Residential Voltage drops on circuits

[Primitivelamps]

Just purchased an AMPROBE INSP-3 wiring inspection tester. Does anyone have experience with this tester, good or bad? Concerned about accuracy. The first house I used it on was a 1947 cape. The up stairs outlets showed Voltage drops or 65.4% and 75.3% respectively (by the way the tenant mentioned that if they plug two or more things into the outlets they trip which would make sense to me with these numbers). Am I right the NEC recommends it be under 5%. The instructional manual that came with the unit list the following as possible problems creating excessive voltage drop:

Bad Splice
Loose screw Termination
Stripped thread on wire nuts or terminal screws
Faulty outlet
undersized wiring(too long of run for wire size)
loose connection at circuit breaker or fuse
corroded connections
overheating due to loose contacts
faulty or poor quality push in connections on receptacles
cold forming on aluminum wiring, causing loose connections
worn switch or circuit breaker contacts

I am a home inspector and would be recommending a licensed electrician to further investigate regardless because this house had other issues. But in the opinion of a licensed electrician(you guys) I guess I am wondering if this tool is accurate enough and what would be the most reasonable number to recommend follow up (anything over 5%?). Any advice is appreciated.

Thanks Terry

 

[Dennis Alwon]

The 5% is correct but it is not enforceable however the problem does need to be addressed.

 

[kwired]

How much test load does this thing introduce to the circuit. If you have drop at the levels you mention, you have a serious problem somewhere.

75% drop on 120 volt means that it is dropping to 30 volts, unlikely to be from too small of a conductor, and very likely to be a weak connection someplace.

 

[Primitivelamps]

You tube of the tester

You tube of the tester

http://www.youtube.com/watch?v=QQesUJEeOdU

 

[infinity]

I am a home inspector and would be recommending a licensed electrician to further investigate regardless because this house had other issues. But in the opinion of a licensed electrician(you guys) I guess I am wondering if this tool is accurate enough and what would be the most reasonable number to recommend follow up (anything over 5%?). Any advice is appreciated.

Thanks Terry

You need to determine at what point is the VD is so high that it's likely caused by one of the hazardous conditions in your list and then report it. A marginally greater than 5% VD in a house with long branch circuit runs probably isn't really a reportable issue.

 

[kwired]

http://www.youtube.com/watch?v=QQesUJEeOdU

Sounds like a salesman that doesn't really know everything about what he is selling. (surprise)

He used wrong terminology in some areas, and claims code requires less than 5% voltage drop. Code does mention 5% in an informational note, but it is not a requirement. He also claimed it takes 40mA to stop the heart, I always understood it to be much closer to the 4-6 mA range therefore the reason GFCI's are set to trip at that range.

 

[GoldDigger]

He also claimed it takes 40mA to stop the heart, I always understood it to be much closer to the 4-6 mA range therefore the reason GFCI's are set to trip at that range.

One difference is that some numbers are whole body current with the current path crossing the area of the heart (arm to arm, arm to leg, etc.) while the smaller numbers are for current flowing directly across the heart rather than through parallel paths.
5 ma seems to be an accepted typical threshold for actually feeling a shock.

 

[Primitivelamps]

How much test load does this thing introduce to the circuit. If you have drop at the levels you mention, you have a serious problem somewhere.

75% drop on 120 volt means that it is dropping to 30 volts, unlikely to be from too small of a conductor, and very likely to be a weak connection someplace.

My thoughts as well. Unfortunately we get sued if we point something out and when we don't point something out. This is my dilemma. What is a safe number to suggest further investigation by a licensed electrician. Based on the NEC recommendation I guess anything over 5 % correct, or do I let 6 or 7 go. Remember I have to cover my A##.

 

[ptonsparky]

Just make the report. 3% could be to high under the right conditions. State the 5% and your source for that information if you wish, but other than that no other comment should be required. Your client would have to be pretty slow on the trigger to not figure out 75% VD is a problem.

