ProSoundWeb Community
Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => AC Power and Grounding => Topic started by: James A. Griffin on May 07, 2014, 09:06:03 PM
-
This is odd.
I've been in my office for almost 10 years, rocking along with no power issues at all. This morning when I came in, everything plugged into one of the AC circuits in one room was off. I checked all the outlets on that circuit and find that Hot and Ground are reversed on all of them. Every other circuit in the office is fine.
To get back going, i just ran extension cables from known good outlets / circuits to power everything I need turned on. But what the hell happened after 10 years of correct wiring?
Wiring melted together somewhere? I plugged something in that has a short and that went back to every outlet on the circuit??
-
This is odd.
I've been in my office for almost 10 years, rocking along with no power issues at all. This morning when I came in, everything plugged into one of the AC circuits in one room was off. I checked all the outlets on that circuit and find that Hot and Ground are reversed on all of them. Every other circuit in the office is fine.
To get back going, i just ran extension cables from known good outlets / circuits to power everything I need turned on. But what the hell happened after 10 years of correct wiring?
Wiring melted together somewhere? I plugged something in that has a short and that went back to every outlet on the circuit??
James, how did you determine the hot and ground are reversed?
-
James, how did you determine the hot and ground are reversed?
I used the handy 3-light gadget....
-
Electricity is magic so that can happen if somebody casts a spell on you. ;D ;D
JR
-
I used the handy 3-light gadget....
They often lie, a lot.... Read this http://ecmweb.com/contractor/failures-outlet-testing-exposed (http://ecmweb.com/contractor/failures-outlet-testing-exposed)
You need to get a meter between a known ground and each contact in the outlet. Also, a $20 NCVT will quickly tell you if your ground is indeed hot. I'll bet you have a ground opened up somewhere. Not great, but not an immediate killer.
-
Electricity is magic so that can happen if somebody casts a spell on you. ;D ;D
JR
Or yes, it COULD be magic. :o
-
You need to get a meter between a known ground and each contact in the outlet. Also, a $20 NCVT will quickly tell you if your ground is indeed hot. I'll bet you have a ground opened up somewhere. Not great, but not an immediate killer.
+1
Except that an open ground would not keep things from powering up.
My first suspect in this situation is one of the first outlets in the circuit-especially if they have been wired with the push in sockets-they are notorious for melting down. IMO a circuit that quits working for no apparent reason is a huge fire hazard. Consider that if there is power at the breaker and none down the line there is a very good possibility that the problem is a bad connection rather than an open connection-a bad connection dropping 120 volts with even 1 amp of draw equals 120 watts which will heat up a device or j box pretty quickly.
If the neutral is open and something beyond the open is plugged in and powered up, both neutral and hot will appear to be hot. Who knows what the outlet tester will claim?
-
Or yes, it COULD be magic. :o
I keep going out and cutting those evil wires out on that branchless tree out front, that seem to convey the devil's handiwork to my home... but somehow (IT'S MAGIC!) they keep re-connecting after a few days. :o
-Ray
-
If the neutral is open and something beyond the open is plugged in and powered up, both neutral and hot will appear to be hot. Who knows what the outlet tester will claim?
Here's a schematic. Let's trace it out.
-
Tracing it out is easy-just not in my head after spending the afternoon in a 110 deg greenhouse.
Normally 1 and 3 should be lit.
Open neutral only 3 would be lit.
If a load is plugged in downstream of the open in the neutral then 2 and 3 would be lit, which is the same result as if you reversed the hot and ground.
Again, my first suspicion-partly because of the ease of checking for it, partly from experience would be push in wiring in receptacles.
-
Tracing it out is easy-just not in my head after spending the afternoon in a 110 deg greenhouse.
Normally 1 and 3 should be lit.
Open neutral only 3 would be lit.
If a load is plugged in downstream of the open in the neutral then 2 and 3 would be lit, which is the same result as if you reversed the hot and ground.
