Oscilloscope Grounding Issues

[Legion]

I've been having some trouble measuring circuits with my scope that are hooked up to my function generator. Since the scope and function generator are both earth grounded, I end up shorting my circuit to ground when I hook up my probes.

Do better function generators offer a floating output signal the way the DC lab power supplies give you the option of have a floating circuit or earth grounded? It seems to me that people would use function generators with oscilloscopes all the time, so having both grounded to earth is a big pain in the ass. Isolation transformers were suggested as a solution to me, but it's annoying to have to spend another couple hundred bucks when I've already spent over $2000 on equipment and I'm just learning basic DC circuits. Does this mean everyone has their scope or function generator hooked up to an isolation transformer? I'm wondering if I had bought something better than an Instek SFG-1013 if I wouldn't have to deal with this issue. Maybe a better one would have a switch to select earth ground or floating?

For the specific issue I'm having, see this thread: https://www.eevblog.com/forum/beginners/full-wave-bridge-rectifier/

[Thor-Arne]

The scope is supposed to be grounded, you are probably hooking up the ground clips to different potential. If both the signal gen and the scope is grounded, and you connect the grounding clips to the same potential it's all good.

And from the Agilent InfiniVision user guide: WARNING Always use a grounded power cord. Do not defeat the power cord ground.

I guess the same goes for your signal gen.

If something needs to be floating it is the DUT, but make sure you put the ground clip at the same potential as the signal gen grounding.

[papo]

Some function generators have floating outputs, others do not. Looks like the one you are using is not floating, meaning that you ought to connect the "negative" terminal to ground. Or if your circuit is floating, at least to whatever point your scope ground lead is connected to. (But be sure to understand what you are doing if you go for this option, same holds for an isolation transformer!)

Rarely do I ever need a floating signal generator, so at least for me, it's not really an issue. I use a differential probe if I don't want to connect the scope probe ground lead to ground, but that's apparently not what you are looking for.

Regards
Matt

[tszaboo]

You actually should connect the scope and the signal generator's trigger output together with a coax for better measurements, delay, consistent triggering and stuff. I don't see the big problem with this. I mean, one point of your circuit will be usually ground and you can always DC bias or AC couple (with a capacitor) the signal generator's output. Often times i take the actual signal to the scope with a T coax splitter.

[Legion]

I tried connecting the ground clips of both the function generator and scope to the negative terminal of the circuit. This way they'd be at the same potential. However, it did not solve the problem.

[papo]

I tried connecting the ground clips of both the function generator and scope to the negative terminal of the circuit. This way they'd be at the same potential. However, it did not solve the problem.

Uhm what exactly is the problem? Essentially if your circuit is ground-referenced somewhere, and you hook up the generator/scope ground somewhere else, you are creating a short between that point and the ground reference. Example: If you have a split supply (+12 V and -12 V referenced to ground), and you connect the ground lead to the negative terminal (-12 V in this case), you are shorting out the negative rail. Instead, connect the stuff to the ground reference itself. Dave has a video on these kinds of problems: here.

Regards

Matt

[Legion]

I tried connecting the ground clips of both the function generator and scope to the negative terminal of the circuit. This way they'd be at the same potential. However, it did not solve the problem.

Uhm what exactly is the problem? Essentially if your circuit is ground-referenced somewhere, and you hook up the generator/scope ground somewhere else, you are creating a short between that point and the ground reference. Example: If you have a split supply (+12 V and -12 V referenced to ground), and you connect the ground lead to the negative terminal (-12 V in this case), you are shorting out the negative rail. Instead, connect the stuff to the ground reference itself. Dave has a video on these kinds of problems: here.

Regards

Matt

Thanks Matt. Yes, I've seen that video. While it was helpful in explaining the problem, I didn't find any information about how to fix the issue. I'm trying to keep this thread more about general grounding problems with scopes and other test equipment. But for my specific issue, it's detailed in this thread: https://www.eevblog.com/forum/beginners/full-wave-bridge-rectifier/

[SoundTech-LG]

You'll see warnings from Tektronix, etc. about using isolation transformers on the scope, but if you use 2, one on the scope and one for your EUT, you should be fine. Depending of course on the voltages involved. If you're testing thousands of volts, it could break down the transformer winding insulation eventually. Nothing is perfect.

[Rudane]

You need to use the add and invert functionality. Two scope probes both referenced to earth ground. Hook one before the device measured, one after the device measured and then use the subtract or add and invert function of your scope. This way you never hook the scope ground to anything but earth ground.

[Legion]

You need to use the add and invert functionality. Two scope probes both referenced to earth ground. Hook one before the device measured, one after the device measured and then use the subtract or add and invert function of your scope. This way you never hook the scope ground to anything but earth ground.

This has given me the best results so far, although the signal still looks a little weird to me. Is this, in general, the strategy used to avoid ground shorts?

[Shock]

[Rudane]

This has given me the best results so far, although the signal still looks a little weird to me. Is this, in general, the strategy used to avoid ground shorts?

Absolutely. This issues is not an old one, and really an isolation transformer isn't needed to make these kind of measurements. If the waveform is weird looking, it may just be a weird looking waveform. Don't feel embarrassed about this either, I've taught Senior Electrical Engineering students who could not do a simple differential measurement. Using a o-scope well seems to be a thing of the past, although I'm not sure why.

And always remember, everyone is learning all this as we go. No one has got it all sorted out.

[w2aew]

I tried connecting the ground clips of both the function generator and scope to the negative terminal of the circuit. This way they'd be at the same potential. However, it did not solve the problem.

Uhm what exactly is the problem? Essentially if your circuit is ground-referenced somewhere, and you hook up the generator/scope ground somewhere else, you are creating a short between that point and the ground reference. Example: If you have a split supply (+12 V and -12 V referenced to ground), and you connect the ground lead to the negative terminal (-12 V in this case), you are shorting out the negative rail. Instead, connect the stuff to the ground reference itself. Dave has a video on these kinds of problems: here.

Regards

Matt

Thanks Matt. Yes, I've seen that video. While it was helpful in explaining the problem, I didn't find any information about how to fix the issue. I'm trying to keep this thread more about general grounding problems with scopes and other test equipment. But for my specific issue, it's detailed in this thread: https://www.eevblog.com/forum/beginners/full-wave-bridge-rectifier/

The way to fix the issue is to use a transformer on the signal generator output. From the other thread, you were trying to examine the behavior of a full wave bridge rectifier. Pick up a 1:1 audio isolation transformer, and connect the output of the function gen to one side, and your circuit to the other.

[tszaboo]

You need to use the add and invert functionality. Two scope probes both referenced to earth ground. Hook one before the device measured, one after the device measured and then use the subtract or add and invert function of your scope. This way you never hook the scope ground to anything but earth ground.

Exactly. There should be only one point grounded in the system. And there is no need for differential probes in this application, as we don't talk about very small, very fast and noisy signals.