My mother is attached to a hard sought and found outside wall mount lamp above her garage. The metal lamp case sets off my current indicator as much as 110-120 volts do. I knew we have touched it in the past with no consiquences unknowing it was a bit hot. So the voltage leaking into it is low. But I set out to remedy the leak, worring that it could get worse especially in wet conditions. I checked for a possible polaritry switch to see if the white wire in the lamp was connected to the hot power source wire. It was not; the polarity is correct. The wire that leads from the switch box in the garage to the lamp is not a proper 14 gage wire. It is a too long and too exposed to wear and tear 18 gage. The seath is frayed. The metal casing of the lamp was touching some of the bare wire. But after that was eliminated the case still tested hot. So my next guess is that the lamp internally has a place where the hot wire in some way touches the case directly or indirectly. I don't know. I looked at the lamp and there is no visual damage or obvious problem. I cannot get it completely apart however.

My mother wants to continue to use the lamp as before since no one has been zapped. I do not know enough about it and would like to know more. So can someone say more about this problem?

Thank you

You need to take it completely down and inspect the whole thing and the wiring feeding it. This is the only way to be sure.

I am going to change the wire feeding the lamp to a new 14 gage run between the switch and the lamp. But before doing that, I temporarily put up a new lamp on the old wire and its case does not seem to be hot. However, the replacement may have been a bad choice as a diagnositic device since it appears to be mostly plastic. I wanted to see if it ruled out a problem with the wiring. That is, if other-replacement lamp's casings do not test hot, all things being the same, then it was the old lamp itself and not the wiring. That appears to be the case--although, as I said, I realized afterward that the mostly plastic case is a problem. I will change the wiring and see if the old original lamp still has a hot casing. But if the wiring to the lamp is to blame, what theoretically is going wrong? If the two wires, hot and neutral, were touching somewhere in the run from switch to lamp, wouldn't the white wire test hot too? It does not.

I would like to add that in going around the house and checking many of the lamps with metal casings, they almost all set the detector off. So either there is a house wide problem or it is not a problem at all. Why would touching the outside of metal lamps set the voltage detector off?

Those testers are famous for giving false readings.

If your using the type of tester what we think, this is called technically a "proximity" tester, marketed as a "non-contact" tester. See attached photo of an Ideal tester.

The reason your detecting voltage ( Not Current) on plastic lampholders is the proximity of both the hot and neutral terminals in a lampholder.

The tester is doing it's job, indicating voltage in the assembly, enough to advice that power is present and should be shut off to work on the assembly.

These types of testers are not for the purpose of troubleshooting anything beyond present voltage.

You must use a voltage tester that contacts terminals to determine if a wire is actually hot or grounded.

You clearly seem to be working on something with tools you are not familiar with.

I do not allow my service electricians to use these testers, period.

Once one indicated no voltage, even though it was tested on a live circuit,and I had an experienced electrician come in contact with 277 volts, took a 14 foot fall after being shocked, (Before OSHA rules were so well enforced as now )After the ambulance ride to the emergency room, and a overnight stay for observation because there was blood in urine from the fall, he was released.

The shock he could take, like most of us electricians. The fall is the killer to us.

The shock can be fatal to laypeople.

Use the proper tools for troubleshooting.