HVAC Electrical Issue: Current Leakage from Furnace

[UrbanFarmEngineer]

Hey Everyone,
I am not an electrical engineer I am a biochemmie and I am here for your help. I have been learning household wiring and troubleshooting electrical wiring problems in my house for the last few months and I am stumped on an HVAC issue and would really appreciate some insights from you guys. I am here to learn, I have been absolutely loving learning about how my home's wiring works, how to find and correct faults and about AC circuits in general. Having said that, I am learning: if I get some of the terminology wrong I request your correction and ask to please help me understand the main subject matter that I am inquiring about:

Here is the issue:

1. I noticed net current on the main power supply from my main panel going to the furnace;
2. When I turned the furnace breaker on I noticed net current at the panel on several other lines coming from different breakers: the sum of the net current on each separate breaker did not add to the net current on the power feed when measured at the the furnace; so
3. Isolated the furnace onto an extension cord to ensure that there were no shunted neutrals in hidden jboxes (its a 50s house) and plugged in to an isolated outlet and all of the stray current on the breakers disappeared but there was STILL net current on the extension cord
4. Looked inside the furnace and saw it wasn't grounded inside the junction box where the main power supply enters: DID that and most of the stray current disappeared
   
   
   I STILL HAVE about 20mA that fluctuates and peaks at 1A ON THE CORD. SO THERE IS STILL SOME LEAKAGE COMING FROM A COMPONENT INSIDE

I have a Klein Tools clamp meter that is calibrated and accurate to 4% so it is not my meter. I have checked it with other clamp meters also and results are same. The furnace is less than one year old and under p&s warranty but the company refuses to send someone to check it out.

Can someone help me to understand which components in the furnace may still be causing this bit of micro leakage? I would really love to learn from you all.

Thank you

 

[nsaspook]

That small current is nothing to worry about. It could be from a AC line EMI/noise filter, indicator lamp or some sort of power safety circuit.

 

[UrbanFarmEngineer]

That small current is nothing to worry about. It could be from a AC line EMI/noise filter, indicator lamp or some sort of power safety circuit.

I know I just want to get to the bottom of it. It' a brand new furnace, less than a year old--and the gas valve just malfunctioned and got stuck in the open position and started letting gas into the house. The current was actually higher until I found the lack of grounding issue, so I want to diagnose anything however small it may be because I am so not impressed with this company, their refusal to do electrical service and how shoddy the whole thing has been.

How would I start to diagnose? The current fluctuates also when the fan goes on, might it be the fan capacitor?
A/C line is on a separate breaker and is powered off right now.

 

[Guineafowl]

Are you measuring all three conductors (L,N,E) inside the clamp? A more usual method would be to clamp L+N, thereby measuring all leakage, both down the earth conductor and through the physical earth.

The pass level for a standard 30mA RCD/GFCI (mandatory for all new lighting/socket circuits over here) is 18-30mA, so a single appliance leaking 20mA is quite a lot, to put it in context. Especially at half the voltage.

Also, I have an old factory-type electrical test station, and the allowable leakage thresholds are 5, 10 and 15mA, so again, something’s not right.

You’d normally use a Megger insulation tester on each part of the (unpowered) appliance to find the fault. However, in your case the IR is so low you might get some results with a normal multimeter.

Start by measuring resistance between L-E and N-E at the power inlet. Assuming you find the fault there, try disconnecting components until the fault clears, then zoom in on the offending part.

 

[berkeman]

Welcome to PF.

It' a brand new furnace, less than a year old--and the gas valve just malfunctioned and got stuck in the open position and started letting gas into the house.

I am so not impressed with this company, their refusal to do electrical service

Wait, so this brand new furnace had a problem that resulted in a stuck-open gas valve that leaked gas into your house, and the manufacturer refuses to service/inspect it?

 

[Averagesupernova]

@UrbanFarmEngineer what you are saying is not very clear to me. Maybe English is not your first language? Your use of the term net current is confusing to me. Net would imply that you are measuring several currents and subtracting some amount of for whatever reason.
-
Without knowing what part of the world you are in it is difficult to help you. I can give you alot of advice if you are in the USA. Beyond that, the standards are different enough that I may not be of much help.

