How much noise floor and other things matter in oscilloscope usability

[Fiorenzo]

Hello,
I am in the process to "buy" an oscilloscope and I really need some clarification because, even if I've read almost everything I have been able to find on the web and on this forum, there are some technical things I cannot sort out by my self.

For example the noise floor:
In what situation really matter to have an oscilloscope with a "low" noise floor?

I am going to use It for many different things, I am a "begginer" but I do digital stuff with embedded electronics and I also aim to learn more things as possible about analog electronics starting from working on an old valve radio that i would like to repair and experiment with.

Actually I bought few days ago a Rigol mso5000 and I did not be aware about its noisy front end.
After a lot of searching on the web and experimenting with such instrument It seem to me that It is at least 3 times more noisy than other comparabile oscilloscopes.

So how much this matter in electronics? How this could preclude its usability?

I have still about 30 days to give It back for free to Amazon but only one week to decide if I want to buy its direct competitor siglent "sds2000x plus" that now Is on offer with the LA probe discount bundle until 30dicember.

The Siglent appear to have a low noise front end but a different way to handle signal recording in its internal memory that i don't understand if it Is bad due to the fact that It cannot "zoom out"...

Also the Rigol has a very fast ADC when working in interleaved mode but at the same time I still don't understand if 8 Gsa/s are really needed with frequency up to 500Mhz and maybe the Siglent with Its 2Gsa/s is enough.... I don't find clear infos on the web apart the nyquist theorem.

I would be gratefull if someone with experience could help me because all this matter a lot for me, electronics has been in my heart from when I was young and I am keen to improve as best as possible day after day....

Thank you

[Kleinstein]

For most uses the noise of the scope is not really an issue. This is because a 1:10 probe is often the mayjor noise source.  Unless the scope is rather high in noise, the noise floor is mainly an issue with the few measurements done with an 1:1 probe or similar signal via coax.
For digital signals itself the scopes noise is usually not an issue, but it may be when looking at the supply ripple.

For comparing the noise, one should take the bandwidth into account. With a higher BW the RMS noise naturally goes up. This effect can make a new higher BW scope look higher noise. So  one usually should compare at the same BW, like the usual 20 MHz BW setting.

2 Gs/s are a bit on the low end for 500MHz BW. This may lead to some aliasing and a few more thoughts about what one is actually seeing at the short time scale end.

[nctnico]

For most uses the noise of the scope is not really an issue.

I strongly disagree with this. An oscilloscope with a lot of noise gives thick traces which make it hard to see the actual level of a signal. Ofcourse you can use high-res or bandwidth limiting but IMHO these should be targeted at cleaning up a signal and not masking a poorly designed analog front end. An oscilloscope with a low noise floor is easier to work with. I used to own an Agilent DSO7104A and it was just horrible to work with for analog stuff due to the massively thick traces it has due its own noise.

Also note that the noise doesn't only apply to the most sensitive V/div setting, it applies similar to all V/div settings. The noise level is usually specified in Volts using the most sensitive V/div setting but it would be more accurate to specify it as a percentage of a division or full range. In the end the V/div setting adjusts an input divider but the actual noise level (what goes into the ADC and what gets added by the ADC) stays the same; it is just scaled differenty.

[Fiorenzo]

I find your replies very interesting.

So now I am thinking: for what type of signal and in what situation the noise floor of the oscilloscope is impacting the anologue measurements?

[Grandchuck]

https://keysightoscilloscopes.wordpress.com/tag/noise-floor/

[Fungus]

For example the noise floor:
In what situation really matter to have an oscilloscope with a "low" noise floor?

When you're measuring very small signals.

I am going to use It for many different things, I am a "begginer" but I do digital stuff with embedded electronics and I also aim to learn more things as possible about analog electronics starting from working on an old valve radio that i would like to repair and experiment with.

It won't make any difference at all on your digital stuff.

