Measure voltage drop with no load at all?

[escopf]

Hello and thanks, always, for your time.

A homeowner of a new-construction house (Florida) has been having her TV's power off intermittently, lights dim sometimes etc. So my first step is to see what branch-circuit drop is. At the outlets farthest from the panel I measure 123v@no load, 121v@minimal load, 114v@7A load and 101v@15A . The line is 15A OCPD so at 12A it would be dropping well over 10%.

I know the NEC recommends but does not mandate a 5% maximum drop, but Florida's building code (FBC Chapter 13) does mandate a 3% on branches (2%_{service drop}/3%_{branch drop} or reverse according to our local inspector).

I called the local inspector to ask if our location was subject to FBC chapter 13 and he said yes, absolutely. I then asked him how it should be measured (my guess would be 180VA/receptacle) and he told me that putting any load on the circuit in question is not how voltage drop is tested for that article. In his words: "Of course the voltage will drop if you're putting load on it, you need to measure the voltage drop across just the branch circuit wires, with no load connected. You want to measure the resistance of the branch circuit, not the load you attach to it".

I have been doing voltage drop calculations for many (too many) years and have never come across this method, and I'm a bit skeptical here. Also I believe that the language in FBC Chapter 13 states "design load".

So, question 1: Is this an accepted method of testing anywhere?
question 2: If the inspector is mistaken what is a recommended/professional way to proceed with the municipality in testing/enforcement?

thanks as always

 

[iwire]

Unless the circuit is fairly long I would be looking for a bad connection and forget about conductor voltage drop.

As far as no load testing for voltage drop? :lol: Yet that is exactly what the GSA inspector did after I renovated IRS offices. He tested with a simply DMM, not surprisingly we met the job specified voltage drop requirements.

 

[iwire]

Did you happen to check what the voltage was like at the panel end with and without load?

 

[mgookin]

Hello and ... what is a recommended/professional way to proceed with the municipality in testing/enforcement?

thanks as always

Why is the city involved?

How old is this house? A builder has to give a minimum one year warranty on construction in Florida. If it's less than a year old, call the builder to have it fixed.

How many sf is the house? How long can that home run be? Are we talking a 8,000 sf house or 1,200 sf?

Are the problems limited to one branch circuit? If so there's a bad connection somewhere or a failed device.

 

[infinity]

So, question 1: Is this an accepted method of testing anywhere?

No load testing? I would leave the tester in the truck being that it's a waste of time. :roll:

 

[ActionDave]

Hello and thanks, always, for your time.

A homeowner of a new-construction house (Florida) has been having her TV's power off intermittently, lights dim sometimes etc. So my first step is to see what branch-circuit drop is. At the outlets farthest from the panel I measure 123v@no load, 121v@minimal load, 114v@7A load and 101v@15A . The line is 15A OCPD so at 12A it would be dropping well over 10%.....

I agree with the others, you don't have a voltage drop problem that can be solved with a calculator you have a problem with the wiring somewhere in the house.

 

[mgookin]

PS: Welcome to the forum neighbor.

 

[escopf]

Did you happen to check what the voltage was like at the panel end with and without load?

Thanks for the reply,

I tested only drop at receptacles unloaded vs. loaded on 3 branch circuits (farthest from panel) and saw similar drop everywhere. The runs are well over 100' and I'd imagine 200' in some places, all #14. The service traverses the house and is an addt'l ~100'. Once I spoke to the inspector and he said there's no problem at all with seeing 98 volts at a television, I put the tester away...

I'm a big Jack fan myself BTW:thumbsup:

 

[escopf]

Thank you for your reply.

Why is the city involved?

Because I plan to call the builder and I'm expecting him to say 'it's fine, let's call the inspector'

How old is this house? A builder has to give a minimum one year warranty on construction in Florida. If it's less than a year old, call the builder to have it fixed.

They closed this past March, new construction

How many sf is the house? How long can that home run be? Are we talking a 8,000 sf house or 1,200 sf?

