How can RF power a multimeter?

[elmo]

I noticed something peculiar today as I was troubleshooting a TIG-welder with a Fluke 179. During fault tracing I decided to test the welder at my bench, as I lit the arc I noticed that the display said "F179", the same screen that shows when you turn on the DMM. I tried it a few more times, and it showed almost every time I tried to ignite the arc. It also showed "batt" a couple of times. The switch on the DMM was in the off-position when this happened. I also noticed that the leads on the DMM was hanging against the cable to the welding torch. I know for a fact that the initial ignition voltage on the welder is 10 kV, I am not sure about the frequency but if the hissing sound of the spark and the transformer inside the welder is any indication I wold say less than 10 kHz.
The display on the DMM turned off as the arc lit, so i wasn't able to test if I could use it.

I guess my question is, how is this possible? Isn't the switch on the DMM supposed to disconnect the battery from the rest of the circuitry?
I presume the power emitted from the welder isn't enough to power the DMM, but again I really don't know.

[Psi]

Simple reflective LCD panels (the kind you find on a multimeter or a wrist watch) use an extremely small amount of power, and the MCU in the fluke is low power as well (since it's designed to prolong battery life).

The battery in the fluke maybe disconnected by a switch but the RF from the TIG is just passing through the case and onto the PCB tracks directly. There's probably overvoltage zeners and other components which are limiting and cleaning up the induced voltages that get onto the power rail.

Note that a MCU will often 'run' from a voltage way lower than its minimum rated voltage.
They just crash often and features don't work correctly.
The RF may even overpower the brownout detector if it has one.

[Time]

Its just EMI and the strong fields from the welder inducing voltages randomly in the meter.  You wouldnt be able to use the meter that way.  Electrical arcs tend to be very noisey.  Random portions of the meter are getting power and causing it to display stuff.  Its a pretty stochastic and unpredictable process which makes using the meter that way not feasible.

[Fraser]

Many modern equipments use power management chips that control a MOSFET that in turn controls the meters main power rail. The power on and off is controlled by either the main processor (that is in low power sleep mode when the meter is off) or by a dedicated power control IC that consumes minute amounts of current when in standby. True On/Off hardware switches that disconnect power, are less common these days as many pieces of equipment are never truly 'OFF' they are in 'sleep' or 'hibernate' mode, much like many PC's.

If you introduce these components to strong electromagnetic fields such as produced by the TIG, you will likely stimulate either the processor or the power control chip, producing the effects that you have described. Its classic RFI behaviour.

Aurora

[siliconmix]

 your welder is transmmiting on all rf frequencies