Rigol DS1052E weird "noise"

[Lunat1c]

Hi,

I didn't want to hijack any other threads with a potential noobie mistake but I was playing around with my recently purchased rigol and for some reason, sometimes when the probe is connected to ch1 or ch2 but not to anything else, I get the signal shown below.

If I press the 'Auto' button, I don't get rid of it. If I disconnect it from the channel, then the signal is normal, once I reconnect it, it gets screwed again. If I then attach the probe to the oscilloscope internal calibration signal, then all I see is gibberish on the screen, I press the 'auto' button and it works fine. Once I remove the probe from the calibration pin, then everything is back to normal. What am I doing wrong?

Sorry for the bad quality pics, that's the best I could do at the moment.

Messed up signal:


Internal Calibration Signal:

[Alex]

Ok, here is what I believe you are experiencing, and a way to verify it too!

What you are seeing on your scope with the probe connected I believe is noise pickup from the probe. It is most lokely around the main frequency, 50/60 Hz (or multiples of). Its amplitude is very low.

When you connect the probe to calibration, the amplitude of the calibration signal generator is much higher. Because the volts per division setting on the scope has not been changed, the signal appears outside the screen (if that was possible).

When you press Auto, the scope changes the volts/div setting and you can correctly see the signal.

When you disconnect the probe from calibration the volt/div is left high from the Auto setting making small signals like the noise picked up disappear from the screen.

The way to test this is simple. After disconnecting the probe from calibration and autoset, press auto set again with the probe open. See and tell us what happens. You should be able to see the noise again.

Alex

[Lunat1c]

Seems like you were right During the calibration the volts/div are at 500mV/div. To see that signal volts per div are about 20mV/div. Thing is when I even move the probe around in free air, then the signal amplitude starts to change. is that normal?

Also it does seem to be mains noise since the frequency is at 50Hz (magnitude at about 50mVpp)

[Alex]

Great, and you learned something! Also, the noise is 50Hz, I see from your pictures the time per division is 20ms and you have exactly one division per period! So it is mains hum pickup. The time you connect the probe to a circuit with some impedance, this noise will be reduced to almost nothing.

Alex

[Lunat1c]

Thanks a lot. One more question, how come it can only be seen when the probe is connected? 

[Simon]

yep 20mS peak to peak sine wave = 50 Hz main hum, try touching the end of that probe with your finger and find out what a great areal you make

[Simon]

Thanks a lot. One more question, how come it can only be seen when the probe is connected?

The probe acts as an areal picking the noise up from your house wiring

[ejeffrey]

The reason the 50 Hz ripple goes away when you touch the calibration terminal is threefold:

First obviously is that the calibration signal is much greater.  Second is that the calibration frequency is higher frequency so your timebase is adjusted so the ripple is harder to see.  Third is that if the coupling is capacitive in nature, it is high impedance, like a current source.  When the current is run through the 1 Mohm input impedance of the scope it generates a large voltage (I*R).  When you connect the probe to the low impedance calibration terminal, the induced current flows into the calibration terminal (probably a few ohms) generating a much smaller voltage.

[Lunat1c]

guys, I just got my fluke 87v multimeter and I am picking the same noise with it in my room. When I switch it to frequency mode it shows a constant 50Hz and the magnitude goes even up to 500mV max (about 60mV constant). Is this normal with every multimeter (my cheapo $20 dollar one doesn't do it)? Will this effect any readings taken with the multimeter?

[Simon]

guys, I just got my fluke 87v multimeter and I am picking the same noise with it in my room. When I switch it to frequency mode it shows a constant 50Hz and the magnitude goes even up to 500mV max (about 60mV constant). Is this normal with every multimeter (my cheapo $20 dollar one doesn't do it)? Will this effect any readings taken with the multimeter?

same reply that ejeffrey gave above applies. Your $20 meter has a lower input impedence

[ejeffrey]

The fluke's input is (IIRC) 10 megaohm, even higher than the scope.  I have several brands of good meters and waving the probes around generates ~100-200 mVAC.

These phantom voltages are the reason many electrician's meters have a low-Z mode, which has an input impedance of a couple kohm in order to short out capacitively coupled voltages present when measuring long cables.

[Simon]

yep I got 3V  out of disconnected cables in my wall, with a 10MR meter

[Lunat1c]

Thanks for your replies ejeffrey and Simon. After some reading it seems that these phantom voltages are to worry about only when you're in the field trying to measure some high voltage conductor and the reading might fool you on whether the conductor is energized or not. As it is, when I measure anything on low power circuits I shouldn't have any problem unless I misunderstood

[vk6zgo]

Yes,in the real world,these things happen,& you simply work around them.
If you are looking at high voltage points,your vertical amplitude setting on the 'scope will be fairly insensitive,so you won't see  a few hundred millivolts anyway.

VK6ZGO