 

[suemarkp]

One of the features of the tool is that it can separately tell you the voltage drop in the hot and the voltage drop in the neutral if a grounding wire is present. Generally, the voltage drop should be equal on those two (because both are about equal in length). If one is significantly larger than the other, that would be an indication of a poor splice or connection. A multiwire branch circuit could also change that (reducing the drop in the neutral), but it should only be a few volts reduction in a multiwire circuit. Not sure if the tool can compensate for that or not.

Overall voltage drop is hard to say what it should be -- depends on circuit length, and in most houses you'll have no idea what path the wires took. If you have a TDR to measure wire length, that could provide a guesstimate. But dropping much more than 6V @ 15A (about 5%) is a fair amount of heat to dissipate in a circuit, especially a short (50' or less) one.

This tool appears to be compensating for other upstream voltage losses from other existing laods (service/feeder conductors) as long as those loads persist through its test pulse. The NEC recommendation of 5% includes overall feeder (and perhaps Service) and branch circuit drop, so if this tool says 5% you've got a lot of drop. But as others have said, there is no NEC mandated voltage drop requirement (only a fine print note recommendation).

 

[Riograndeelectric]

I would be interested in other members input on the Amprobe circuit tester.
I am looking to buy a new circuit analyzer.
I have a older model Ideal sure test and comparing the different manufacturers

 

[kwired]

My thoughts as well. Unfortunately we get sued if we point something out and when we don't point something out. This is my dilemma. What is a safe number to suggest further investigation by a licensed electrician. Based on the NEC recommendation I guess anything over 5 % correct, or do I let 6 or 7 go. Remember I have to cover my A##.

Just make the report. 3% could be to high under the right conditions. State the 5% and your source for that information if you wish, but other than that no other comment should be required. Your client would have to be pretty slow on the trigger to not figure out 75% VD is a problem.

Exactly. Are you inspecting or troubleshooting?

You could test it one day and tell them it passes your testing, the next day a connection could fail, do you want to assume liability for that connection failure that you said was not there? Such tests only represent conditions at the time of the test.

Maybe two days later they start up a heavy load that wasn't running when you were there and it drops the voltage to levels that you would not have passed.

I find these type of tools to be useful for finding problems, but not they have limitations. Voltage at an outlet is not only dependent on conditions of the branch circuit, but also feeders, and service ahead of the branch circuit. Additional loads on other circuits will change the service and feeder properties.

 

[hardworkingstiff]

My thoughts as well. Unfortunately we get sued if we point something out and when we don't point something out. This is my dilemma. What is a safe number to suggest further investigation by a licensed electrician. Based on the NEC recommendation I guess anything over 5 % correct, or do I let 6 or 7 go. Remember I have to cover my A##.

If you are concerned about covering yourself, maybe an attorney should answer your question.

If it was me, I'd just report the equipment I used to test, the test results, reference the NEC recommendations, note it's a recommendation and not a requirement, and move on to the next inspection.

 

[Primitivelamps]

What would cause the loaded voltage to be so high on a normal outlet? I understand it fluctuates but 143?

One of the features of the tool is that it can separately tell you the voltage drop in the hot and the voltage drop in the neutral if a grounding wire is present. Generally, the voltage drop should be equal on those two (because both are about equal in length). If one is significantly larger than the other, that would be an indication of a poor splice or connection. A multiwire branch circuit could also change that (reducing the drop in the neutral), but it should only be a few volts reduction in a multiwire circuit. Not sure if the tool can compensate for that or not.

Overall voltage drop is hard to say what it should be -- depends on circuit length, and in most houses you'll have no idea what path the wires took. If you have a TDR to measure wire length, that could provide a guesstimate. But dropping much more than 6V @ 15A (about 5%) is a fair amount of heat to dissipate in a circuit, especially a short (50' or less) one.

