Floating scope ground

[Mick Bailey]

I used to have the loan of an expensive Tektronix isolated input module, but that source has dried up. I know some people disconnect the ground lead off the mains plug, or run off a mains iso transformer, neither of which are safe. Anyone have any tricks for this? I was wondering about using a 1:1 transformer on the probe.

[flyingdutchman]

The recommended solution is to use an isolation transformer to power the equipment under test. This is what I do when I have to probe stuff like switching power supplies.

I have a project somewhere for a differential input capable of withstanding mains voltages, it was published by Elektor quite some time ago. It's neither cheap nor easy to build, but probably performs better than a transformer.

[Tom Phillips]

...Anyone have any tricks for this?...

In some situations I have used both inputs of a two channel scope with the channels set to "add" and one channel inverted. The scope ground 1s still connected to the system ground of the device under test and the setup can still display the voltage across any two points in the circuit where you connect the probes. If those points are within the max voltage rating of the scope probe/input used of course. (And your scope has the add channels and invert functions) Tell us what you are trying to measure.

[g1]

Why is using an iso transformer on the scope more dangerous than using it on the unit under test?
I thought isolation transformers used to be standard equipment for repair shops?
Most things you do as a tech are dangerous to some extent, to stay completely safe we could never open up a unit .
What do you use to looks at primary side of SMPS?

[Mick Bailey]

Why is using an iso transformer on the scope more dangerous than using it on the unit under test?
I thought isolation transformers used to be standard equipment for repair shops?
Most things you do as a tech are dangerous to some extent, to stay completely safe we could never open up a unit .
What do you use to looks at primary side of SMPS?

I use an iso transformer for SMPS primaries, but still don't go touching everything around. Textronix says;

"A common but risky practice is to float
the oscilloscope through the use of
an isolation transformer that does
not carry the ground through to the
secondary or by disconnecting the
oscilloscope’s AC mains power cord
grounding connector.
“Floating” a ground referenced
oscilloscope puts all accessible
metal including the chassis, casing,
and connectors at the same
voltage as the test point that the
probe reference lead is connected."

[flyingdutchman]

Why is using an iso transformer on the scope more dangerous than using it on the unit under test?

Because the entire 'scope chassis will be live or thereabouts, and there are lots of exposed parts one might inadvertently touch when adjusting the 'scope controls.

When you use an isolation transformer to power the equipment under test, let's say an smps, you can safely connect your probe's ground clip to the negative terminal of the primary reservoir cap, and the precautions from that point on are not much different from those one has to adopt when working on a valve amp.

[g1]

Thanks guys! My scope doesn't have metal controls, and I rarely need to isolate it (no more SMPS primaries for me). However, should I need isolation in the future, I will use the xfrmr for the unit under test rather than the scope.

[km6xz]

I do not like to float test instruments unless they are intended for it such as insulated case meters. A differential probe is not too hard to find used or surplus and solves the problem. There are lots of solutions, such as differential probe amplifiers, self-contained diff. probes, a diff amp plug in for a Tektronix mainframe scope which are pretty cheap nowadays, or using two matched probes with a standard bench scope that has channel inversion and channel summing. For weak signal work, the external preamp will have better noise figure and CMRR than a bench scope but for most work....switching supplies, bridging power amps, class D amps etc, the two channel bench scope is fine. For switching supplies, the 100x probes come in handy.

[Enzo]

g-one, your scope may not have metal knobs, but when you reach for that plastic knob, all it takes is your knuckle brushing against the frame of the scope or the BNC barrel on the probe connector, and...

[flyingdutchman]

and...

ZAP!

The only exception I can think of is one of those Philips/Fluke "scope meters", they are completely insulated and can be used fully floating.

[Mick Bailey]

In some situations I have used both inputs of a two channel scope with the channels set to "add" and one channel inverted. The scope ground 1s still connected to the system ground of the device under test and the setup can still display the voltage across any two points in the circuit where you connect the probes. If those points are within the max voltage rating of the scope probe/input used of course. (And your scope has the add channels and invert functions) Tell us what you are trying to measure.

Just monitoring bridged amp output with a single channel and using the other channel to monitor the input. An isolated probe means I can clip across the amp output without blowing it, or frying my probe. The differential technique means I've used up both inputs of my scope.

[bob p]

I keep my scope grounded at all times. On my scope I just flip the display selector to: A - B, and take a differential measurement.

Tom gets the bonus points for coming in first.

To add my own personal experience (that's only of marginal value): I bought a brand new isolation transformer and I can't remember ever actually using it. My tasks have always been simple enough that 2 channel differential measurements have always done the job for me.

[dbdoc]

We got a Fluke scopemeter a couple of years ago at work to look at video signals in the field. It is actually terrible for looking at video, but it is completely isolated whether using on battery or external power supply. And the 2 inputs are isolated from one another. So you can not only use anything for reference, you can have 2 different references at once. Now the only problem is remembering I can't do this on my other scopes.

[bob p]

> Now the only problem is remembering I can't do this on my other scopes.

right -- i'm not sure that i'd be happy learning to use a scope that's double insulated and battery powered. sounds like a good way to pick up bad habits...

[Mick Bailey]

I just tried a 1:1 audio transformer with a resistive attenuator and it works quite well as an iso module for the probe. Bandwidth isn't really an issue for what I need this for - having the other scope channel free is the main thing. Looks like something I'll build into a proper box.