6,000 S.F. only one floor so longer runs than usual. The service traverses the house (~100') and the branch circuitry is between 100' and 200' in some places

Are the problems limited to one branch circuit? If so there's a bad connection somewhere or a failed device

I've measured similar drop on 3 branches.

thanks again

 

[escopf]

PS: Welcome to the forum neighbor.

Thanks for having me!

 

[ggunn]

Without current flow there can be no voltage drop.

 

[iwire]

I tested only drop at receptacles unloaded vs. loaded on 3 branch circuits (farthest from panel) and saw similar drop everywhere. The runs are well over 100' and I'd imagine 200' in some places, all #14. The service traverses the house and is an addt'l ~100'. Once I spoke to the inspector and he said there's no problem at all with seeing 98 volts at a television, I put the tester away...

In my opinion you cannot talk about voltage drop on a branch circuit unless you have also checked the source loaded and unloaded.

This problem could be a bad utility connection.

I'm a big Jack fan myself BTW:thumbsup:

:thumbsup:

 

[gar]

160729-0905 EDT

escopf:

From your measurements I get 9/7 = 1.3 ohms, and 22/15 = 1.5 ohms for the total loop impedance. Total loop impedance includes power company transformer, all wire, connections, and breakers in the loop where the current flows. Average your two measurements and the result is 1.4 ohms.

By comparison I measure 2/8 = 0.25 ohms at a location in my family room. At my main panel I measure 0.6/8 = 0.075 ohms measured at the panel input terminals (before main fuses). My branch circuit measurement includes the 0.075 ohms plus main fuse, bus bars, breaker to subpanel, #8 copper to subpanel, breaker to final branch circuit, #12 copper to duplex outlet.

The voltage measurement was made from one half of the duplex and the test load was plugged into the other half. This means my measurement of voltage was essentially at the input terminals of the duplex.

You need to do voltage change measurements at various points with your changing on-off test load.

If the voltage change at the main panel is a large fraction of what it is at the final load point, then there could be a problem before the main panel.


.

 

[charlie b]

If voltage drop is the issue under consideration, then you can use a procedure not unlike what the local inspector tried to describe. But he described it incorrectly. With no load (and indeed with the power off), you can measure the resistance of the conductor(s) from the breaker to the receptacle. From that, you can calculate what the voltage drop would be, at any load value you care to name. Yes, this would work. But it is too much work. You can take the easier path by imposing a load and measuring voltage at the receptacle.

I agree with Bob's suggestion of measuring voltage both at the receptacle and at the breaker. If you get 101V at the receptacle with a 15 amp load, but you also get 102V at the breaker, then the voltage drop is taking place upstream of the panel. In other words, it is the utility's problem.

 

[escopf]

Apologies for the late reply; I didn't see the page 2 button right away....

In my opinion you cannot talk about voltage drop on a branch circuit unless you have also checked the source loaded and unloaded.

Yes, I agree. I tested 3 different branches against each other (at receptacles) while one was loaded and only the loaded one dropped significantly. I call that the lazy-man's panel test, but agreed, when I go back I need to take the cover off and be more thorough (and simple...)

My concern now is being able to compel the builder, if I find that this system is indeed faulty, to address the issue, especially if his electrician/AHJ employs the 'local' testing method.

thanks for your time

 

[escopf]

Thanks. This was very helpful:

160729-0905 EDT

escopf:

From your measurements I get 9/7 = 1.3 ohms, and 22/15 = 1.5 ohms for the total loop impedance. Total loop impedance includes power company transformer, all wire, connections, and breakers in the loop where the current flows. Average your two measurements and the result is 1.4 ohms.

By comparison I measure 2/8 = 0.25 ohms at a location in my family room. At my main panel I measure 0.6/8 = 0.075 ohms measured at the panel input terminals (before main fuses). My branch circuit measurement includes the 0.075 ohms plus main fuse, bus bars, breaker to subpanel, #8 copper to subpanel, breaker to final branch circuit, #12 copper to duplex outlet.

