Rigol 1052E noise at 100Mhz, 3Mhz, etc..

[abieneman]

hi All,

This is probably a newbie question, but since this is my first DSO, here it is - Why is there a constant noise at 100Mhz, 3Mhz, 2Mhz, etc.?

I am using the probes that came with the unit and shorting end of the probe to the ground on the probe. Probe is set to 1X.

I'd like to know if this is expected behavior and if it is, how do i then measure small <10mV signals?

No bandwidth limiting:



Bandwidth limiting turned on:



Turning digital filter on to low pass mode and turning cutoff frequency to below 100Mhz eliminates the noise



However as i turn the horizontal knob counter clockwise, as soon as sampling frequency changes to 100 Ms/s following noise appears at 3 Mhz:



Eliminating that with digital filter again and continuing to decrease sampling, noise reappears when sampling frequency changes to 50 Ms/s. This time it's around 2 Mhz:



Continuing in same fashion noise reappears again:



and again:




Thank you!

[rf-loop]

(this is one possible thinking road only )

Probe is antenna.

If you connect probe GND to probe center pin you make simple  loop antenna.

Turn radio on and listen around 103MHz. There is maybe roc'n roll. (also you can listen this Rigol with radio)
(Of course in this situation in normal lab it can check with spectrum analyzator. (rigol own FFT is nothing useful for this case)

This is what I first think after I read your question and look these pictures.

103MHz - 100MHz is 3MHz (in oskilloscopes we name it some kind "alias" and this is some kind of posible explanation in this case)


But, this all is only first thinking from this what you have tell. maybe I think different if there is more perfect data and explanation.

take probe off. I think you can not see this. Change CHx mode to GND you can not see this. After you connect this probe "antenna" to scope you see some 103MHz signal. Where from this signal is coming? Find it. (you can use radio, directional loop antenna connected to scope etc..)

maybe I am wrong but with this information it is one and fun possible explanation.
----

Off topic:

If some people use oscilloscope (or other equipments) in area where is high level RF field, there need special knowledge how to do (accurate) measurements. Some times it is very difficult also for true professional peoples and specially for some "professionals".

[alm]

This is where those shorter ground accessories like bayonet ground adapter and ground springs come in handy. If grounding the channel without probe connected doesn't show the problem, you could try shorting the probe tip to the grounding ring with a piece of aluminum foil all around (like Dave did in one of those unusual scope phenomena videos). If that fixes the problem, it's the loop area of the ground lead that's acting as antenna, and you should use shorter ground connections at high sensitivities in your RF environment.

If grounding the channel doesn't fix the problem, it is a problem with the vertical amplifier or the shielding of the scope itself.

[wd5gnr]

What do you see if you set the probe coupling to GND? I agree it sounds like you are picking up something nearby.

[abieneman]

Thank you for your replies. The problem indeed seems the radio pickup:

Probe off, center pin shorted to ground:




Probe on with end wrapped in aluminum foil:



Probe on with regular ground adapter clamped to probe pin:



Same but touching large metal chair:



Probe with ground spring adapter connected to 9v battery in AC coupling mode:



it looks like the ground loop is picking up something around 103 Mhz which is most likely a local 103.4FM radio station. This exercise certainly showed me how fragile a sensor system can be relying on measuring small signals - any length of wire can become an antenna and pick up all sort of frequencies. I guess i would need to put a low-pass filter on my sensor - i am sampling at 40kSa/s, so no sense in passing anything over 4-10 Khz

[DJPhil]

I guess i would need to put a low-pass filter on my sensor - i am sampling at 40kSa/s, so no sense in passing anything over 4-10 Khz

It should have a bandwidth limiting option built in, handy for just these occasions. I'm not sure what it's cutoff would be, but 20MHz is common. I haven't used a Rigol yet (still saving up for one), so I'm not sure where in the menus it would be.

It might not be good enough to filter everything, but it's as cheap and easy as it gets.

Hope that helps.

[wd5gnr]

The 20MHz bandwidth limiting is on the channel menu. Press CH1 or CH2 and there it is.

[abieneman]

Bandwidth limiting indeed helps. If you look at my original post, the second photo is with limiting turned on which cuts the peak to peak from 28 to 6 mV. That built in is not enough to cut through the 3, 2, 1.5, 0.5, .. Mhz noise, so i had to use digital filtering as well.

[DJPhil]

Bandwidth limiting indeed helps. If you look at my original post, the second photo is with limiting turned on which cuts the peak to peak from 28 to 6 mV. That built in is not enough to cut through the 3, 2, 1.5, 0.5, .. Mhz noise, so i had to use digital filtering as well.

Ah, my apologies. Reading comprehension failure on my part. 

[clonecrp]

Great Scope Huh ?