Weird TIG arc instability

It sounds just like what happens when you forget to turn on the argon. My guess is its a gas flow problem- either you forgot, or the valve was sticky, or there was some teflon tape in the hose, or a kink, or something along those lines.

I have plenty of gas. I can feel it coming out and I have used half of the bottle without this problem.

Cannot be. I used half of the bottle before it got this way.

Sounds to me like you're getting discharge in places other than from tungsten tip to workpiece. That suggests a deposit of metal in the various parts of your torch, cup, lens, etc.

You might try gritblasting those parts where a discharge could occur from tungsten to metal deposit to workpiece.

You may need a new torch.

If there were conductive paths in the torch, he'd be getting zapped by the HF. Hmm, that does give me a thought, that his HF is turned up way too high. Some machines, like the bigger Lincolns, have so much HF that you can actually weld thin materials without any real welding current. The HF does tend to hop around a lot, too. But, the horrible mess of his electrode and the total melting of the gas lens is not something the HF system could do. Maybe if he was resting the torch or gas lens cup against the workpiece, the welding current could jump to the dirty cup and then to the work. That would almost certainly shatter the cup within seconds, though. And, there'd be obvious arcing from the cup to the work.

There is nothing in the torch handle that is grounded. There couldn't be, or the HF would arc to it for sure.

Jon

I looked a the gas lens and indeed there were signs of brass melting around the periphery. There also copper deposits in the tungsten in the region where it emerges from the lens. I cleaned out all the threads on the inside of the torch. When I was reassembling the torch I saw that there is a spot for the wrench on the lens and I have been installing it only finger tight. So now I made sure the lens is in snug. The wild arc behavior is gone. Thanks for all the help.

Not if the conductive material were only in and around the cup, not extending anywhere near where the user holds the torch. This seems to be the most likely region of contamination.