 

[Guineafowl]

@UrbanFarmEngineer what you are saying is not very clear to me. Maybe English is not your first language? Your use of the term net current is confusing to me. Net would imply that you are measuring several currents and subtracting some amount of for whatever reason.
-
Without knowing what part of the world you are in it is difficult to help you. I can give you alot of advice if you are in the USA. Beyond that, the standards are different enough that I may not be of much help.

Something like this:

Any net current would be earth leakage. The AC clamp meter will miss DC leakage, which is increasingly common. There are various forms of RCD to account for modern loads:
Type AC - mains freq. AC leakage only
Type A - above + pulsed DC
Type F - above + higher frequencies
Type B - above + smooth DC.

You select the one appropriate for the load, eg electronic, VFD, car charger, etc.

 

[UrbanFarmEngineer]

Are you measuring all three conductors (L,N,E) inside the clamp? A more usual method would be to clamp L+N, thereby measuring all leakage, both down the earth conductor and through the physical earth.

The pass level for a standard 30mA RCD/GFCI (mandatory for all new lighting/socket circuits over here) is 18-30mA, so a single appliance leaking 20mA is quite a lot, to put it in context. Especially at half the voltage.

Also, I have an old factory-type electrical test station, and the allowable leakage thresholds are 5, 10 and 15mA, so again, something’s not right.

You’d normally use a Megger insulation tester on each part of the (unpowered) appliance to find the fault. However, in your case the IR is so low you might get some results with a normal multimeter.

Start by measuring resistance between L-E and N-E at the power inlet. Assuming you find the fault there, try disconnecting components until the fault clears, then zoom in on the offending part.

THNK YOU for this. This is extremely informative and points me in the right direction because as I mentioned I am just learning.

I am measuring the leakage at the point where I installed the extension cord which is on the line side of the furnace's power switch. I am measuring L & N. I haven't gone into the furnace yet because I didn't know where the best place to start would be. I am learning.

Regarding what you said: "Start by measuring resistance between L-E and N-E at the power inlet. Assuming you find the fault there, try disconnecting components until the fault clears, then zoom in on the offending part." Is there any risk of damage if I say disconnect a component and them power on the furnace. Foe example if I remove the fan capacitor and power the furnace up? (I suspect the fan capacitor because the stray current really starts jumping around when the fan is starting up)

 

[UrbanFarmEngineer]

Something like this:
View attachment 342803

Any net current would be earth leakage. The AC clamp meter will miss DC leakage, which is increasingly common. There are various forms of RCD to account for modern loads:
Type AC - mains freq. AC leakage only
Type A - above + pulsed DC
Type F - above + higher frequencies
Type B - above + smooth DC.

You select the one appropriate for the load, eg electronic, VFD, car charger, etc.

@Averagesupernova

Thank you guys for this, I really appreciate the explanation! English is my first language but I am just learning about HVAC so I may not be using the correct terminology and please do correct me where I make mistakes as it helps me to learn.

Here is how I am taking the measurement:
I am so far measuring on the line side of the main power switch to the furnace and my clamp is set on AC and I am measuring the hot and neutral inside the clamp. I've run an extension cord to isolate the furnace so I know that it is not neutral current leaking from some burried jbox in my 1950s house.

Is there a better way to do this? What do you suggest?

 

[Guineafowl]

At this stage I would stick to dead testing. Hopefully, the multimeter will show a low resistance reading between either L-E or N-E. You can then disconnect bits until the reading jumps back up.

Testing at 500V with a Megger, a pass is a resistance reading of $1 M\Omega$ typically. With a 6-9V meter test voltage, you’d expect it to be even higher. Let us know what the readings are.

If you mean the start capacitor for the fan, that should be nowhere near earth, but we’ll get to that. Don’t start powering anything on until you’ve exhausted the dead testing.