To meet you from the other direction with the math: The resistance of #14AWG according to this chart is 2.7 Ohms per 1000' (ambient temp.). I suspect (fear?) close to ~200' run length of branch. So 200' multiplied by 2 conductors gives 400' total length. That gives just over 1 ohm on the branch. And it does not yet include devices, service, utility pot, etc. Is this math OK?

I have to get back in there and test/log thoroughly. I'll post back here in a few weeks with updates.
My concern, if wiring is indeed faulty, is that the builder's electrician and/or AHJ will investigate using the local 'no load' method :jawdrop:

thanks again

 

[K8MHZ]

From my reading of the link to the Florida building code, 'design load' is 180 VA for receptacles. It also says that NEC 220 shall be used for other loads as the 'design load'.

That's 1.5 amps per 'yoke'. What I am not clear about is if you have to add all the receptacles up, or is 180 VA good for the entire circuit, regardless of the number of receptacles.

 

[K8MHZ]

Thanks. This was very helpful:



To meet you from the other direction with the math: The resistance of #14AWG according to this chart is 2.7 Ohms per 1000' (ambient temp.). I suspect (fear?) close to ~200' run length of branch. So 200' multiplied by 2 conductors gives 400' total length. That gives just over 1 ohm on the branch. And it does not yet include devices, service, utility pot, etc. Is this math OK?

I have to get back in there and test/log thoroughly. I'll post back here in a few weeks with updates.
My concern, if wiring is indeed faulty, is that the builder's electrician and/or AHJ will investigate using the local 'no load' method :jawdrop:

thanks again

Knowing the resistance is of no value in figuring voltage drop if you don't know the load. You could have 10 ohms in the conductors and your high impedance DVOM won't be able to see a voltage drop. If using a 100k ohm impedance meter, the meter will be looking at a load of 100,010 ohms instead of 100,000.

 

[gar]

160730-1722 EDT

escopf:

Your math is correct. Your calculated 1 ohm is in the ballpark of my calculation of 1.4 ohms from your measured data. And as you stated the 1 ohm does not include all the other items that are in the actual loop.

I believe that you mentioned in one of your posts that you measure the main panel voltage by picking a circuit that is not the one under test (the test one being the one to which you apply the load), and using this other circuit to see what the voltage change is at the main panel. This is a useful technique so long as there are no changes in load on that main panel voltage measuring circuit. Even if there are changes, so long as they are not too rapid, then you can still get the information you need. This method does not give you the absolute voltage at the main panel unless there is no load on this voltage reference circuit. But, that does not matter because we just need to see the change in voltage at the main panel for a known load current change on the panel.

Measure voltage drop with no load at all?

This makes no sense as most have commented in one way or other. You need a precise definition from the local inspector as how this "no load voltage drop" is to be measured or determined.

A homeowner of a new-construction house (Florida) has been having her TV's power off intermittently, lights dim sometimes etc.

You appear to have good troubleshooting skills. Somehow we need clearer information from the homeowner about what actually happens, and what correlates with the problems of this quote.

.

 

[escopf]

160730-1722 EDT
I believe that you mentioned in one of your posts that you measure the main panel voltage by picking a circuit that is not the one under test (the test one being the one to which you apply the load), and using this other circuit to see what the voltage change is at the main panel.

I use 3 circuits, this way I'm guaranteed to double-up on a phase (in case only one service conductor is acting up)

160730-1722 EDT
You need a precise definition from the local inspector as how this "no load voltage drop" is to be measured or determined.

He gave me one, and it seems to be the same method iwire has seen an inspector use: With no load on the branch, test the panel lugs, then test the farthest point on branch (DMM) and that's your official drop. I'd guess that using this method one might even see voltage gain from time to time (due to cables running next to each other).

Still, I'll have to go back and take better measurements. Will post back here.

thanks