Again, my first suspicion-partly because of the ease of checking for it, partly from experience would be push in wiring in receptacles.
One of the projects I've considered is a master "goof board" with every variation of mis-wired and open-wire recept/outlets possible. Then I could plug a 3-light outlet tester from recept to recept and demonstrate the light patterns. Or I could build a matrix with 3 x 3 relays that would use one duplex outlet that would be programmable for every possible wiring failure and mis-connection. I've got a little Rasberry PI by kids gave me for Christmas just begging for a project, so maybe that would be fun. Gotta remember to "break" the relays before the "make" since it could easily cross-connect the incoming wires if I'm not careful with the programming. Yeah, I'll put a series limiting resistor on each leg in case something goes wrong in my logic. I admit to being a hack at programming this stuff, so I need a backup plan.
-
Or I could build a matrix with 3 x 3 relays that would use one duplex outlet that would be programmable for every possible wiring failure and mis-connection. I've got a little Rasberry PI by kids gave me for Christmas just begging for a project, so maybe that would be fun. Gotta remember to "break" the relays before the "make" since it could easily cross-connect the incoming wires if I'm not careful with the programming. Yeah, I'll put a series limiting resistor on each leg in case something goes wrong in my logic. I admit to being a hack at programming this stuff, so I need a backup plan.
No way I would use a Pi for this application. The Pi is like a late 90's computer with a modern low end video card. Way over kill and complex for the task at hand. If it MUST be chip and relay driven, with minimal programming knowledge, then use an arduino variant.
-
No way I would use a Pi for this application. The Pi is like a late 90's computer with a modern low end video card. Way over kill and complex for the task at hand. If it MUST be chip and relay driven, with minimal programming knowledge, then use an arduino variant.
Very interesting. A quick search found an 8-channel Arduino output interface with 10-amp relays rated for 250 VAC. Cost is $19.59 with free shipping to the USA. I can't buy just the relays for that kinda money. http://www.dx.com/p/arduino-8-channel-12v-relay-module-expansion-board-142797#.U2ywpijVvpc (http://www.dx.com/p/arduino-8-channel-12v-relay-module-expansion-board-142797#.U2ywpijVvpc)
Gosh, I think this controller needs 9 relays for the outlet goof matrix, so perhaps I should do a Karnaugh Map and try the simplify the circuit a bit. I used to do Karnaugh Maps all the time in the 70's while designing relay logic for packaging machines. It's been a long time, but I'm betting it still works...
-
A few knife switches and fuses.... (http://www.enasco.com/prod/images/products/85/AC098376l.jpg)
Programming an embedded controller only makes sense if it needs to be automated or make decisions based on reading conditions.
JR
-
A few knife switches and fuses.... (http://www.enasco.com/prod/images/products/85/AC098376l.jpg)
Programming an embedded controller only makes sense if it needs to be automated or make decisions based on reading conditions.
JR
Half of the world's new inventions are made because they're cool, not because they are necessary or the simplest way to do something. I think Mike was going for the automatic interlock element - the micro automates opening and closing so you can't goof the switches up.
Unless the Frankenstein element of knife switches is what you're going for, it seems foolish to use them at line voltage. A dual throw with center off switch would seem safer.
-
I used to do Karnaugh Maps all the time in the 70's while designing relay logic for packaging machines. It's been a long time, but I'm betting it still works...
No. Everything has to be done with computer simulation and CAD, CAM and 3D modelling now - just ask any of the cool kids!
Steve.
-
Very interesting. A quick search found an 8-channel Arduino output interface with 10-amp relays rated for 250 VAC. Cost is $19.59 with free shipping to the USA. I can't buy just the relays for that kinda money. http://www.dx.com/p/arduino-8-channel-12v-relay-module-expansion-board-142797#.U2ywpijVvpc (http://www.dx.com/p/arduino-8-channel-12v-relay-module-expansion-board-142797#.U2ywpijVvpc)
Gosh, I think this controller needs 9 relays for the outlet goof matrix, so perhaps I should do a Karnaugh Map and try the simplify the circuit a bit. I used to do Karnaugh Maps all the time in the 70's while designing relay logic for packaging machines. It's been a long time, but I'm betting it still works...
