Lights in the house flicker/strobe/sometimes don't work when shop mini split is on

[jmillican1983]

My shop uses a sub panel fed from a 100a breaker in my main panel in the house. I have a 3 head 36k btu Lennox mini split system in the shop. When the mini split is turned on in the shop, two separate lighting circuits in the house have issues. They're canless LED retrofit lights that screw into the sockets inside each ceiling can. I have these lights on 4 circuits throughout the house, but only two circuits are affected when the shop mini splits are on. The some of the lights on each circuit will exhibit various symptoms ranging from not lighting at all, flickering/strobing, delayed illumination. If I turn the light switch off and back on, sometimes the lights that were having issues start working fine and different lights on the same circuit exhibit the above issues. The problem goes away when the mini split is turned off. I replaced all of the lights on one of the affected lighting circuits (with the same type), without success.

The wiring in the main panel for these affected lighting circuits (coming off the breaker) is near the wiring for the shop sub panel (4ga copper). Is it possible that there is some EMI being emitted by the Mini split, affecting the frequency of the voltage in the lighting circuits and upsetting the led drivers?

I tried a regular LED bulb in place of one of the canless led lights just to see how it would work and there were no issues. I tested the canless lights on a different circuit by screwing them into my bathroom vanity light. no problems there. I did check the voltage at a few different light sockets on the affected circuits and I have 121 volts between hot and neutral. I also checked to make sure each of the hot wires coming into my main panel had the same voltage and they do. 122v each. I checked the torque of the screw on each breaker and all the neutral screws. I inspected each light switch for loose screws and loose neutral wires at the wire nuts in the junction boxes. No problems were found.

Is it possible that the mini split is creating interference and affecting those lighting circuits due to the proximity of the wiring in my main panel? If so, what would be the best way to correct that?

 

[infinkc]

how is your ground and neutral connected in the subpanel? how long of run is it?

definitely sounds like feedback to me

 

[jmillican1983]

how is your ground and neutral connected in the subpanel? how long of run is it?

definitely sounds like feedback to me

The ground and neutral are bonded in the sub panel. 3 wire sub panel feed with ufer ground in the shop. Code acceptable where I live. It’s about a 230ft run between the 2 panels.


Sent from my iPhone using Garage Journal

 

[cybrdyke]

On the EMI speculation....no.
Did you share neutral wires between these circuits?
What controls are on the lighting circuits?

CD

 

[jmillican1983]

On the EMI speculation....no.
Did you share neutral wires between these circuits?
What controls are on the lighting circuits?

CD

I’m not sure if the two lighting circuits share a neutral. I will check on that. One circuit uses a standard switch, the other runs on three 4-way switches. No dimmers involved.


Sent from my iPhone using Garage Journal

 

[wyliesdiesels]

....

 

[wyliesdiesels]

My shop uses a sub panel fed from a 100a breaker in my main panel in the house. I have a 3 head 36k btu Lennox mini split system in the shop. When the mini split is turned on in the shop, two separate lighting circuits in the house have issues. They're canless LED retrofit lights that screw into the sockets inside each ceiling can. I have these lights on 4 circuits throughout the house, but only two circuits are affected when the shop mini splits are on. The some of the lights on each circuit will exhibit various symptoms ranging from not lighting at all, flickering/strobing, delayed illumination. If I turn the light switch off and back on, sometimes the lights that were having issues start working fine and different lights on the same circuit exhibit the above issues. The problem goes away when the mini split is turned off. I replaced all of the lights on one of the affected lighting circuits (with the same type), without success.

The wiring in the main panel for these affected lighting circuits (coming off the breaker) is near the wiring for the shop sub panel (4ga copper). Is it possible that there is some EMI being emitted by the Mini split, affecting the frequency of the voltage in the lighting circuits and upsetting the led drivers?

I tried a regular LED bulb in place of one of the canless led lights just to see how it would work and there were no issues. I tested the canless lights on a different circuit by screwing them into my bathroom vanity light. no problems there. I did check the voltage at a few different light sockets on the affected circuits and I have 121 volts between hot and neutral. I also checked to make sure each of the hot wires coming into my main panel had the same voltage and they do. 122v each. I checked the torque of the screw on each breaker and all the neutral screws. I inspected each light switch for loose screws and loose neutral wires at the wire nuts in the junction boxes. No problems were found.