 

[UrbanFarmEngineer]

Welcome to PF.



Wait, so this brand new furnace had a problem that resulted in a stuck-open gas valve that leaked gas into your house, and the manufacturer refuses to service/inspect it?

Hey, thank you for the warm welcome.

Thankfully the company that sold it to me did replace the gas valve but they refuse to do any sort of work, even a diagnostic for anything electrical.

Sorry, if my comment before sounded scary! lol. No I am not sitting in a house slowly filling with gas... thankfully!!!

 

[UrbanFarmEngineer]

At this stage I would stick to dead testing. Hopefully, the multimeter will show a low resistance reading between either L-E or N-E. You can then disconnect bits until the reading jumps back up.

Testing at 500V with a Megger, a pass is a resistance reading of $1 M\Omega$ typically. With a 6-9V meter test voltage, you’d expect it to be even higher. Let us know what the readings are.

If you mean the start capacitor for the fan, that should be nowhere near earth, but we’ll get to that. Don’t start powering anything on until you’ve exhausted the dead testing.

OK got it. I will start with what you have advised and and post my results. I will try to post some pics to better explain myself.

I really appreciate your patience and the patience that everyone who has replied has demonstrated in explaining this to me, i.e., someone who is learning. People on other forums are not nearly as patient and many are just crazy. But this approach really facilitates learning and helping each other and as I said I really appreciate that.

 

[Averagesupernova]

Do you have the ground wire inside the clamp on your ammeter as well as the hot and neutral?

 

[UrbanFarmEngineer]

Do you have the ground wire inside the clamp on your ammeter as well as the hot and neutral?

No. Just the hot and neutral.
Clamp meter setting is on AC.
Measurement is taken on line side of main furnace switch.

This is how I have been checking the household wiring. Should I do something different in this case or is my method OK for this application as well? ...or is there a better way (even if this method is OK)?

 

[UrbanFarmEngineer]

@UrbanFarmEngineer what you are saying is not very clear to me. Maybe English is not your first language? Your use of the term net current is confusing to me. Net would imply that you are measuring several currents and subtracting some amount of for whatever reason.
-
Without knowing what part of the world you are in it is difficult to help you. I can give you alot of advice if you are in the USA. Beyond that, the standards are different enough that I may not be of much help.

OK I perhaps the proper term I should be using is "earth leakage" instead of "net current." Thank you @Guineafowl the current goes from the furnace right back to the panel. So let me know if that is still the correct term.

 

[Averagesupernova]

If you have a current indicated on your meter with only the hot and neutral in the clamp then you have an imbalance. This means current is flowing where it is not supposed to be. Now you mention 20 mA. When you attached a cord to your furnace and plugged it in, where did you plug it in? A good working GFCI outlet should have tripped instantly if your readings with the clamp meter are actually correct.
-
What model Klein do you have?
-
I have this one: https://www.physicsforums.com/threa...-when-measuring-current.1045614/#post-6801421

I haven't used it much since I got it. Prior to the Klein I've had an old AW Sperry. I dropped it and now the needle won't move.
-
Edit. Scroll a bit farther to the pic when you click on that link. I am not sure why it goes to a few posts earlier than the one wanted to link to.

 

[berkeman]

Looked inside the furnace and saw it wasn't grounded inside the junction box where the main power supply enters: DID that and most of the stray current disappeared

Can you say more about this? Who installed this furnace? Do you have a copy of the Service Manual or Installation Instructions for this furnace? One would think that the wiring to the house AC Mains would be covered by those documents.

 

[Averagesupernova]

It's usually pretty straight forward. The furnace manufacturer tries to make it easy. They don't want the installer to accidentally mess up some of their own internal wiring.
-
Leaving ground unhooked is just plain laziness in my opinion. Or basic stupidity. Not knowing enough to know where that wire goes and still thinking they can handle the job doesn't show much for smarts.

 

[Guineafowl]

^ and there lies the rub. Domestic electrics will work if you simply hook up L/N, ignoring wire sizes, cable runs, breaker specs, and the most basic protection against fire/shock - the earth/ground wire.