This tool appears to be compensating for other upstream voltage losses from other existing laods (service/feeder conductors) as long as those loads persist through its test pulse. The NEC recommendation of 5% includes overall feeder (and perhaps Service) and branch circuit drop, so if this tool says 5% you've got a lot of drop. But as others have said, there is no NEC mandated voltage drop requirement (only a fine print note recommendation).

 

[GoldDigger]

View attachment 7909
What would cause the loaded voltage to be so high on a normal outlet? I understand it fluctuates but 143?

You have an anomalous situation that the tester is unable to handle. The measured voltage with no load is 120, which is fine. The tester does not actually put on a high current load, but instead extrapolates from the voltage difference seen when a low current test load is applied.

You will not see separate line and neutral drops because no ground reference is available. If you were to attach a separate ground wire to a grounded point you would see the two readings.

But the interesting thing is that the tester is convinced that applying a load is causing the voltage to INCREASE instead of decrease. That is why the calculated "loaded voltage" reading is higher than the no load reading of 120.3 VRMS.

I do not have any good ideas what could cause the meter to behave in this way, but I would definitely get an electrician with an actual high power load to check the circuit out. Or have my meter repaired.

Take a look at the manual, here.

Voltage under load: This is the actual voltage
available if a steady load of 10, 15 or 20 amps is
applied to the line.

 

[Primitivelamps]

You have an anomalous situation that the tester is unable to handle. The measured voltage with no load is 120, which is fine. The tester does not actually put on a high current load, but instead extrapolates from the voltage difference seen when a low current test load is applied.

You will not see separate line and neutral drops because no ground reference is available. If you were to attach a separate ground wire to a grounded point you would see the two readings.

But the interesting thing is that the tester is convinced that applying a load is causing the voltage to INCREASE instead of decrease. That is why the calculated "loaded voltage" reading is higher than the no load reading of 120.3 VRMS.

I do not have any good ideas what could cause the meter to behave in this way, but I would definitely get an electrician with an actual high power load to check the circuit out. Or have my meter repaired.

Take a look at the manual, here.

Thanks for the feedback. As I mentioned this tester is brand new to me. The interesting part is only the upstairs outlets tested like this in this 1940's cape. The other outlets in the home tested fine using the meter. I have used the unit on my own 200 year old colonial and all was well except a couple outlets in one room that showed a 13 % voltage drop. Not a surprises because I think there are some poor splices in the attic and this is the farthest room from the panel. I just could not wrap my head around the number being higher with a supposed large VD.

 

[GoldDigger]

I just could not wrap my head around the number being higher with a supposed large VD.

Yes, there is no way that a measured loaded voltage could end up higher than the measured no-load voltage. (With a high resistance neutral in an MWBC you could get an increase of the voltage on the OTHER side of the MWBC, but never on the side you are measuring.)

But this is a calculated voltage, and so indicates to me that there may be some sort of highly reactive or non-linear load plugged into the circuit somewhere that, perhaps in combination with one or more high resistance connections, is causing the meter to calculate incorrectly.

Just plugging in a 100 watt lamp and measuring the voltage will give you an idea of what the real voltage drop is. And just using an external ground reference to measure the line and neutral voltages under test separately would give you an idea of which side of the circuit the hypothetical bad splice is in.

Finally, plugging two or more loads into the circuit would not cause the breaker to trip based on overcurrent just because the circuit resistance is high. The only way the bad circuit could cause a higher current would be if the voltage drop affected a motor or other non-resistive load.

This really is a puzzler.

 

[Primitivelamps]

Yes, there is no way that a measured loaded voltage could end up higher than the measured no-load voltage. (With a high resistance neutral in an MWBC you could get an increase of the voltage on the OTHER side of the MWBC, but never on the side you are measuring.)

But this is a calculated voltage, and so indicates to me that there may be some sort of highly reactive or non-linear load plugged into the circuit somewhere that, perhaps in combination with one or more high resistance connections, is causing the meter to calculate incorrectly.

