Hey there,

Just thought you'd enjoy hearing about a trouble call I went on this weekend...

A client called me Friday afternoon, frantic. Their computer/cash register in one location keeps crashing, and could it be the circuit that it's powered by? BTW, I installed that circuit a couple of months ago. They need to get the PC working ASAP. If it's not fixable, could I recommend a UPS/power conditioner unit?

Background--the customer is setting up a retail operation in converted industrial space. I installed a 10kVA dry xformer to take 480 to 240/120, and bonded neutral/center tap of xfrmer to earth ground. Installed a loadcenter and about 150' of EMT. Ran two 20A circuits, one for computer and one for convenience outlets. Separate neutrals. Ran isolated ground for computer ckt.

I get on site and see that there's an extension cord plugged into the outlet I installed. It's a nice heavy cord, looks like 12 gage, and it disappears under the bottom shelf of the counter on which the computer is sitting. The PC is about 10 feet from the outlet. A surge protector/power strip is plugged into the far end of the extension cord, and the PC and some other stuff is plugged into the outlet strip. The power strip looks pretty beat up.

I whip out the trusty Fluke T-5 to check voltage. A-OK at the outlet, 119V, ditto at the strip. I jiggle the plugs and the power cords and flip the power strip switch thinking that maybe it's an intermittant-- bad plug or dirty switch. Everything seems OK--there's no loss of power/dropouts that I can see on the tester.

The client says he's had a problem with the computer since he set it up at that location. Maybe the power is dirty; we're in an industrial area and I'm wondering what's going on in the factory next door, which is fed from the same customer transformer and main switchgear.

I get the scopemeter (Fluke 123) and check power at the outlet. Perfect sinewave, clean as a whistle. I go to the power strip and get a "fat" sine wave--the trace is a sine wave, but kind of thick and fuzzy. Hmmm, what the heck is this?

I unplug the power strip and insert the probes into the end of the extension cord. Same "fat" sinewave. I'm thinking it's something to do with the extension cord, and maybe a new cord will solve the problem.

So I unplug the cord from the outlet and figure I'll look it over just to see what I can see. I'll bet that some of you know exactly what the problem was!

The extention cord was a nice one--12 gage, romex brand, AND IT'S 100 FEET LONG! All coiled up under the bottom shelf of the counter...one big choke!

I suggested that the client get power strip with a 10' cord.

I'm always amazed at how little people know about electricity. Alomg these same lines, a friend of mine was a sparky at a nearby AF base said that some maintenance men once needed a long extension cord so some Einstein decided that a 250' coil of romex would work great. They put cord caps on the ends of a 250' roll and used (uncoiled) about half of the roll to get where they needed to go. They then started using the big drill motor or whatever they were using.

Imagine their suprise when the coil of romex burst into flame. The coil was acting as a nice choke and the heat generated led to a short circuit, with all the attendant fireworks.

AT any rate, I wish they were all this easy!

Cliff

LOL!Cliff, that usually makes a 'statement'.


BTW, i read an article, some time ago, that actually recommended coiling extention cords serving sensitive equipment. I believe that it was refered to as a 'Belden Wrap', i wish i could pick it out in this mountain of trade mags.......

Cliff:
In your case, it's easy if you have good troubleshooting skills, as well as the proper testing equipment at hand. Excellent call!

Mike (mamills)

I didn't think that "excessive" heat was produced by a coil of wire unless there was some ferrous metal inside the coil. This was the cause of the McCormick place fire in Jan of '67 or '68. The National Housewares show was there and a large SO cord with a 40 or 50 amp load was in use. It was too long and the excess cord was coiled up and dropped over the 411 box on the end of the cord. The box was heated to the point that packaging material ignited and the place burnt down. It is right on the lake, but the fire department couldn't get water because of frozen hydrants. It went to a 5-11 and 5 special alarms. I don't think Chicago has ever sounded that many alarms for a single fire before or since, with of course the exception of the "Chicago Fire".
Don(resqcapt19)

Amp-Man,
Did the outlet strip with the 10Ft cord fix the problem?
Was the fat part of the 60 Hz sine wave some sort of high frequency interference picked up by the 100Ft cord?If so were you able to locate the source of the interference?
Just curious.

Chris

A couple of points strike me about excessive heat caused by coiling up an extension cord.

First, both the line and neutral wires are parallel throughout and the current is flowing in opposite directions at any moment in time. The magnetic fields would tend to cancel and the inductance would be very low.

