The Silent Killer - 120VAC Hot Chassis

[K1JJ]

Today I plugged in my 813 rig into the 120VAC socket to watch the pretty tubes glow and the fans run.  It ran fine. I switched it off and then picked up the rig to slide it over on the bench. With it in my arms I touched the cabinet of another working rig at the same time.  BAM!!  I dropped Hollywood right on the table after getting a blast across the chest of 120VAC.

I thought, WTF?  There's not even any HV on.  I knew immediately it had to be 120VAC.  I then added a jumper ground between the 813 rig and the other cabinet and the jumper blew itself apart.  The VOM showed 120VAC between the 813 rig and station ground.

To make the story short, inside the 813 rig  I found a terminal strip for AC that had one of its lugs touching the chassis and grounded one leg of the 120AC to the chassis, rather than floating.  I corrected the problem and all was FB again.

In hindsight instead of relying on a 3-wire plug, I should have had a station ground connected to the 813 rig. This would have blown the breaker.  Without the station ground I had relied on nothing for protection for the last few days.  The chassis was hot and I didn't even know it.  What a dummy!

I wonder how many of us run a voltage test on their rig's AC wiring to be sure it is not tied to chassis by accident?  This happpened to me 30 years ago when I was wiring floodlights on a trailer.  One of the 120VAC leads shorted through a metal wall and put the whole trailer shell hot to ground. The only way I knew was when my dog jumped up on the metal steps and yelped.

Again, a good ground would have exposed the problem, but sometimes we can get lazy and lax and depend on a 3-way cord in the early stages of building.  In my case, a simple VOM test would have shown the hot chassis.

Well, at least the problem is fixed.   Just wanted to give a heads up.

T

[w8khk]

Very thankful that you are OK, Tom.  Reminds me of how we lost Ross Hull, another great, albeit before my time.

I am curious why the short between hot and chassis did not raise a flag if the ground wire of the three-wire cord was connect to the chassis.  Is there a problem with the power cord, the plug, or possibly the source receptacle?  I suspect you have already addressed that issue.

One thing I feel is very important is verifying the wiring and reliability of all receptacles in the shack.  For one thing, I use only 20 amp receptacles, as they are much better quality, and tighter connections to all the pins.  When kitchen appliances are plugged into 15 amp receptacles, the plug pins get rather warm; not so on the 20 amp receptacles.  So I prefer to avoid voltage drop on all receptacles in the shack.  I make my own power strips, either using metal receptacle boxes, or replacing the receptacles in power strips manufactured with standard duplex receptacles.

To verify the wiring, especially the ground, I use an "Ecos" Ground Test Device.  We used these religiously when doing computer support for Hewlett Packard systems, starting in the late 70s.  There are the simple three-neon bulb testers that supposedly verify the receptacle is wired correctly, but they do nothing to examine the quality of the ground connection.  The Ecos device generates a pulse on the ground pin, and measures the amplitude of the pulse between the ground and neutral, thereby indicating whether there is a low-impedance on the ground wire back to the distribution panel.  Failure reveals a loose connection or defective wiring, indicating that the ground may not provide a current path to safely trip the breaker in the event of a ground fault in the equipment attached to the defective power source.

The Ecos testers come in various configurations.  Attached is a photo of the one I use.

EDIT:  These Ecos testers obviously cannot be used with GFCI receptacles.....

[K1JJ]

Hi Rick,

That's a good point.... why didn't the 120VAC ground wire take out the circuit when it was first plugged in?  I traced it out and found that the AC strip I was using only had a two wire plug on the end!  It was from 30 years ago and when I didn't pay much attention to three wire plugs.  I went thru my shack and out of about 20 strips, two had no grounds.

