Bob- sorry I just noticed your question regarding "what is a USB isolator?"
I will start out with what is USB?
USB is a bi-directional serial data system with several capacities, the system automatically adapts to the slowest capacity device. It also provides some limited power to run things that dont need much- like thumb drives, small hubs, etc.
Now what is an isolator?
An isolator acts like "wire" with no physical electrical connection from one side to the other. So anyone would reasonably ask- how do you send power across such a thing? Answer: good old transformers. Well really modern high-tech transformers, RF transformers instead of 60hz transformers. The power capacity of an isolator is approximately 95% of the original capacity of the computer it is plugged into. In a nutshell - plug it in and dont worry about it.
So- how does data get across the "gap"? - optical isolation. Data going each direction drives LED's that get connected to photo-detectors on the other side. No electrical connection.
Since power is transferred from one side to the other, the isolator only works when the power source is on one side (the male usb plug) and can only provide power on the other side (the female usb plug).
And that last little thing- signaling the capability of each side to the other- that is done with optical isolation also.
What you end up with is a USB cable that can tolerate very high voltages from one side to the other. Usually over 1000 volts before the system fails catastrophically. Some isolators are over 3000 volts, then I wouldnt want to be working on a computer with 3000 volts on the other side of that little cable would you? I dont even want 50 volts..
How much differential ground voltage can USB take with standard cables? Conceptually none, but in reality about 0.75 volts. How much can old fashioned RS232 serial tolerate? About 7 volts. Thats a factor of 10 difference, where the "new and improved" serial system is not as tolerant as the old serial system. Both of which were designed based on the assumption that there was no ground differential voltage between the sender and the receiver.
When I was at Tim's shop, I measured the ground differential voltage - which was usually around 0.005 volts, with a maximum of 0.025 volts or so during air cuts. We did not cut anything so no dust was going through the ductwork which can change that 0.025 to a much higher number.
What I was checking for is ground loops, that was clearly eliminated as a problem source. That leaves two suspects- Windows 8 and dust collection static. Tim has tried another computer without any change in reliability so that leaves us with dust collection static buildup.
An isolator might help with that problem, but its a shot in the dark.
Good news is USB isolators can be purchased for under $50 on Amazon.
Now to deal with the real problem- static charge in the dust collector. For that, I am devising a meter to read how much static we have. Then I intend to try different things to lower it, and rather than just guess how effective each possible solution is, I intend to measure it and remove the guesswork.
Hopefully that is a full answer on USB isolators, and how it pertains to this installation.
I will now rotate my hat back from "electrical engineer" to "woodworker".
D
"The best thing about building something new is either you succeed or learn something. Its a win-win situation."
--Greg Westbrook