Second, even if that were not the case, remember that inductive current is wattless -- it does not generate any heat. The heat in the cables is a function only of the resistance of the copper and the current, i.e. (I^2)*R.

The only way I can see for the current to increase drastically would be if the inductance introduced happened to match any capacitance on the load to form a tuned-circuit series resonant at 60Hz. And that would need some large values of L and C.

I think it's more a case that an enclosed coiled cord fully loaded just can't dissipate the power caused by normal resistive heating fast enough.

The resonance point may have a bearing on the high-frequency noise, though. The small amount of inductance coupled with stray capacitance could easily set up a tuned circuit, and if it happens to be resonant at a frequency on which a lot of noise is present, you've effectively got a very efficient antenna for that noise.

Paul:
I just tried an experiment wiht 250 feet of 12-2 still coiled up. Put a 10 amp load on it and the coil began to heat up. So far so good your statements hold up. the voltage drop was about 10 volts.
Everything was as you say, until I put the gauss meter on top of it and it went nutz.
Radially it read about 150mg but when set on top of it about a foot from the top of the coil " out of range"
The heating I assume was due to the lenght of wire in the coil 250 feet at 10 amps, but the size of the magnetic field was unbelievable. The load was strictly resistive
all 100 watt bulbs.
This was not any double blind study just hooking some wire to see what would happen, but the magnetic field surprised me.

Anyone can jump in, because what Paul said makes sense but I am not sure.
-Mark-

Mark,
Very interesting. I guess with the coil you made an electro-magnet, after all windind wire around something is how you make one.

On Voltage Drop. I went on to a job site one day and I saw one of my helpers using a Hole Hawg that sounded real low and was just barely making it through the wood. He said the drill was bad and it was going to take forever to finish. Then I noticed that he had 2 100 ft 16 gauge cords (left from HVAC crew) pluged into a 50 ft 12 gauge off the temp pole. I had him take off the 200ft of 16 and add 100 ft of 10 and amazingly the Hole Hawg worked perfectly. Imagine that.

Interesting experiment.

Thinking about it some more, I guess I would expect some inductance, because the line/neutral fields aren't going to cancel out precisely due to the wires not being perfectly uniform throughout the whole 250' coil.

I suppose it comes down to a question of just how much inductance it is possible to get in this way.

How about trying it again just taking the hot wire through one wire of the Romex and wiring the neutral straight to the lamp?

Paul;
I will try that and let you know the results... By the way your experiment with the tap-water sounded interesting, but any time water is basic/acidic it will conduct as you mentioned, and I believe highly mineralized water is basic ?

-Mark-

Paul:
Here are the results of my un-scientific scientific test, test conditions 250 foot roll of 12-2 using only the black conductor and a short piece of white to complete the circuit.
The load purely resistive 9.9 amps @120 volts.
Empirical data: no physical signs of heating not that my hand could detect.
Voltage drop was 8 volts this was an improvement from yesterdays test.
Now heres the part you have been waiting for, Out of range, on the gauss meter at 18 inches and down to 159 at 1 meter. These readings were taken on top of the coil on the side of the and from underneath the workbench a spherical cube if you will the reading were the same.
The voltage drop difference because I was using half the wire as yesterday but the magnitude of the gauss reading is not inconsistant with what I have been reading from the manual from the ' Magnetic Sciences International', this is what they claim for readings of Knob and Tube wiring that is seperated by the width of 2 two by fours, or joists. In fact from 6 meters away I was reading 6mg, which I thought was incredible, since they claim 0.0-0.3 to be the ambient Normal range. Outside near the power lines in the street I was only getting 1.7mg.
Conclusion ???
Hey I only did the tests, best guess is these people from the EMF and Mag. Sciences are definitely on to something.
I originally bought the gauss meter from Mike Holts Newsletter, and have learned alot about how just a shorted neutral can cause a Net current on the water line and a disturbing magnetic effect. Its really quite interesting.
So what do you think , Paul ?

Electric Eagle:
When I plugged in that whole roll of wire I was really surprised how fast it heated up, but when thought about the overall length is 500 feet but surprisingly too was it only dropped 10 volts at a 10 amp load.
Can imagine your surprise when seeing the 'Hole-Hog' barely limping along wiht 400 feet of extension cord. I would have been quite P*****.

-Mark-

Mark,
I was a little ticked off, mainly because it was an almost new Hole Hawg, but the guy was new and inexperienced. I just explained what was happening and made it clear to not just ingore it when something doesn't seem right, fix the problem.