 

[Averagesupernova]

^ and there lies the rub. Domestic electrics will work if you simply hook up L/N, ignoring wire sizes, cable runs, breaker specs, and the most basic protection against fire/shock - the earth/ground wire.

Oh but it gets better than that. Recessed ceiling lights, AKA can lights, have max bulb size (watts) ratings and it's based on whether it is covered in insulation or not. It's not that straight forward. You don't have to be messing with wires and circuit breakers to set up a dangerous situation. Most bulbs are going to higher efficiency with LEDs and such but for a long time I suspect alot of can lights were not running 'safe'.

 

[Tom.G]

Does your ammeter still read a net current flow if you measure with all three wires, Hot, Neutral, Safety Ground , thru the clamp-on?

1) If it does, then see if you can put the clamp-on around the Gas line to see if the current is flowing there. If you get current flow there, then the Safety Ground (Green wire in the USA) is not connected.

2) If it does not, then the Grounding is likely OK but you still have leakage somewhere in the furnace.

You state that you suspect the fan motor capacitor. These are often metal-cased and mounted either to the motor or nearby.

There could be an internal problem with it. With the power off, remove the capacitor from its mounting and wrap it in a cloth or something so the case can not touch any metal. Leave the wires connected.

Then turn the power back on and see if the strange current readings have disappeared.

Motors can also fail in a similiar manner, when they do though, they will typically trip a circuit breaker as soon as they are energized.

Another possible fault location is if there is an "ignitor" rather than a pilot light to ignite the main burner. These tend to have a higher failure rate than the rest of the parts.

Cheers,
Tom

P.S. Please keep us informed on what you find (or fix!), We like to learn too.

 

[UrbanFarmEngineer]

Can you say more about this? Who installed this furnace? Do you have a copy of the Service Manual or Installation Instructions for this furnace? One would think that the wiring to the house AC Mains would be covered by those documents.

Let me take some pics. I don't use a mobile phone so I will have to get a friend over that has one and take pics to better express what I am saying.

When the HVAC guy came over to fix the valve, I put everything back exactly to how it was so they couldn't say that I "voided the warranty" by altering the unit and when I pointed out the ground wire not being connected to the ground lug/ground screw (again sorry if terminology is wrong let me know if I need to explain better) the HVAC guy looked shocked/panicked and immediately attached it.

I think that it is what @Averagesupernova was saying, just negligence.
The Furnace is a York TM9V Series. Here is the installation manual. It says right on the diagram "Connect ground lead to screw" inside the electrical entry junction box...dang.
I don't have any explanation for that one except that the installer didn't do it..
https://www.manualslib.com/manual/865111/York-Tm9v-C.html?page=12#manual

My Ammeter is Klein Tools CL120. I don't have a professional one because I don't use it for work I just got it because I am just trying to fix up my crazy 1950s house wiring and bring it up to code so I thought the CL120 would be sufficient... let me know what you guys think.

I see more helpful comments.. thank you I will have to get back in a bit to respond.

 

[UrbanFarmEngineer]

Does your ammeter still read a net current flow if you measure with all three wires, Hot, Neutral, Safety Ground , thru the clamp-on?

1) If it does, then see if you can put the clamp-on around the Gas line to see if the current is flowing there. If you get current flow there, then the Safety Ground (Green wire in the USA) is not connected.

2) If it does not, then the Grounding is likely OK but you still have leakage somewhere in the furnace.

You state that you suspect the fan motor capacitor. These are often metal-cased and mounted either to the motor or nearby.

There could be an internal problem with it. With the power off, remove the capacitor from its mounting and wrap it in a cloth or something so the case can not touch any metal. Leave the wires connected.

Then turn the power back on and see if the strange current readings have disappeared.

Motors can also fail in a similiar manner, when they do though, they will typically trip a circuit breaker as soon as they are energized.

Another possible fault location is if there is an "ignitor" rather than a pilot light to ignite the main burner. These tend to have a higher failure rate than the rest of the parts.