Is it possible that the mini split is creating interference and affecting those lighting circuits due to the proximity of the wiring in my main panel? If so, what would be the best way to correct that?

ok who installed this feeder? #4 cu is not rated for 100a and 230' is too long even if it was.

You need to put a smaller breaker in the main panel

The ground and neutral are bonded in the sub panel. 3 wire sub panel feed with ufer ground in the shop. Code acceptable where I live. It’s about a 230ft run between the 2 panels.


Sent from my iPhone using Garage Journal

When was the feeder installed? if after the 2008 code cycle then it is actually NOT code permissible. the 2008 code cycle prohibited 3-wire feeds to detached structures.... and the grounding electrode/UFER does nothing to change this....

the 230' is too long for #4 cu even if the breaker was lower

 

[nadogail]

I have found that LED lights are very sensitive to fluctuations in voltage from motor startups.

 

[jmillican1983]

ok who installed this feeder? #4 cu is not rated for 100a and 230' is too long even if it was.

You need to put a smaller breaker in the main panel



When was the feeder installed? if after the 2008 code cycle then it is actually NOT code permissible. the 2008 code cycle prohibited 3-wire feeds to detached structures.... and the grounding electrode/UFER does nothing to change this....

the 230' is too long for #4 cu even if the breaker was lower

Current Oregon Amended code allows a 3 wire feed to a detached structure with a UFER ground.

I switched to a 90A breaker but the lights in the house on the same two circuits are still acting up when the shop mini split is turned on.

 

[wyliesdiesels]

which code is that? a local AHJ cant take away prohibitions in the NEC, they can only add to them.

BTW theres a real good reason it was prohibited. It creates a potential for shock

 

[kinglake]

I don't know why you think the NEC is the bible and can't be edited, but many states do so. The NEC has no enforcement or regulatory powers over states or jurisdictions. Oregon has their own electrical code, and while heavily based on the NEC, there are significant edits if you will. Entire sections are removed, others sections edited and exceptions added to yet others. They have explicit exceptions added to allow 3 wire feeders in certain scenarios.

 

[walrus]

I switched to a 90A breaker but the lights in the house on the same two circuits are still acting up when the shop mini split is turned on.

A smaller breaker isn't going to do anything for the flickering

 

[mcbane]

I have seen the same issue running cheap LED bulbs when the house is on generator power. Your LED can retrofits are cheap China junk. Just replace them with higher quality parts. Nothing to do with breaker size.

 

[jmillican1983]

A smaller breaker isn't going to do anything for the flickering

I was being sarcastic about switching the breaker. I'm going to try higher quality lights next. Thank you for the input.

 

[Terry D]

which code is that? a local AHJ cant take away prohibitions in the NEC, they can only add to them.

BTW theres a real good reason it was prohibited. It creates a potential for shock

Actually they can. At least here they do. St Louis county is in the 2014 code, they only require us to AFCI protect the bedroom circuits, not the rest. We can also use that #4 copper for a 100 amp to that outbuilding, but I agree, not at that distance

 

[infinkc]

The ground and neutral are bonded in the sub panel. 3 wire sub panel feed with ufer ground in the shop. Code acceptable where I live. It’s about a 230ft run between the 2 panels.


Sent from my iPhone using Garage Journal

Others may chime in, but i dont believe that you should have the Ground and Neutral bonded in the sub panel. Both the ground should be connected from the main and the Ufer on the shop, but neutral should be isolated.

 

[Busted_Knuckles]

Once all the sub-panel connections, grounds, neutrals, when the all connections are checked and whatever, take a look at your main breaker/main breakers, or any breaker down the line that you have issues with that circuit.

Ive seen this once, where a 100a,... I cant even recall... some super deep in weeds panel I had never seen before ( very old as well ), the breaker failed internally as far as the contacts are concerned. Maybe it was ITT or one of its affiliates ( foreign to me and my area of operation. I found one good used unit on Fleabay, and cured it.

This is diagnostics 101.

 

[jmillican1983]

Others may chime in, but i dont believe that you should have the Ground and Neutral bonded in the sub panel. Both the ground should be connected from the main and the Ufer on the shop, but neutral should be isolated.

