Scopes and balanced connections

[fum]

good day,

I've been thinking a bit about oscilliscopes, and their proper application, and my thinking is that how you test is often as important as what you test.

I've got an old 20mhz dual trace hitachi scope, that does not have a balanced input.

When attempting to look at equipment that has balanced output, how would one properly look at the signal on output?

I've heard that using a battery powered scope often makes this easier, as it is floating from ground reference.

Does the use of the scope change if you are looking at something that's output is balanced by the use of an output transformer, vs. something that has solid state output?

does the output need to be properly loaded?

I may be trying to make this into something more difficult than it is, but thought I'd ask for some collective wisdom.

Regards

ju

 

[bcarso]

For a lot of so-called balanced outputs the signals are equal in magnitude and opposite in polarity. So referencing the scope to common and looking at each in turn shows you something at least.

When it's a dual channel scope sometimes the B channel has a polarity invert switch and the scope has an A+B display mode. Then once you've adjusted one of the other channels variable gain looking at a common-mode signal (both probes connected to the same signal generator, which can be the local sqaure wave one on the scope for adjusting the probe compensation) to null, you can use the probes as differential inputs---i.e., look at each side of the balanced signal. You may still need to reference one or both probe grounds to device common if the device probed is floating relative to the safety-ground referenced scope. This may create other problems, although putting a small resistance in series with the scope probe ground should alleviate them.

The output transformer equipment should allow you to look across + and - with a single channel---it should not care much about grounding the - side, if that's all that the trafo's secondary is connected to.

Battery operated scopes are ok unless the internal switching supply kicks out a lot of noise. This was the case with the big Ni-Cd based power supply for the Tek 475, as I discovered to my chagrin, and it made it essentially unusable for my purposes. I'm sure they are doing a much better job in the LCD display portables now.

 

[BYacey]

You could plug your scope into an AC isolation transformer, or lift the ground. This can create a potential safety issue, so care and caution must be taken. The iso transformer is the preferable method.

 

[bcarso]

And lifting the ground leaves you with usually some very stiffish line bypass caps, so the whole scope is sitting at 60VAC and will readily blow out semiconductor junctions if the unit under test is referred to safety ground and you touch a circuit point with the scope ground clip.

 

[CJ]

Pretend you didn't read this:

(rip out the center prong on the pwr cord)

We know resume our regular programming. :grin:

 

[bradb]

The whole scope and balanced line intersection has been my center of confusion for a while.

If pin 1 on a balanced line is not ground, just a shield, then how to you measure the voltage of pin 2 or 3 using an oscilloscope, meaning what is the voltage on 2 and 3 IN REFERENCE to?

Would you connect the probe between pins 2 and 3?

still getting a grasp on this whole balanced, floating and differential stuff, its almost like the concept of a ground needs to be discarded

 

[DrFrankencopter]

Measure between 2 and 3....that's all it takes.

Cheers,

Kris

 

[PRR]

> what is the voltage on 2 and 3 IN REFERENCE to?

To each other.

Actually, if it is really a BALANCED output, both sides are equal but opposite in reference to ground. So in any ideal world, you can measure from pin 1 to pin 2, and know that the signal from 2 to 3 is the same only twice as big. This is nearly true for many transformerless outputs, ignoring the fact that pin 1 is not necessarily a clean ground. (It is defined as a shield, not a signal path.)

If it is known to be transformer floating (or an excellent transformerless imitation of a transformer), you ground pin 2 and read from 2 to 3.

Since most of these outputs really-really expect a differential input on the other end of the wire, and diff-inputs are fairly common, you really should measure with a differential input. This may mean routing through a very excellent "balanced line input" to the nearest unbalanced patch output, or getting a diff-input scope.

> does the output need to be properly loaded?

"Need to be?" No, they all do something unloaded.

There are two main cases:

1) load it the way the maker tells you to

2) load it the way YOU plan to use it.

#2 may be more relevant to YOU. But if it sucks that way, don't complain to the maker until you try it the way THEY tell you to load it. If their way works good and your way is less good, re-consider your plans.