[This message has been edited by Electric Eagle (edited 05-22-2002).]

Motor-T,
I wonder what the Gauss meter would show with no load/current.My educated guess would be that the reading would be very small.
This would support that these non-contact detectors are electric field detectors and don't rely on the presence of a magnetic field(current flow) for voltage detection.
Chris

Motor-T, you da man!

*explosions and shouts of "Eureka" imminate from motor-T's garage...*

I'd certainly expect a reasonable magnetic field with just the one conductor in use from the coil of cable.

How about a third stage to the experiment? What I have in mind is using both conductors but arranged so that the currents are flowing the same direction (e.g. load connected between the black from one end and the white from the other, power on to the remaining two opposite ends).

That way we'd have the same resistance as in the original experiment, and the same heating effect from I^2*R in the coil. With an effective 500' coiled up, you should see an increase in the magnetic field strength.

Paul:
Why not, ' i'll have a go at it', I'll let you know what happens.

-Mark-

Et al:
Ok I had my go at it, this test was the third of an infinite series of un-scientific-scientific test.
The roll of 12-2 was set up such that the current was all in the same direction. as usual the 10 amp load. 120 volts.
Results:
This is unbelievable, the coil was vibrating, I kid you not,(little Jack Paar lingo), The bench was vibrating, the Gaussmeter went bazerk. It was reading 17.5mg from at 6 meters, the voltage drop was 13 volts or 107 at the terminals.
I think for my next test i will get 250 volt incandescents and try 250 volt source.
Now when I say the coil was vibrating Not so much visual but you could feel it with the hand and the bench too, should get great results with 250 volts. Eh ?

-Mark-

Motor-T,

Wood bench or metal bench ?

Wow...

*...on edge of seat...*

Wow is right, Virgil! Some experimentalists that would make Edison and Tesla proud!!

As to whether getting rid of the long extension cord took care of the problem, I haven't heard back that it didn't. I even called the client and left a voicemail message--"Everything going all right?", and haven't heard back.

No news is good news...

Cliff

I had a feeling the field strength would increase a little! By the time you reach experiment #294 it'll be 13.8kV on 2000' coils and magnetic objects levitating all over town!

Cliff,
As a matter of interest, did you note the frequency of the interference you were picking up on that 100' extension?

Amp-Man, a few questions;

<OL TYPE=1>


[*] Is the PC crashing to a "Blue Screen" (AKA The Blue Screen Of Death)?,


[*] If Blue Screens, does this bring up an Exeption OE within NDIS and NWDIR, or similar NIC responsive Kernals / DLLs?,


[*] Do the blue screens become less frequent when either the monitor is powered off for a while, or when the TVSS strip is not used?

</OL>

I have serviced this type of problem a few times and found it to be faulty MOVs in TVSS strips and failing power supplies in the Monitor.
These hard errors were results of randomly occuring drops in line voltage and current, which probably do not go more than 4 cycles per drop. Just enough to create havoc on the motherboard, but not enough to bring the PC power supply into "Low Power" state.

I have only seen a handful of Low Power situations, and these commonly result in a "Warm Boot" restart, rather than any types of Protection Faults or Blue Screens.
These normally restart the machine and lock it into error status at the POST on the more expensive machines. Cheap-o's will simply reboot to the Operating System once the "Power Good" signal is sent from the Power Supply to the CPU [taking it out of restart loop].

I like your testing and results! Keep up the work!

Paul stole my ideas! <joking>

I could see the mG levels being through the roof with the first two setups! Can't figure why they are so high when the circuit is connected as normal - especially when driving a Linear load!
Something with Harmonics would be a No-Brainer to figure odd mG readings, or maybe I am not thinking correctly here!

Also curious about the test bench - is it Ferrous metal [Iron, Nickel or Cobalt based]? Something that would cause an Inductive Effect to the coiled cables.

As you know, if the coil was wound with two wires carrying currents of equal intensity in opposite directions, this will [somewhat] nullify the Magnetic Field created by Inductance. True for AC and DC.
If there is a difference in Intensity, this will create a significant Magnetic Flux. If this current was AC, this would Induce currents into nearby conductors [basis of GFCI's]. If there is some method of "Harnessing" the Magnetic Flux, this would result in XL influencing the circuit's Currents, resulting in a noticable Voltage Difference on the "Load" side of this XL.

Paul hit the nail on the head when mentioning the heat build up. This is the result of Series Resistance in the Conductors, so there is a transfer of True Power at these points - we feel the result of the True Power transfer as Heat [or smell it as smoke ].

