Why is some test gear grounded?

[e100]

My mains powered scope is grounded which causes ground loop interference when measuring mV outputs from grounded sensors.
My battery powered ungrounded scope doesn't have a ground loop problem and doesn't seem any more dangerous than my mains powered scope.

This doc says that you shouldn't use a mains isolation transformer to deal with ground loops.
www.tek.com/Measurement/App_Notes/Technical_Briefs/tds3000-float/eng/51W_10640_1.pdf]http://direct.[url=http://www.tek.com/Measurement/App_Notes/Technical_Briefs/tds3000-float/eng/51W_10640_1.pdf]www.tek.com/Measurement/App_Notes/Technical_Briefs/tds3000-float/eng/51W_10640_1.pdf[/url]
Is this because of the exposed BNC metal connectors being at some potential other than earth?
Would this issue disappear if scopes had the same insulation design as multimeters with all plastic construction, recessed connectors and an internal mains isolation transformer?

Mike

[alm]

Yes. You would also have to redesign the case to provide enough insulation, a typical DMM has a very different case from an oscilloscope. Look at the Tektronix TPS2000 series or the Fluke ScopeMeters. Expect degraded performance and a much higher price. Oscilloscopes have the added complication of orders of magnitude more bandwidth and multiple channels, so you have to isolate each channel. This is usually accomplished with a combination of optical and magnetic coupling, to cover the full range from DC to 100 MHz or so.

Does your battery powered scope have insulated BNC connectors, insulated probes (as shipped with Fluke ScopeMeters, I believe they are built by MC/HCK) and a proper insulated case designed to withstand mains voltages? Otherwise I would advise against using it for anything other than SELV, ie. low voltage, low power circuits.

[Psi]

Would this issue disappear if scopes had the same insulation design as multimeters with all plastic construction, recessed connectors and an internal mains isolation transformer?

interesting thought, i guess so.

But there is always going to be exposed metal on the scope probe. So maybe its impossible to get the device to comply with current regulations.

[alm]

Just use these probes from MultiContact. No exposed metal.

[ejeffrey]

Other solutions:

1) Float your sensor.  Battery power is best, transformer is next best.  SMPSs will have non-negligible capacitance to ground even at 50/60 Hz, so they may or may not solve the problem.

2) Use a differential probe.  This will completely solve the problem.

3) Use two probes in A-B.  This has poor common mode rejection at high frequency, but is likely good enough and has the advantage of being essentially free.

4) Add a differential output amplifier to your sensor.  This has the same effect as #2, but you don't need to try to hit the full bandwidth of the scope, and it may be more useful if you want to connect the sensor to something other than a scope.

5) Add a high gain amplifier to your sensor so that a few mV don't bother you. This loses you dynamic range, and may require impractically accurate or stable zero signal correction, but may be the simplest solution.