Tweet
My brain is getting whacked by test equipment grounding issues.
Turns out my signal generator output has one lead grounded. My oscilloscope has a ground. Both grounds are connected through the three-prong power plug.
If I want to inject a signal somewhere in my circuit, I have to think carefully. One lead of the signal, being ground, will automatically make a short circuit to wherever the oscilloscope ground is.
I can use an isolation capacitor, but then I still have an "AC" short between those two points.
I can use two oscilloscope probes in "differential" mode to try to avoid attaching the oscilloscope ground, but sometimes you get a poor waveform without having a ground reference somewhere in the circuit -- I think. And then each probe will still measure zero AC volts relative to the grounded signal lead -- not necessarily bad, but can be confusing, especially when you leave differential mode and expect floating probe signals but get distinct measurements relative to ground.
I could also use an "isolation" audio transformer to isolate my injected signal from the signal generator. Is this worth it? What does it mean to apply such a floating signal across a differential amplifier for example? Do you need a "ground path" somewhere anyway.
I don't have power isolation transformers, but if I did, would I want to use one on my oscilloscope and one on my signal generator, so each is floating?
I would be interested in a thorough discussion of such grounding issues.
-SB
Turns out my signal generator output has one lead grounded. My oscilloscope has a ground. Both grounds are connected through the three-prong power plug.
If I want to inject a signal somewhere in my circuit, I have to think carefully. One lead of the signal, being ground, will automatically make a short circuit to wherever the oscilloscope ground is.
I can use an isolation capacitor, but then I still have an "AC" short between those two points.
I can use two oscilloscope probes in "differential" mode to try to avoid attaching the oscilloscope ground, but sometimes you get a poor waveform without having a ground reference somewhere in the circuit -- I think. And then each probe will still measure zero AC volts relative to the grounded signal lead -- not necessarily bad, but can be confusing, especially when you leave differential mode and expect floating probe signals but get distinct measurements relative to ground.
I could also use an "isolation" audio transformer to isolate my injected signal from the signal generator. Is this worth it? What does it mean to apply such a floating signal across a differential amplifier for example? Do you need a "ground path" somewhere anyway.
I don't have power isolation transformers, but if I did, would I want to use one on my oscilloscope and one on my signal generator, so each is floating?
I would be interested in a thorough discussion of such grounding issues.
-SB
Comment