Poor Power Factor [50% or less] will draw an excessive level of currents, which overloads or "Saturates" the conductor materials with too much current. In this case, this can result in Heat and corresponding Voltage Drop, due to the increased currents drawn through a fixed [and increasing with heat] Resistance.

Enough of the babble!

Good luck with the T Shooting!

Scott S.E.T.

I thought that meant that Windows was operating normally

Sorry, couldn't resist that one...

Scott35-- I'll bet BSODs would be history if you just got a new Wallmart Super-Surge plug strip. They actually had some that were marked 'Internet Ready' and 'Fully Y2K Compliant'!

I recall hearing about someplace that was advertising a "Y2K compliant" screwdriver a while back!

Although I'm going off on a slight tangent here, the interference problem with sensitive equipment is considered a big problem in some telecoms places.

At the Goonhilly Satellite Station we had a site 11kV primary with back-up generators feeding xfmrs for normal 240/415V 3-phase to each building. But the xfmr banks and distribution gear were split to provide "clean" supplies for comms equipment and "dirty" supplies for everything else.

To answer the workbench question the bench is made of wood, nothing magnetic about it.
But I just got a new meter true rms, crest etc. and a huge choke. Adjustable from 1-130 volts. Will be posting the results as they come in.
Somebody asked if there was a gauss reading with no load small but it was still there and yes even without a load the non-contact tester still indicated voltage present.
As far as the field goes with the normal currrent orientation the 250 foot coil is exactly that it was a coil, and created a good size magnetic field. smaller somewhat when the flow was in the same direction, but still very noticeable.

-Mark-

Mark,
Thanks for all your efforts and reporting the test results.I believe it was very interesting and educational.

Chris

<quote>

1. Is the PC crashing to a "Blue Screen" (AKA the Blue Screen of Death)?,


I thought that meant that Windows was operating normally

Sorry, couldn't resist that one...
<quote>

LOL!!! Seems like most MS Operating Systems need to Blue Screen once things become stable
And this from a die-hard MS O/S user! [me..]

<quote>
Scott35-- I'll bet BSODs would be history if you just got a new Wallmart Super-Surge plug strip. They actually had some
that were marked 'Internet Ready' and 'Fully Y2K Compliant'!
<quote>

ROTFLOL!!!
I've got to get one! At least the box, or sticker with the very informative text!

No ends of insanity with that Y2K rollover craze!

I need a Plugstrip which works with HTML stuff and accesses CGI's without problems! Does it work well with Netscape?

Thanks to Paul and Bjarney for the laughs!

Scott S.E.T.

paul uk
RE;
Thinking about it some more, I guess I would expect some inductance, because the line/neutral fields aren't going to cancel out precisely due to the wires not being perfectly uniform throughout the whole 250' coil.

I suppose it comes down to a question of just how much inductance it is possible to get in this way.

Why are the wires not uniform throughout the 250 coil when it is presumably manufactured double insulated pvc flex type cable?

If the cable was to have some kind of inductance would this not leave the cpc susceptible to an induced voltage creating leakage currents to earth?

dont mean to sound nasty just intrested to know thats all!!

on 40 metre extension cords, printed on the real is 4 amps unwound and 13 wound.

ps i like the experiments

James;


I guess the best explanation I could give is the cable is not designed to be something "Precise" - like the difference between Plain-'Ol Multiwire Thermostat cable (18/6) and CAT-5e UTP cable.

Scott35

p.s. glad you brought this thread back to life! Very interresting stuff here!

I wanna check out Motor-T's Garage! Never get us out of there!


Any invitations, Mark?

S.E.T.

Yes, this has certainly revived an old thread. Everyone remember the water experiment as well?

James,
You raise a good question there. Even if the spacing between conductors were exactly uniform throughout the cable, don't forget that we're talking about a coil of Romex in this case.

Once you wind the cable in this way you get not only the interaction of fields between the two conductors at any point along the run, but also between conductors in nearby turns of the coil. Thus you might have hot & neutral next to each other between adjacent turns at one point, but two hots or two neutrals next to each other where the winding direction changes.

The fields won't cancel out completely, so there will be a little measurable inductance.

Regarding coupling to the CPC (*), there will always be a little leakage anyway where we have one circuit conductor grounded and the other hot, due to capacitance in the cable.

(*) The term CPC that James used means Circuit Protective Conductor, equivalent to the EGC in America.



[This message has been edited by pauluk (edited 02-18-2004).]