Cheers,
Tom

P.S. Please keep us informed on what you find (or fix!), We like to learn too.

OK this is extremely helpful too.
Thank you for this post.
Please give me some time to respond. I have a bunch of things going on at the same time at the moment and I want to give a detailed response to your questions.

 

[berkeman]

when I pointed out the ground wire not being connected to the ground lug/ground screw (again sorry if terminology is wrong let me know if I need to explain better) the HVAC guy looked shocked/panicked and immediately attached it.

Okay, that makes sense. No idea why the initial installer missed that, but glad it's fixed now.

I am just trying to fix up my crazy 1950s house wiring and bring it up to code

That is a good goal and project. If you can, I'd suggest that you talk to your local building inspection department to see what-all is involved in pulling a permit for the work you are doing. That way when you are done and have passed the final permit inspection, you can use that in the future whenever you sell the house. At least around here where I live (northern California), a home inspection is usually done prior to selling a home, and code violations can be caught at that stage and hold up the sale of the home.

 

[Averagesupernova]

a home inspection is usually done prior to selling a home, and code violations can be caught at that stage and hold up the sale of the home.

Just about anything can hold up the sale of a house. Truth is, a home inspector has absolutely no teeth in enforcing anything. Many cases an electrician is called on because of something a home inspector flags and in reality the house is up to code to the point of when it was built. The law requires nothing more. Electricians generally hate home inspectors. The consensus is usually they catch things that are not issues and miss things that are. The most famous one is two wires coming off a single breaker is not permitted. This is false. Some breakers are built to accept more than one wire per.
-
I would personally rather buy a house that was wired to code when built and had absolutely nothing done since. I've seen too many hacks that turn out to be worse than leaving it as it was originally. All the supposed updates done and it is reflected in the price and usually turns out most of it needs to be done over.

 

[berkeman]

and in reality the house is up to code to the point of when it was built.

Good point, ungrounded outlets being an example of that (if the house were built long enough ago).

 

[UrbanFarmEngineer]

Okay, that makes sense. No idea why the initial installer missed that, but glad it's fixed now.


That is a good goal and project. If you can, I'd suggest that you talk to your local building inspection department to see what-all is involved in pulling a permit for the work you are doing. That way when you are done and have passed the final permit inspection, you can use that in the future whenever you sell the house. At least around here where I live (northern California), a home inspection is usually done prior to selling a home, and code violations can be caught at that stage and hold up the sale of the home.

Totally agree. I have a permit for the stuff that required a permit in my province in Canada (Ontario) and I just looked up inspections on DIY electrical work and they offer them here. Thanks so much for this tip! Very valuable, especially since I do eventually want to sell this house.

 

[Nik_2213]

Tangential, my very new, wall-hung c/h combi-boiler kept 'locking up' every 2~~3 days, refusing to respond to remote control / room-stat. Each time resolved by traditional switch boiler off, swear a baker's dozen, switch on again.
Solved by clipping a split ferrite core onto power feed to manual controller / base-station.
Note no re-wiring required, the two halves just clip on...

 

[UrbanFarmEngineer]

^ and there lies the rub. Domestic electrics will work if you simply hook up L/N, ignoring wire sizes, cable runs, breaker specs, and the most basic protection against fire/shock - the earth/ground wire.

RIght? Flick the switch Lights go on and it's"all good" haha that is what I think happened in my case before I moved in. I think it was a DIY job and I didn't look inside until sparks started coming out of a switch and smoke started coming out of a fixture. Then when I started to open boxes and have a look at the rest ... it was bad. really bad.

 

[UrbanFarmEngineer]

Tangential, my very new, wall-hung c/h combi-boiler kept 'locking up' every 2~~3 days, refusing to respond to remote control / room-stat. Each time resolved by traditional switch boiler off, swear a baker's dozen, switch on again.
Solved by clipping a split ferrite core onto power feed to manual controller / base-station.
Note no re-wiring required, the two halves just clip on...

really!
What do you suspect the original problem was?
and what size core did you use?