Just plugging in a 100 watt lamp and measuring the voltage will give you an idea of what the real voltage drop is. And just using an external ground reference to measure the line and neutral voltages under test separately would give you an idea of which side of the circuit the hypothetical bad splice is in.

Finally, plugging two or more loads into the circuit would not cause the breaker to trip based on overcurrent just because the circuit resistance is high. The only way the bad circuit could cause a higher current would be if the voltage drop affected a motor or other non-resistive load.

This really is a puzzler.

I called AMPROBE and sent them the picture. The tech on the phone could not explain and was contacting an "engineer" or someone else on staff. They said they would call me back today or tomorrow at the latest. I'll let you know what they say.

Thanks

 

[kwired]

Yes, there is no way that a measured loaded voltage could end up higher than the measured no-load voltage. (With a high resistance neutral in an MWBC you could get an increase of the voltage on the OTHER side of the MWBC, but never on the side you are measuring.)

But this is a calculated voltage, and so indicates to me that there may be some sort of highly reactive or non-linear load plugged into the circuit somewhere that, perhaps in combination with one or more high resistance connections, is causing the meter to calculate incorrectly.

Just plugging in a 100 watt lamp and measuring the voltage will give you an idea of what the real voltage drop is. And just using an external ground reference to measure the line and neutral voltages under test separately would give you an idea of which side of the circuit the hypothetical bad splice is in.

Finally, plugging two or more loads into the circuit would not cause the breaker to trip based on overcurrent just because the circuit resistance is high. The only way the bad circuit could cause a higher current would be if the voltage drop affected a motor or other non-resistive load.

This really is a puzzler.

I think the first thing to check with the raise in voltage is bad neutral. Just because this device does not actually load the circuit doesn't mean condition of the neutral will not effect the reading. Why you think it is impossible for the tested side of a MWBC to increase in voltage? If that is the side that has higher impedance then it will have more voltage drop across it. The actual voltage of each side of the MWBC with a bad neutral will entirely depend on impedance of loads that are in series. Since this tester does not actually put a 10,15, or 20 amp test load in the circuit and instead puts a much smaller load and calculates what would happen with a larger load, it is a much higher impedance than if you were to place a 10-20 amp test load in the circuit, and if the other side of the MWBC is a lower impedance by nature it should result in voltage rising on the tested side of the circuit.

ADD: this would not have to be a MWBC neutral that failed, could be feeder or service neutral also.

 

[GoldDigger]

I think the first thing to check with the raise in voltage is bad neutral. Just because this device does not actually load the circuit doesn't mean condition of the neutral will not effect the reading. Why you think it is impossible for the tested side of a MWBC to increase in voltage? If that is the side that has higher impedance then it will have more voltage drop across it. The actual voltage of each side of the MWBC with a bad neutral will entirely depend on impedance of loads that are in series. Since this tester does not actually put a 10,15, or 20 amp test load in the circuit and instead puts a much smaller load and calculates what would happen with a larger load, it is a much higher impedance than if you were to place a 10-20 amp test load in the circuit, and if the other side of the MWBC is a lower impedance by nature it should result in voltage rising on the tested side of the circuit.

ADD: this would not have to be a MWBC neutral that failed, could be feeder or service neutral also.

With all due respect to a Senior Member, what you are proposing violated Kirchoff's Laws and several other basic principles of "equivalent circuits". Regardless of the complexity of the circuitry on the other side of the receptacle, it can be modelled as a voltage source and a series resistance. The only ways that the voltage at the interface could go up when the current draw increased would be:
1. The equivalent circuit has a negative resistance, or
2. The act of drawing the current caused some other device to switch on or off, thus changing the equivalent circuit.

The voltage between neutral and ground could certainly rise, but that would reduce the measured line to neutral voltage, not increase it! And the line to ground voltage could not increase at all.

One thing that occurred to me was that if the tester applies a current pulse rather than a linear load, the presence of a resonant circuit on the other side of the receptacle somewhere could cause ringing which could be misinterpreted as an increase in voltage.