T

[KD1SH]

   Working with lots of old boat-anchors, I come across "floating" chassis all the time, but that's not quite the same as a "hot" chassis.  In the former, the voltage typically appears at the end of a lot of various resistances in the circuit, so it won't hold up under any load at all, and gives you a bit of a "tingle" when you touch a grounded object at the same time.  Still, it's a risk, and I make it a point to change all of my boat-anchor non-polarized cords over to three prong, polarized ones.  Right near my workbench is one of those steel "Lally" posts that hold up the floor of the house.  It's grounded, at least somewhat, since it's planted in the concrete floor, and sometimes I'll brush an elbow across it while touching one of those floating chassis, and get a tingle.  I should wrap it with plastic or something, to prevent that.
   A few years ago, working at a company that made large "moving probe" pc board testers, I got a nasty surprise.  The machines were pretty big, so that sometimes you had to crawl partially inside one to get access when troubleshooting.  We had rolling carts, on which we mounted a computer monitor and other items needed to run the machine, and they also had power strips mounted on them, into which we plugged the power cord for the machine, so you could turn off the whole rig from the power strip.  One day I had to troubleshoot something in the machine, so I turned off the power strip, at which point everything on the machine went dead - the cooling fans, indicator lights - and all the energized contactors opened up with audible clicks.
   So, since the machine was quite obviously powered down, I slid myself inside and across the power supply deck, which consisted of a bunch of open-frame supplies on an aluminum plate.  Zing!  120 AC right into the soft skin on both forearms!
   Now what the....?  All lights were off, the fans were off, and the contactors had opened up.  Where did that AC come from?  A few minutes with my multimeter found the culprit.  The power strip wasn't an off-the-shelf hardware store plastic thing, but a sturdy industrial one, which some knucklehead in our shop had wired up himself, and, in the process, foolishly wired the power switch so that in the off position it opened not the black "line" wire, but the white "neutral" wire instead!  All returns inside the machine were - as they should be - through the neutral wire, and not the ground, so that shutting off the power strip shut down the machine, since the return was opened, but  live 120 AC was still present everywhere in the machine.  It just had nowhere to go, until my forearms provided the path to ground.
   Okay, I know - "lock out/tag out" and all that - it was as much my fault as the knucklehead who wired the power strip, but often we find ourselves making faulty assumptions about the competence of others, and put our lives at risk.  It's not dead until the plug is pulled!  (Except for the caps).

[KA3EKH]

Years ago, I was doing an install of a Harris Z-5 for a dumpy little FM station. I do things like install all the modules, connect the ground strap and transmission line and do a proof. Being cheap they hired the cheapest electrician they could find to do the three-phase connection. I always recommend a separate disconnect and EMT between the main disconnect, transmitter disconnect and the transmitter itself.
They found an electrician that was willing to just do a PVC pipe between the main panel in the building and the transmitter and run the wires and connect them.
By the time I got to the site their electrician had already been there and finished his work. He used four black #8 wires pulled from the barker in the main panel, three to the barker and one to ground to the terminal strip in the transmitter with the ground wire lugged to the transmitter frame.
I had not installed the ground or transmission line yet and decided to check the three-phase voltage on the input of the transmitter so turned on the panel barker and using my Simpson 260 looked at the voltage leg to leg on the input and noticed something was not right. Taking the meter and reading from the transmitter frame to the electrical panel discovered 120 volts between the transmitter case and the ground.
There idiot electrician did not tape or label any of the wires and had swapped one of the three hot legs with the ground and connected a hot leg to the transmitters ground.
Being the ground strap, any control lines or transmission line were not connected yet it did not blow the three pole 60A breaker but would imagine it would have become a very effective electrocution machine for anyone who had touched it and anything else that was at ground potential in the transmitter shack.
From that point on I not only always check the electricians work but also attach the station grounding before ever applying power.

[KD6VXI]

That's pretty sad.  NEC requires us to phase tape our wires if not using color coded wire.

Glad you're OK bud.

--Shane
KD6VXI

[WD5JKO]

Tom, Glad you were OK, and the rig didn't get tossed across the shack.