There seems to be a lot of debate about this. I was told, in order for the breakers to trip during an over-current event, the neutral needed to be bonded in the sub panel. That being said, a bonded neutral in the sub panel is technically allowing the loads from the main panel a return path to the UFER ground in the shop which wouldn't be a good setup. Some say the current takes the easiest path to ground. Some say the load is trying to return to the source. Either way, the inspector was satisfied with the neutral bonded. There is a lot of info online, including here that simply says a 4 wire feeder should have the ground and neutral isolated, and a 3 wire feeder has them bonded. There is other info I read that says they shouldn't be bonded. At this point I do not know who to believe, but the county inspector is probably the most reliable source I have.

 

[jmillican1983]

This seems to make sense to me:
The reason you need to bond the neutral bar in the case of a three wire feeder, is to provide an effective ground-fault current path. Remember, electricity is trying to return to the source, not to the ground (earth). In the event of a ground-fault, you want the fault current to be able to have a low resistance path back to the source. This low resistance path should allow the fault current to be great enough, so that a circuit breaker (or other protective device) can activate and open the circuit.

 

[Terry D]

If you only have a 3-wire feeder to that sub panel, you have to have the neutrals and grounds bonded together

 

[wyliesdiesels]

There seems to be a lot of debate about this. I was told, in order for the breakers to trip during an over-current event, the neutral needed to be bonded in the sub panel.

whoever said that is wrong. the neutral bond in the main service panel works just fine for the fault current pathway. theres millions of subpanels in this country with isolated neutral busses where breakers can and do properly trip upon ground or line fault.

If what you were told is true, then the NEC wouldnt require a 4-wire isolated neutral subpanel in a detached or attached structure....

That being said, a bonded neutral in the sub panel is technically allowing the loads from the main panel a return path to the UFER ground in the shop which wouldn't be a good setup. Some say the current takes the easiest path to ground.Some say the load is trying to return to the source.

no, that is incorrect. a grounding electrode(rod or UFER) has NOTHING to do with the low impedance fault current pathway. the earth is a poor conductor of electricity. Grounding electrodes are for shunting lightning, limit voltage to earth, and shunting high voltage when primary lines contact secondary lines.

Also electricity doesnt take the easiest path to ground. It takes all paths to the source. This is why you dont want a bonded neutral in a subpanel, especially in a detached subpanel, where there are other metallic pathways (such as phone, cable, water, and gas pipes) connecting the buildings together.

Either way, the inspector was satisfied with the neutral bonded. There is a lot of info online, including here that simply says a 4 wire feeder should have the ground and neutral isolated, and a 3 wire feeder has them bonded. There is other info I read that says they shouldn't be bonded. At this point I do not know who to believe, but the county inspector is probably the most reliable source I have.

a 4-wire feed should connect to a subpanel with an isolated neutral.
a 3-wire feed should connecto a subpanel with a bonded neutral

grounding electrodes have no bearing on either of those.

This seems to make sense to me:
The reason you need to bond the neutral bar in the case of a three wire feeder, is to provide an effective ground-fault current path. Remember, electricity is trying to return to the source, not to the ground (earth). In the event of a ground-fault, you want the fault current to be able to have a low resistance path back to the source. This low resistance path should allow the fault current to be great enough, so that a circuit breaker (or other protective device) can activate and open the circuit.

yes, since there is no EGC, the neutral is used...

same concept for old 3-wire dryer and stove circuits

 

[jmillican1983]

Getting off topic at this point, but I will bite.

whoever said that is wrong. the neutral bond in the main service panel works just fine for the fault current pathway. theres millions of subpanels in this country with isolated neutral busses where breakers can and do properly trip upon ground or line fault.

What would happen in this scenario?:

-3 wire feed to sub panel
-Neutral and ground isolated in the sub panel
-An electrical device/appliance fed from the sub panel fails and the ground wire becomes energized
-The ground bar in the sub panel is bonded only to the UFER ground
-The UFER ground is not a good conductor (high resistance)

Wouldn't the high resistance path to ground through the UFER potentially limit the fault current to the point that the breaker for the failed device/appliance wouldn't trip, therefore energizing every metal junction box and the sub panel frame as well? The fault current has nowhere else to go if the neutral is not bonded.

Bonding the neutral allows that fault current a low resistance path back to the source, which would allow the fault current to increase dramatically and trip the breaker.

