Unreliable voltage detector

[Richard Evans]

I have a Greenlee GT-11 voltage detector, practically new. It's about the size and shape of a magic marker, but instead of a felt tip it has a flat blade with which to probe for live voltage. Stick it into a receptacle or even touch it to the outside of a live cable and it flashes and beeps. At least, it used to.

Last night, I applied it to a cable I was pretty sure was live and it was mute. I then tried it in a receptacle I knew was hot and it was still mute. I tested the batteries. They were good, but I replaced them anyway. Still nothing. I then resigned myself to buying a new one and tossed it into a box of batteries, where it immediately began beeping. I took it out and started testing it again and it beeped accurately, until I turned it off and back on, then nothing again.

So, the problem appears to be in the switch, though the thing is so simple I don't know how it could be failing intermittently.

Bottom line: Once the thing is on and known to be working, it is reliable until turned off. Once turned off, it has to be tested on a known live circuit when it's turned on again. (Which is recommended practice anyway.)

My concern now is: Once it's on and working, can I depend on it to work for as long as it's on? Could it just stop working without warning?

Unless someone has some simple solution, I guess I'm off to buy a new one.

[Bill]

I suggest you learn to use a multimeter if you are going to be working with electricity or call a professional.

[kcarlson46]

I have one of those non-contact voltage detectors. Number 1 rule in using it is to check it against a known live cicuit to confim that it is working, then go test the circuit you need to (usually to make sure all the wires in the box you want to work on are dead), and then go retest against the known live circuit.

Ken

[dpb]

Richard Evans wrote: ...

Fluke has a model that has continuous self-test if you're in the market.

Unless can discover an intermittent connection at switch that is repairable or some similar, seems like best solution. Being unsure of reading is worse than no measurement and simply assuming everything is live imo.

--

[Ralph Mowery]

Either that or a light bulb detector. Something that makes direct connection to the wires. I work as an electrician in a large plant and we were given detectors similar to that around 10 years ago (the company furnishes all the tools). I never did trust the things. Especially not enough to grab a wire. I carry one in my tool pouch, but seldom use it, I usually use the Fluke t-100 meter I have. Guess that I am old school and still like the Simpson 260, but I don't usually carry it as it is too big to fit in my back pocket.

[Ralph Mowery]

using it is to check it against a known live cicuit to confim that it is working, then go test the circuit you need to (usually to make sure all the wires in the box you want to work on are dead), and then go

That should be the # 1 rule always nomater what you use.

[Charlie]

I just use a bug light. It used to be a neon pilot light. Entirely self contained.

However, if you really want to use your toy detector, check the battery contacts in the device. They should be clean and shiny and have lots of spring so that you get a nice tight fit.

Charlie

[AZ Nomad]

high impedance multimeters work very poorly on house wiring. The capacitance between a pair of wires reads the same as a connection.

I've resorted to wiring a spare lamp socket as a test lamp to have a low impedance "test light".

[trader4]

Since it's almost new, I'd contact the manufacturer. On many cheap things like this, they just send you a new one. It may be under warranty too. If that doesn't work, then I guess I'd take it apart and do some investigating, but I sure wouldn't trust it unless you can find and fix what ails it.

[John Grabowski]

*I had a Greenlee and found it to be less than perfect. I now use a Fluke and love it. No more false readings.

[The Daring Dufas]

My clamp-on amp meter has a detector built in. I push the data hold button with it turned off and the continuity beeper buzzes with AC hum. It's very reliable.

TDD

[HeyBub]

Send it to Greenlee with a note:

"Thought you might like to investigate the failure mode on the enclosed GT-11 with a view toward making a more reliable unit.

"I wouldn't say no to a replacement..."

They're a pretty good company - I'll give 8-to-5 you get a new probe.

[Stormin Mormon]

That doesn't save much time, or effort. Since you go through all that routine.