A few months ago I was on the air (75m) with a D104 in one hand transmitting. I needed to read my log better so I reached for my fluorescent desk lamp to move it over the log. At that very moment, I had 120v arm to arm, and was frozen. A moment later (seemed longer), the GFCI outlet tripped, and everything went dark. I was fine, but a bit surprised. I rebuilt that lamp and added a 3 wire cord, and jumper ground wire all the way to the bulb reflector. One of the wires in the lamp at a flex point had wore through the insulation, and I happened to move that joint enough to get a solid connection.

How many people have been fried from shock when one side of the path went through the D104 neck?

Jim
Wd5JKO

[KD1SH]

  On some hollow-state rigs, you do not want to go pulling out the microphone plug with one hand on the rig and one hand on the metal shell of the plug (ask me how I know this).  Depending on the microphone wiring and which connector pins break contact first, you can get one hell of a wallop.

How many people have been fried from shock when one side of the path went through the D104 neck?

Jim
Wd5JKO

[AG5UM]

I just put an ohmmeter from my station ground (ground rod) to the 3rd prong- ground of the house outlet in the room.
My ground rod is not connected back to the power pole ground, the power pole is about 50 yards away from the house.
I get an ohm reading of 5000 ohms, the transmitter is not plugged in at all,etc.
this does'nt seem right, I tried  digital and analog meters.
The Simpson shows a half volt of AC with this connection, maybe thats effecting the ohm reading,
5000 ohms?
I realize you need proper test equipment, "Ground testers, Megger,etc...",
I was just curious what I would get, ...probably just false readings..
but I think I'll clean my connections again....I think I'll work on improving my grounding system.
AG5UM

[Opcom]

Hi Rick,

That's a good point.... why didn't the 120VAC ground wire take out the circuit when it was first plugged in?  I traced it out and found that the AC strip I was using only had a two wire plug on the end!  It was from 30 years ago and when I didn't pay much attention to three wire plugs.  I went thru my shack and out of about 20 strips, two had no grounds.

T

If I had a dollar for every time I've seen a grounded outlet strip with a missing ground pin on its plug. People rip them off and plug them into old outlets or 2-wire extension cords, then people like us sometimes inherit them.
It's easy to understand why this happens and from necessity of getting something done they get put into use and it's later forgotten that they need a new plug installed. Might be a good idea to check the grounds here. Much has been done in haste.

[KD1SH]

We keep a handful of those three-pin/two-pin outlet adapters, with the little screw tab cut off, up here in the lab where I work, so someone can "float" an oscilloscope when needed, but I often worry that some of the younger engineers and techs don't fully appreciate how dangerous a floating scope can be.  Then again, I can't count the number of times I've seen people clip the probe ground onto a live circuit with the scope not isolated.

Hi Rick,

That's a good point.... why didn't the 120VAC ground wire take out the circuit when it was first plugged in?  I traced it out and found that the AC strip I was using only had a two wire plug on the end!  It was from 30 years ago and when I didn't pay much attention to three wire plugs.  I went thru my shack and out of about 20 strips, two had no grounds.

T

If I had a dollar for every time I've seen a grounded outlet strip with a missing ground pin on its plug. People rip them off and plug them into old outlets or 2-wire extension cords, then people like us sometimes inherit them.
It's easy to understand why this happens and from necessity of getting something done they get put into use and it's later forgotten that they need a new plug installed. Might be a good idea to check the grounds here. Much has been done in haste.

[KD1SH]

Yes indeed, it's pretty much impossible to get a legitimate resistance reading like that - stray leakage currents, and soil galvanic "battery" action will fool your Simpson, or any other ohmmeter that relies on minute voltage drops across a resistance.

I just put an ohmmeter from my station ground (ground rod) to the 3rd prong- ground of the house outlet in the room.
My ground rod is not connected back to the power pole ground, the power pole is about 50 yards away from the house.
I get an ohm reading of 5000 ohms, the transmitter is not plugged in at all,etc.
this does'nt seem right, I tried  digital and analog meters.
The Simpson shows a half volt of AC with this connection, maybe thats effecting the ohm reading,
5000 ohms?
I realize you need proper test equipment, "Ground testers, Megger,etc...",
I was just curious what I would get, ...probably just false readings..
but I think I'll clean my connections again....I think I'll work on improving my grounding system.
AG5UM

[AG5UM]

KD1SH,
good explaination, There is no substitute for understanding what your doing and having the right equipment to do it.
Donnie
AG5UM

[W1RKW]

Here's a good do's and don'ts piece of info from Tektronix on floating.