Tell me how the breaker would trip if the neutral was not bonded in the sub panel?

 

[Terry D]

Getting off topic at this point, but I will bite.



What would happen in this scenario?:

-3 wire feed to sub panel
-Neutral and ground isolated in the sub panel
-An electrical device/appliance fed from the sub panel fails and the ground wire becomes energized
-The ground bar in the sub panel is bonded only to the UFER ground
-The UFER ground is not a good conductor (high resistance)

Wouldn't the high resistance path to ground through the UFER potentially limit the fault current to the point that the breaker for the failed device/appliance wouldn't trip, therefore energizing every metal junction box and the sub panel frame as well? The fault current has nowhere else to go if the neutral is not bonded.

Bonding the neutral allows that fault current a low resistance path back to the source, which would allow the fault current to increase dramatically and trip the breaker.

Tell me how the breaker would trip if the neutral was not bonded in the sub panel?

What wylies is talking about millions of sub panels not having the neutral bonded is because they have a 4-wire feed to them. A equipment ground is carried from the main panel to the sub panel. In this case, you dont bond the neutral at the sub panel. that will put current on your equipment ground back to the main panel. If someone would undo that ground with current on it there would be a potential shock hazard. The ground rod is just for lighting stikes

In your case where there in no equipment ground carried from the main panel, you have to bond the neutral, the same way it is done in your main panel, to give it a low resistance path, or any ground faults would have to rely on just that ground rod to try and trip the breaker, which even if it was possible, the time it took could cause death. Without doing this, your ground buss has no reference to ground other than the rod

The scenario that you mentioned above would be dangerous

 

[wyliesdiesels]

Getting off topic at this point, but I will bite.



What would happen in this scenario?:

-3 wire feed to sub panel
-Neutral and ground isolated in the sub panel
-An electrical device/appliance fed from the sub panel fails and the ground wire becomes energized
-The ground bar in the sub panel is bonded only to the UFER ground
-The UFER ground is not a good conductor (high resistance)

Wouldn't the high resistance path to ground through the UFER potentially limit the fault current to the point that the breaker for the failed device/appliance wouldn't trip, therefore energizing every metal junction box and the sub panel frame as well? The fault current has nowhere else to go if the neutral is not bonded.

Bonding the neutral allows that fault current a low resistance path back to the source, which would allow the fault current to increase dramatically and trip the breaker.

Tell me how the breaker would trip if the neutral was not bonded in the sub panel?

First of all your scenario is not possible. If the panel has a 3-wire feed, then it wouldnt have an isolated neutral. that was never how things were wired. it would have a bonded neutral and it wouldnt have a separate ground bar. neutrals and grounds would go to the same bar, as would the GEC.

but the issue with the bonded neutral in the sub has to do with the bonding and if the neutral connection to the main panel (or upstream subpanel) becomes compromised, then the neutral current can use an alternate pathway such as water piping.

heres an example. the neutral, in a feeder to a detached structure, develops a bad connection. the subpanel it feeds, has a bonded neutral, thus any metallic pathways or objects that are bonded to the panel enclosure are at the same potential as that neutral bus. someone leans against the panel while touching a metal water pipe that is bonded at the main panel upstream. now neutral return current is flowing through the person and onto the water pipe to get back to the main panel and ultimately the transformer.

so having that bonded neutral sets the stage for a potential for shock should the neutral become comprimised.

this is the same reason range and dryer circuits were no longer allowed to be 3-wire with a bonded neutral at the appliance frame.

as to how breakers would trip if the neutral was not bonded in the subpanel. thats simple. the neutral is bonded at the main and the EGC in the subpanel feeder carries the fault current back to the neutral bond in the main. this causes high current to flow and the branch breaker sees this and trips.

 

[jmillican1983]

First of all your scenario is not possible. If the panel has a 3-wire feed, then it wouldnt have an isolated neutral. that was never how things were wired. it would have a bonded neutral and it wouldnt have a separate ground bar. neutrals and grounds would go to the same bar, as would the GEC.