 

[Nik_2213]

I had a part-baggy of them, dating from final attempt to make some [REDACTED] WiFi range-extenders work for more than a week at a time without reset.
Ferrites are about the size of top finger-bone, two halves just clip together over 8mm OD 'Twin+E' flex. So, probably 9mm ID, with a range of sizes available.

Look for such under A**n's passive components, ferrite cores: They're about 2/$
I'd clipped one onto the Dyson 'hoover' lead, three onto the fridge-freezer lead, sundry else-where.

Yes, multiple turns through a toroid are much 'better' electronically, but such tight turns would put undue strain on the flex. Also, with many UK appliances having 'moulded-on' plugs, I'd have to cut such off...

I reckon culprit was the big fridge/freezer's 'stat' cycling, but never proved it: Partial correlation at best...
For a long time, other suspect was the kitchen's suspended ceiling, whose strip-lights became an endless race to keep up with their flickery tubes and blinky-blink starters. I've since replaced all such with pin-compatible LED strips, just needed to swap the blinky starters by supplied dummy 'place-holders'. So, with hind-sight, was not them....

FWIW, I abandoned those [REDACTED] WiFi repeaters, hard-wired Cat_5 cable to a couple of WAP/routers. After which, every-one was happy until BossCat, a BIG 'Spotty', figured how to un-latch patch-cables that lacked protective boots....
Poltercats !!!

 

[UrbanFarmEngineer]

OK furnace finally failed. I didn't even get to start the dead testing yet.

Right now, when I switch it on and call for heat, I can hear it starting up and it reaches a steady hum but the fan won't start. I took out the fan motor capacitor which is a 5uF +-5% and tested it. The reading really bounced around a lot then settled on 4.62uF (just out of range) then dropped to 0 after that.

I noticed that there is a little rubber cap that is supposed to cover the contacts and it was not on, just dangling down the wires. I put the rubber cap back on and no stray current on the power line. So I am thinking that what @Tom.G . said may be the solution.


Just for your entertainment, enjoyment and for a little laugh on a Friday afternoon:

The company is really mad and says their techs insist that there is nothing wrong with the electrical but none of the techs did any sort of electrical diagnostics AT ALL and upon visual inspection they missed the fact that the ground lead was not connected in the electrical entry box inside the furnace. None of the techs that came even had the equipment to do electrical testing and the company is telling me that if I touch the furnace any more I will void my warranty. DO I laugh? DO I cry? DO I just buy a new capacitor and install it myself and pretend it was like that the whole time... One of the company representatives who I talked to on the phone tried to insist to me for a good long time that there was in fact NO fan motor capacitor inside my furnace for a while...

Is this incompetence?
Is it gas-lighting?
Is it a miscommunication somehow?
It's been easier to learn HVAC from scratch than it has been to deal with the "techs" and "experts."

--------OK Enough entertainment for now--------------

Anyhow this thread has been extremely useful and I have learned a lot. Thank you all for your detailed explanations and your troubleshooting ideas. I have actually become really fascinated with HVAC and hope to continue learning about it when this dumpster fire of a problem is done.

I will post updates when the company sends their techs again on Monday. Happy weekend!

 

[berkeman]

I would recommend that you avoid trying to troubleshoot this furnace if it is failing this badly under warranty. If the tech who responds to your warranty complaint sees that you've been taking it apart trying to troubleshoot it, they can deny service for cause.

(Full disclosure -- my wife works in the home warranty industry, and deals with DIY homeowner "fixes" on warrantied appliances regularly...)

 

[UrbanFarmEngineer]

I would recommend that you avoid trying to troubleshoot this furnace if it is failing this badly under warranty. If the tech who responds to your warranty complaint sees that you've been taking it apart trying to troubleshoot it, they can deny service for cause.

(Full disclosure -- my wife works in the home warranty industry, and deals with DIY homeowner "fixes" on warrantied appliances regularly...)

I know.
It is just so frustrating when you watch them fumble around and not even have the tools to diagnose.
You are right tho. It just makes things so much harder.