is really no DC safety ground in most systems--like grounded metal automobile chassis serves both as the "-" lead and safety ground).

    I am sorry I don't have better answers... The little Honda's (eu1000i and eu2000i) are not large and will not smoke safety grounds like a big short to your battery would (and, frankly they are not that expensive to replace if something goes terribly wrong)--so I would tend to, in your case, pick one place to put the ground rod (probably at your distribution panel--just like any normal home A/C wiring) and carry the three wires back to the generator (Hot/Neutral/Safety-Ground) and plug it in there. And don't bother to use the ground lug on the generator chassis.

    I would guess that it is possible that a battery charger could have grounded their DC "-" output to the Green Wire (or Neutral which is grounded back at the panel). You could confirm that with any ohm meter (Look for anything less than ~1,000 ohms as a real potential problem--Should be >100,000 ohms resistance between output and any AC line plug connection).

    I can't imagine any decent model of a modern battery charger that would ground reference their DC output to the AC line cord (anything is possible--so I can't say it would never ever happen). For example, charging an old positive ground automobile would give some very fine fireworks with a negative grounded battery charger... :p

    The other issue is battery chargers can use a Diode Bridge rectifier that would take the majority of the current near the peak of voltage sine wave--this wave form can over heat transformers, generator coils, and electrical wiring because of these very high (and short) peak currents. This is what your generator manufacturer was warning you about. If your battery charger is 1/10 of the generator capacity, then it will probably never be a problem. If your charger is 50%+ of your generator capacity--then it may be an issue with the charger.

    Basically, any good brand of battery charger (like Iota), is going to be a power factor corrected switch mode power supply--the old days of a transformer and full wave rectifier is long gone (I looked through the Iota website--but I did not find their Power Factor rating--but they need to meet the new FCC/Utility specifications).

    I am sorry I can't really do more--but over the web is not the easiest to diagnose strange failures...

    One thing you can do is break the ground wire to the generator and put (for example) a 100 watt 120 VAC filament light bulb in series and measure the voltage drop. If you get any significant voltage across the light bulb (measure both AC and DC--they should both be near zero volts or at least below 2-5 volts--by the way, different meters will measure AC+DC wave forms differently--so I cannot be too exact here) then you may have some unwanted AC or DC connections to ground through the generator ground.

    You do have to be kind of careful with DC batteries, battery chargers, and grounding... As an example, I had a large telecom system that was powered by -48 VDC -- but we were reading +48 VDC--turned out that the -48 VDC power supply/battery charger output was not ground referenced--but when the customer accidentally touched a wire (or tool) to the +48 VDC lead (which was supposed to be ground) he "saw the sun" in front of him. Turned out the DC power supply has some pretty big DC caps referenced to AC ground--enough storage that they supplied a good sized arc to a short to ground on what was a floating supply--shorting a by-pass capacitor used to keep the output voltage smooth and stable.

    The normal trick to check ground connections was to use a 120 VAC filament light bulb (~40-60 watt) to see if something "floating" or suppling a voltage with lots of current behind it (in the above case, the light would have quickly and safely discharged the output capacitors when the bulb was connected from the positive lead to ground).

    I probably have typed too much--you can ignore the stories and why too detailed discussions above if they (or my typing) are confusing.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • icarusicarus Solar Expert Posts: 5,436 ✭✭✭✭
    Re: Generator/dc grounding issues

    Bill,

    Thanks for all your valuable information! You have given me lots to chew on. Sometimes simple things can get quite complicated! I'll give it some thought and get back. (I'm away from the site for a couple of months, so it will be a while before I can make any changes. I'm trying to get my ducks in a row now.

    Tony
  • RedfordRedford Solar Expert Posts: 38 ✭✭
    Re: Generator/dc grounding issues
    BB. wrote: »
    Using two ground rods (one for AC the other for DC)--in my mind is a "cheat". You could just end up with two energized ground rods (could still get a shock if both grounding systems were touched/shorted together accidentally or on purpose).
    (clip)
    --so I would tend to, in your case, pick one place to put the ground rod (probably at your distribution panel--just like any normal home A/C wiring) and carry the three wires back to the generator (Hot/Neutral/Safety-Ground) and plug it in there. And don't bother to use the ground lug on the generator chassis.

    I have my system set up like that, except I have an EU3000is which has a 125v 30A dedicated outlet. I acquired a plug for it and ran 12 gauge (3 strand) wire to the D box hooked up for hot/neutral/safety ground. Then I ran 35 feet of 4 gauge braided copper wire on which I have 2, 8' grounding rods hooked to at about 10 feet apart.

    I asked my friend if I should also ground my generator to this grounding system (of which is currently separate from the DC circuits in my house) and he said is wouldn't hurt.

    This winter when I expand my current system I was going to deal with this topic and was unsure of just how to do it....so many thanks! I'm sure I'll have many questions in the future!
  • icarusicarus Solar Expert Posts: 5,436 ✭✭✭✭
    Re: Generator/dc grounding issues

    I'm not clear. You say you have your system set up "like that too" Does that mean one ground rod or two? Do you have a pv array and or charge set up as well?

    Icarus
  • RedfordRedford Solar Expert Posts: 38 ✭✭
    Re: Generator/dc grounding issues

    I have an array and a battery bank but right now the only charger I have is a micro-processor controlled deep-cycle marine charger which plugs into a 120v socket. (Of which I'm going to replace with an MX-60 and an inverter this winter) For my array, I have a small 10 amp (generic) charge controller. I have not grounded any of my DC loads as of now because they only run two small lights and a 300 watt inverter via 12vdc power.

    So, when the sun is out the panels charge the batteries and when the genny is running the automotive charger charges the batteries at a rate of 2 or 15 amps.

    The Eu3000 is indicated as having a floating neutral so I ran a dedicated line to my distribution box. The D-box is grounded via a grounding field. I figured that it was fine to carry the generator neutral to the Distribution box and provide neutral to ground bonding there.

    I think there is an OSHA regulation that some industrial generators be neutral bonded to chassis, at least that's what my Honda dealer told me. In this case I was told that I would have to switch the neutral conductor (upon entry to main box) to prevent the neutral from being bonded to ground in two locations.

    Anyhow, I was trying to figure out the same DC/AC grounding question when I was "mentally" wiring my system upgrades for this winter, and came upon this thread which made a lot of sense to me.
  • RedfordRedford Solar Expert Posts: 38 ✭✭
    Re: Generator/dc grounding issues

    Last night I got to thinking about this topic and read up on it a bit. In the manual that came with my generator it never actually states that it has a "floating neutral". Nor could I find a panel on the generator that states so....(but it was real dark when I got home last night)

    The manual states that:
    Honda portable generators have a system ground that connects generator frame components to the ground terminals in the AC output receptacles. The system ground is not connected to the AC neutral wire. If the generator is tested by a receptacle tester, it will not show the same ground circuit condition as for a home receptacle.

    That's verbatim from the owners manual. I'm definitely not a certified electrician but I would take that to mean that the AC neutral is floating as it is not connected to ground.

    Originally I was going to "cheat" and hook up a separate ground field to my DC loads so as to avoid any conflicts with the AC ground. I think maybe I just have to read up a bit more on this topic before I go ahead with my expansion.
Sign In or Register to comment.