[KD1SH]

  There's some good advice in there.  But, while nobody can deny that "floating" a scope by isolating it from facility ground is dangerous, Tektronix would love us all to spring for their high-dollar differential probes.  Some of those things cost more than the scope you'd connect them to.
  I've never floated a scope here in my own shop.  I've got a couple of X100 probes, and I'll use them in A-B "pseudo differential" mode when needed.  In the lab where I work we practically never measure anything greater than 12 volts, so floating scopes aren't really a big deal, but it's still something to keep an eye on.  I've tried to talk them into getting some real active differential probes, but they won't go for it.

Here's a good do's and don'ts piece of info from Tektronix on floating.

[KD1SH]

  I've checked grounds rods with a light bulb; one end to the ground rod and the other to the hot side of a wall outlet.  Watch out, though: if you've got other equipment connected to the ground rod, it'll be above ground unless that ground rod has perfect continuity, which is unlikely.  And you can't rely on the voltage drop across the bulb to measure the ground resistance, since the temperature coefficient of an incandescent filament is huge.
   Watch out, too, if you've got pets or kids walking around outside near that ground rod - you'll be setting up field currents in the soil all around the rod.  Of course, if you're planning a fishing trip, you'll have no shortage of worms.
(edit: obviously a GFI would take offense at this)

KD1SH,
good explaination, There is no substitute for understanding what your doing and having the right equipment to do it.
Donnie
AG5UM

[N0WEK]

I had the 3 phase problem years ago at a friends commercial wood shop. I did most of his wiring when I was around on the weekends and he had several 3 phase woodworking machines, radial arm saw, two Rockwell table saws and air compressor. When he bought a new jointer, also 3 phase, I wired it up. With 3 phase motors you have a 50/50 chance of the motor running backwards and I usually just hooked them up and if it ran backwards just reversed two of the lines to correct it. Everything checked out and off I went.

When I stopped back the next weekend he mentioned that they'd been cutting a large piece that had clamps attached and one of the clamps had come between the jointer and the table saw and they'd gotten some sparks, also a couple of tingles between the two, supposedly grounded, machines.

I checked it out and there was 125 volts between the jointer body and ground. I'd swapped the ground with one of the phases, the machine ran fine that way but they'd gone a whole week with 125 volts on a steel machine three feet from a hard ground on the table saw, not to mention a concrete floor. Very lucky nobody got killed.

A few years later he bought a different building that had all the 3 phase disconnects and breaker panels in an old section of the building. I wired up all the lighting, 3 phase and the rest in the new section. I checked everything this time and it ran fine for a year or so.

The office was in the old section with a wood floor and on the other side of the wall from the incoming electrical service panels which were right next to the front door. The guys mentioned that they'd been getting shocks from the grounded breaker panels. I spent a fair amount of time looking for bad bonding and lifted grounds and found nothing. It finally developed that nothing was wrong with the panels. There was a short between the hot on a receptacle in the office and a piece of galvanized roofing that was acting as wainscoting on the office wall. When it rained and the roof leaked by the front door the hot piece of sheet metal that was a piece of the office wall leaked voltage into the wet floor and they were getting zapped by touching the hard ground at the panels. They'd gone for more that a year with a 3 ft. x 10 ft. piece of sheet metal at 125 volts with no shocks until the floor got wet and they got grounded.