VERY LOGICAL!
-
Unless the Frankenstein element of knife switches is what you're going for, it seems foolish to use them at line voltage. A dual throw with center off switch would seem safer.
As much as I love big knife switches, putting line voltages on them and flipping around in a demonstration is probably a little over the top. One easy(er) way to do this could be relays and diode logic on a rotary selector switch. I worry about the break-before-make time of the matrix relays though. I just don't want to haul around a huge demo board with a few dozen outlets in various states of mis wiring. Not elegant enough and takes up too much room in the pack. Ideally this this be in a little quad box with some sort of remote controller.
Just dabbling with how to even do this is fun. I think showing all the ways that 3-light outlet testers can mislead you would be highly instructive.
-
Just dabbling with how to even do this is fun. I think showing all the ways that 3-light outlet testers can mislead you would be highly instructive.
How about 3 rotary switches with 4 positions each. So each of the 3 contacts in the duplex outlet could be switched to H-N-G-X
That would be foolproof and pretty cheap to build, I think.
-
Yup KISS.... I actually gave some thought to designing a smart outlet strip that would sense the mains connections and self correct if needed, but not really a viable product.
JR
-
Very interesting. A quick search found an 8-channel Arduino output interface with 10-amp relays rated for 250 VAC. Cost is $19.59 with free shipping to the USA. I can't buy just the relays for that kinda money. http://www.dx.com/p/arduino-8-channel-12v-relay-module-expansion-board-142797#.U2ywpijVvpc (http://www.dx.com/p/arduino-8-channel-12v-relay-module-expansion-board-142797#.U2ywpijVvpc)
Gosh, I think this controller needs 9 relays for the outlet goof matrix, so perhaps I should do a Karnaugh Map and try the simplify the circuit a bit. I used to do Karnaugh Maps all the time in the 70's while designing relay logic for packaging machines. It's been a long time, but I'm betting it still works...
I would think you could simplify it, considering the tester can't tell a ground, neutral reversal. You really only need hot and Ground/neutral switched to 3 contacts. This works out to a 2x3 relay matrix. I would also consider just using LEDs to show the conditions and no real power. 3 RGB leds in an outlet overlay + 3 leds showing the tester output. Maybe a 3rd rgb LED showing, OK, wrong, Deadly. An ardiuino can also drive an LCD display with some up down buttons. You could show expected meter readings? Maybe turn it into a game?
-
If a game you could create a random miswiring and require correct fault identification. It might be wise to use a lower voltage, or a specially built current-limited milliamp mains-voltage supply.
-
If a game you could create a random miswiring and require correct fault identification. It might be wise to use a lower voltage, or a specially built current-limited milliamp mains-voltage supply.
I was thinking sll low voltage with simulated meter readings on the LCD.
-
Simulations are good-but with some people it doesn't seem to click until they see the same tester they have been using give misleading readings-there is perception that it is a complex magic little box as opposed to 3 simple lights with a key to interpret them.
Three terminals-black, white, green each with three possible connections-hot, neutral/ground, or no connection at all gives 9 possible combinations. But it could be done as simply and cheaply as 3 SPDT on-off-on toggle switches to a receptacle. Feed it with a GFCI for a measure of safety if you want to leave the wiring visible (shrink tube connections so nothing bare for first line of defense).
-
I would think you could simplify it, considering the tester can't tell a ground, neutral reversal.
But I HAVE thought about adding a variation to this demonstration on how to tell when the ground and neutral wires are swapped. Can't tell that with a 3-light tester, but you can figure it out using a voltmeter and a load, and perhaps with a SureTest or INSP-3 Ground Impedance Tester. A swapped G-N creates some really strange ground-loop hum issues in sound systems.