[Bob(but not THAT Bob)]

I've had one of these for 30 yrs and I've always been able to trust it:

Of course, it was around 25 bucks back then

[Red Green]

"HeyBub" wrote in news:4oKdne8PaI8eta3XnZ2dnUVZ [e-mail]:

....Although the shock I received was very uncomfortable, at least I wasn't on a ladder, fell and got injured. Lawyers around here eat that shit up.

[David Combs]

Any suggestions on what to read to get a good understanding of all this?

And, what words to search for on wikipedia?

Thanks!

David

[AZ Nomad]

ohm's law

[bud--]

There is an explanation of "phantom voltage" at:

probably doesn't say much more than Ralph posted.

Digging out some old notes:

A good analog meter will have an input impedance (AC scales) of about "5,000 ohms per volt". On a 150V scale the resistance of the meter will be 750,000 ohms.

To read 120V, the current through the meter will be 0.00016A (160 microamps).

A digital meter will have a higher input impedance, say 10 megohms (probably higher).

To read 120V, the current through the meter will be 0.000000012A (12 microamps). This is a very small current.

=============== A couple years ago I measured the capacitance between 2 insulated conductors in a 3 conductor with ground Romex as 15 pf per foot (other 2 conductors open).

For 20 feet that is 300pf.

At 60 Hz, the 20 feet of Romex will have capacitive reactance of 8.8 megohms

==================== With 20 ft of the Romex above, one Romex wire to hot, 2nd Romex wire to digital meter and other digital meter lead to neutral you have the 8.8 megohm capacitive reactance is in series with 10 megohm meter. With 120V across the 18.8 megohm total the meter reading is about 64 volts.

Actually that is wrong because the capacitive reactance and resistance don't just add. If added correctly the total impedance is 13.3 megohms and the meter reading is about 90 volts.

===================== The 8.8 megohm capacitive reactance in series with a 750,000 ohm analog meter will give a much smaller voltage because the capacitive reactance is much higher than the meter resistance. The analog meter will read about 10 volts.

The capacitive reactance is so much higher than the meter resistance, the current is pretty constant if you switch to lower scales, and the reading will stay about 6% of full scale. (This can be as odd as having any voltage at all.)

======================== The example above would be similar to measuring the voltage at a light bulb socket (no bulb) where there is a 20 ft run of old ungrounded Romex to a light switch that is turned off. The run to the light switch has a hot supply wire and the switched return wire. The capacitance between the wires gives the meter reading.

======================= All measuring devices have limitations. Knowing what the limitations are is part of the challenge of making measurements. The digital meter measures an actual voltage (and probably measures low). But the "phantom" voltage has no practical significance.

[blueman]

When I was saving up presents to buy my first VOHM back when I was a young kid in the earl-70s, all consumer versions were analog. In those days, the more you paid the higher the input impedance -- because it was useful for electronics (so that the measurement wouldn't affect the reading) and since getting high input impedence was expensive back then. I think there were some models (out of my price range) that had megohm type input impedance using FETs but I the best I could afford was one with about 20Kohm/volt.

Interesting that now the situation is reversed -- all the cheap (Chinese-made) multimeters are digital with megohm input impedances while you have to pay extra to get an "old-fashioned" analog one.

I think I recently saw my old Archer (Radio Shack) multimeter lying around in a box at my parents' house. I probably should retrieve it. By the way, I think I paid about $20 for it on-sale back in the day which was a heck of a lot of money then for a 10 year old kid who was getting 10 cents a week in allowance (and it wasn't the lowest end one either).

[bud--]

The first meter I bought when I was a kid was a 20Kohns/volt Lafayette kit. I suspect a bunch of us learned a lot of electronics from kits. I don't know how kids learn electronics now.

The main alternative back then was a VTVM - thats a vacuum tube volt meter for the younger types. I probably still have one of those around.

It always amazes me that the 20Kohm/volt meter movements are 50 micoamps full scale. Hard to imagine you can make something that sensitive rugged enough for field work.

Those meter are typically 5Kohms/volt on AC scales, which I used in my calculations. If I remember right you need higher current to get the diodes in a more linear operating area.