I thought You had said earlier that the neutral and ground should NOT be bonded in the sub panel with a 3 wire feed. I read back through it and didn't see where you said that, but post #6 was edited out of this thread. Either way, I was making the case that they should be bonded. I agree that my scenario was creating a dangerous situation, making the argument to keep my neutral and ground bonded at the sub panel.

as to how breakers would trip if the neutral was not bonded in the subpanel. thats simple. the neutral is bonded at the main and the EGC in the subpanel feeder carries the fault current back to the neutral bond in the main. this causes high current to flow and the branch breaker sees this and trips.

This only works for a 4 wire feed. A 3 wire feed would not trip the breaker if the neutral was not bonded in the sub panel. Again, I was making the case that an isolated neutral-ground in the sub-panel is dangerous and won't allow a breaker to trip.

I understand the hazard created with a 3 wire feed if the neutral wire fails between the main and sub panel. Wouldn't that same hazard exist if the neutral failed coming into the main panel from the utility provider?

 

[Terry D]

Let me add something to what I said. On a 3-wire feed to a sub panel, when I was saying that the neutral has to be bonded, I meant it has to be bonded if the neutrals and ground were on separate bars. If the neutrals and grounds were on the same bar including the GEC, The only thing that would be accomplished by bonding would be to bring the metal enclosure and any piping coming out of it to ground potential.

 

[wyliesdiesels]

I thought You had said earlier that the neutral and ground should NOT be bonded in the sub panel with a 3 wire feed. I read back through it and didn't see where you said that, but post #6 was edited out of this thread. Either way, I was making the case that they should be bonded. I agree that my scenario was creating a dangerous situation, making the argument to keep my neutral and ground bonded at the sub panel.

no i never wouldve said that.

im quite familiar with the different scenarios. go read my write ups on this subject here

https://www.garagejournal.com/forum/showthread.php?t=356460

This only works for a 4 wire feed. A 3 wire feed would not trip the breaker if the neutral was not bonded in the sub panel. Again, I was making the case that an isolated neutral-ground in the sub-panel is dangerous and won't allow a breaker to trip.

not quite

if the panel is fed with metallic conduit, then it would be a 3-wire feed and have an isolated neutral but breakers would still be able to trip.

I understand the hazard created with a 3 wire feed if the neutral wire fails between the main and sub panel. Wouldn't that same hazard exist if the neutral failed coming into the main panel from the utility provider?

yes and theres another more dangerous hazard that a failed neutral in a 3-wire feed can cause. i forgot to mention this earlier. if the neutral becomes open, current would still flow but the voltage on each hot leg will vary wildly and can end up being a lot higher or lower than 120v. the total voltage between the legs will always be 240v but one leg could be 80v while the other could be 160v, for example and could cause a fire...

but in the service drop from the utility, we arent concerned about breakers being able to trip because there are no breakers on the utility feed. so its a different animal

 

[jmillican1983]

but in the service drop from the utility, we arent concerned about breakers being able to trip because there are no breakers on the utility feed. so its a different animal

If the neutral is open in the service drop from the utility, any fault current has lost it's low resistance path back to the source, creating a risk for shock due to breakers not tripping. This hazard is not limited to 3 wire feeds.

Also, here is the deleted comment I was referring to:

Originally Posted by wyliesdiesels View Post
whoever said that is wrong. the neutral bond in the main service panel works just fine for the fault current pathway. theres millions of subpanels in this country with isolated neutral busses where breakers can and do properly trip upon ground or line fault.

I misunderstood your statement to mean "isolated neutral busses" on 3 wire feeds,

 

[wyliesdiesels]

If the neutral is open in the service drop from the utility, any fault current has lost it's low resistance path back to the source, creating a risk for shock due to breakers not tripping. This hazard is not limited to 3 wire feeds.

youre not making any sense here. there is no ground fault current protection on the line side of a service drop from a utility. There is no breaker on the pole, to trip. there isnt even anything that fault current could energize between the service drop handoff and the pole so im not even sure why you brought this up. its neither here nor there.

Also, here is the deleted comment I was referring to:

Originally Posted by wyliesdiesels View Post
whoever said that is wrong. the neutral bond in the main service panel works just fine for the fault current pathway. theres millions of subpanels in this country with isolated neutral busses where breakers can and do properly trip upon ground or line fault.

I misunderstood your statement to mean "isolated neutral busses" on 3 wire feeds,

except i didnt delete that comment. its still there

 

[jmillican1983]

My mistake.