-
Simulations are good-but with some people it doesn't seem to click until they see the same tester they have been using give misleading readings-there is perception that it is a complex magic little box as opposed to 3 simple lights with a key to interpret them.
True that... I want to be able to plug in any brand/model of tester and show the kind of mis-wiring issues that confuses them. It also needs to be able to create the really dangerous RPBG (Reverse Polarity Bootleg Ground) condition.
-
One other condition that you occasionally run into is a bad connection that tests OK under no load, but appears to "switch polarity" as soon as a load is applied. Those can make you think you are losing your mind.
I had gotten away from using the handy little testers-its almost as easy to and more informative to use a voltmeter-but with the new tamper resistant receptacles it is all but impossible to use a voltmeter without getting into the box.
-
I had gotten away from using the handy little testers-its almost as easy to and more informative to use a voltmeter-but with the new tamper resistant receptacles it is all but impossible to use a voltmeter without getting into the box.
I now use an INSP-3 or SureTest Analyzer all the time for just those reasons.
-
One other condition that you occasionally run into is a bad connection that tests OK under no load, but appears to "switch polarity" as soon as a load is applied. Those can make you think you are losing your mind.
I can simulate a bad connection with my outlet goof tester by adding something like a 10K resistor in series with each line as desired. That means I need a 5-position/1-pole rotary switches for each line, so I could label them H-N-G-R-O for Hot, Neutral, Ground, Resistor, Open. As long as the rotary switch is break-before-make, then this should be safe for any combination of switches. Of course, this can easily energize the chassis of any plugged-in gear to 120-volts, so that has to be considered. If I get husky rotary switches rated for 10-amps, I can actually put something like a heater load on this demo receptacle to show how to find swapped G-N lines with a meter.
I think this design is Safe, Simple and Sexy... I'll start looking for parts.
-
I think this design is Safe, Simple and Sexy... I'll start looking for parts.
I can get a 6-position rotary switch with 20 amp contacts on eBay for $2 each.
http://www.ebay.com/itm/GE-ASR7173-510-1165248-6-POSITION-ROTARY-SWITCH-120-240-VOLT-88145-/360703536601?pt=LH_DefaultDomain_0&hash=item53fb9b2dd9 (http://www.ebay.com/itm/GE-ASR7173-510-1165248-6-POSITION-ROTARY-SWITCH-120-240-VOLT-88145-/360703536601?pt=LH_DefaultDomain_0&hash=item53fb9b2dd9)
I just need some cool looking pointer knobs and an aluminum box to mount everything in. Oh yes, got to isolate the receptacle ground itself so I don't energize my own test chassis. Plus I'll put a GFCI breaker in the power feed so if something goes wrong it shouldn't kill me. Anyone know of a panel mounted GFCI?
-
Cooper or P&S makes a dead front GFCI that mounts in a device box or maybe use one of the portable GFCI cords and hardwire it in as your power cord?
Might also consider a .1 ohm/10 watt resistor to simulate roughly 60 feet of #12 wire in the neutral circuit. I guess that just depends on the wiring where you use it for a demo. If you wind up somewhere with a bootleg ground, you won't get much difference between your neutral and ground!
-
Sorry to have started this ball rolling and then not replied to the many good comments. I've been out for nearly a week.
Update on what I've done here:
1) While I was gone, I killed the breaker. Upon return, I unplugged everything from the bad circuit and flipped breaker back on.
2) I have two 3-light testers, a Greenlee GT-10 and a Woodhead 1750. Both provide correct readings on known good circuits.
3) Both testers show the first 3 outlets in the circuit to be good. All outlets downstream of the 3rd show Hot & Ground reversed.
I deduce from this that either a) outlet 4 is bad or b) wiring between outlets 3 and 4 is bad. Are there other possibilities?