[w8khk]

 I've checked grounds rods with a light bulb; one end to the ground rod and the other to the hot side of a wall outlet.  Watch out, though: if you've got other equipment connected to the ground rod, it'll be above ground unless that ground rod has perfect continuity, which is unlikely.  And you can't rely on the voltage drop across the bulb to measure the ground resistance, since the temperature coefficient of an incandescent filament is huge.
   Watch out, too, if you've got pets or kids walking around outside near that ground rod - you'll be setting up field currents in the soil all around the rod.  Of course, if you're planning a fishing trip, you'll have no shortage of worms.
(edit: obviously a GFI would take offense at this)

KD1SH,
good explaination, There is no substitute for understanding what your doing and having the right equipment to do it.
Donnie
AG5UM

I do not think a GFI (GFCI) would even care about ground currents caused by your light bulb.  In fact, a GFCI will function properly in a 2-wire (hot and neutral) connection, and the only ground connection in a GFCI is between the ground screw terminal for the green wire and the grounding pin part of the receptacle.  I am not suggesting that this wiring arrangement meets code, but the GFCI (Ground Fault Circuit Interruptor) is simply composed of the neutral and hot wires passed through a toroid, with a sense winding on the toroid to determine if there is a difference between the hot and neutral currents.  So long as all the current from the hot wire returns via the neutral wire, the currents cancel, thus there is no voltage on the sense winding, and all is well.  If some of the current from the hot wire  finds a different path to ground (such as through a person's body) then the sense winding signal activates the GFCI and the latching switch is released, removing the power to the receptacle.

[KD1SH]

   That could very well be the case; I must admit that I have no GFCI's here in the house, and so never had a chance to experiment with them.  I assumed that, given a GFCI protected outlet, any current sensed by the device in the hot lead that isn't balanced by an equal return current in the neutral, would trip the GFCI.  The GFCI, as you say, has no "real world" method of knowing whether current is flowing to ground or not - and indeed doesn't even care if there is a ground - it only knows that what went out didn't come back, and if it didn't come back, it must have gotten lost and is out causing mischief somewhere.

I do not think a GFI (GFCI) would even care about ground currents caused by your light bulb.  In fact, a GFCI will function properly in a 2-wire (hot and neutral) connection, and the only ground connection in a GFCI is between the ground screw terminal for the green wire and the grounding pin part of the receptacle.  I am not suggesting that this wiring arrangement meets code, but the GFCI (Ground Fault Circuit Interruptor) is simply composed of the neutral and hot wires passed through a toroid, with a sense winding on the toroid to determine if there is a difference between the hot and neutral currents.  So long as all the current from the hot wire returns via the neutral wire, the currents cancel, thus there is no voltage on the sense winding, and all is well.  If some of the current from the hot wire  finds a different path to ground (such as through a person's body) then the sense winding signal activates the GFCI and the latching switch is released, removing the power to the receptacle.

[KD6VXI]

Yes indeed, it's pretty much impossible to get a legitimate resistance reading like that - stray leakage currents, and soil galvanic "battery" action will fool your Simpson, or any other ohmmeter that relies on minute voltage drops across a resistance.

I just put an ohmmeter from my station ground (ground rod) to the 3rd prong- ground of the house outlet in the room.
My ground rod is not connected back to the power pole ground, the power pole is about 50 yards away from the house.
I get an ohm reading of 5000 ohms, the transmitter is not plugged in at all,etc.
this does'nt seem right, I tried  digital and analog meters.
The Simpson shows a half volt of AC with this connection, maybe thats effecting the ohm reading,
5000 ohms?
I realize you need proper test equipment, "Ground testers, Megger,etc...",
I was just curious what I would get, ...probably just false readings..
but I think I'll clean my connections again....I think I'll work on improving my grounding system.
AG5UM

My FLUKE megger / insulation tester goes to 1.5kv IIRC.

Pretty  good, considering it's a couple small (in relation) batteries.

--Shane
KD6VXI

[WB3JOK]

This is a detailed report of the location and repair of a ground fault on a casino "boat".
Lots of misplaced amps, but no smoke 
Interesting reading (at least I thought it was) 
http://www.mikeholt.com/download.php?file=PDF/Blue_Chip_Casino_Michigan_Indiana_mjt1.pdf