For the radio? If it turns out to be a problem you can easily add a preamplifier and make it even better than a lower-noise oscilloscope.

So how much this matter in electronics? How this could preclude its usability?

It's not a showstopper. You can still do everything, just maybe not as easily for a few specific things.

The real questions are: How often do you do those things? How much would you have to spend to get a lower-noise 'scope with the same abilities as your Rigol? Is the extra money well-spent?

[nctnico]

For example the noise floor:
In what situation really matter to have an oscilloscope with a "low" noise floor?

When you're measuring very small signals.

Actually not because the relative noise floor remains more or less constant!
I just tried on an R&S RTM3004. At 1V/div I get 17mV stdev. At 10V/div I get 170mv stdev (Stdev= RMS with DC removed). The noise floor scales with the V/div setting.

[Fiorenzo]

Grandchuck I read that article and many others but they doesn't give example of practical situation that are impatted from the noise floor of the oscilloscope. Which kind of signal need an oscilloscope with a low noise frontend?
I ask these because I do not have a clear comprehension of such subjects.

[Fungus]

For example the noise floor:
In what situation really matter to have an oscilloscope with a "low" noise floor?

When you're measuring very small signals.

Actually not because the relative noise floor remains more or less constant!
I just tried on an R&S RTM3004. At 1V/div I get 17mV stdev. At 10V/div I get 170mv stdev (Stdev= RMS with DC removed). The noise floor scales with the V/div setting.

If the signal is large, you can see it, it's just that the traces on screen are thicker.

If the signal is small, it gets lost inside the trace.

[Fungus]

Which kind of signal need an oscilloscope with a low noise frontend?

A 1mV signal.

(for example)

[nctnico]

For example the noise floor:
In what situation really matter to have an oscilloscope with a "low" noise floor?

When you're measuring very small signals.

Actually not because the relative noise floor remains more or less constant!
I just tried on an R&S RTM3004. At 1V/div I get 17mV stdev. At 10V/div I get 170mv stdev (Stdev= RMS with DC removed). The noise floor scales with the V/div setting.

If the signal is large, you can see it, it's just that the traces on screen are thicker.

Yes, and thick traces just suck. Try to make a cursor measurement on a trace that is 20% of a division even with V/div set to 1V/div. Needless to say that small variations of a signal are also lost in a noisy oscilloscope.

Don't confuse needing a pre-amplifier to look at signals that are outside the V/div range of an oscilloscope. That is a different subject!

[Fiorenzo]

Ok, you all are giving me many usefull information, but now give a look to the attached photos.
The first image shows the apparent noise of CH1, It is set in AC mode, 1X attenuation, without bandwith limit and is also sampling in normal mode with the probe attached and grounded to its own tip.
The second image shows the same configuration but with the probe attached to the output of and old transformer just to check its ripple.

I see a lot of noise floor in this photos.
Do you agree? It seem too much to me, but as I said before I am not an expert in analoge stuff so maybe I am doing something wrong... Or maybe this oscilloscope is particularly noisy...

[Fiorenzo]

Which kind of signal need an oscilloscope with a low noise frontend?

A 1mV signal.

(for example)

What circuits works with such a low signal?  This is my question from the beginning.
I understand that noise sucks but it is the practical application of a low noise oscilloscope that i cannot figure out.
For example, if I work with on an old valve radio am I going to encounter such kind of low signal? It Is only an example....

I don't want to buy an oscilloscope and find myself after a year that It is going to particularly limit my study in electronics, this are costly instruments, I can spend some more money than that spent for the Rigol but I must do the right choice and now I don't feel confident with my actual knowledge.

[bdunham7]

I haven't compared those two models directly, but comparisons with two lower models of Rigol and Siglent were enough to convince me that low noise was an important factor in some cases.  The Rigol was a lot noisier than typical analog scopes, whereas the Siglent is comparable with a good basic CRO.  The Siglent also has a for-real 500uV/div capability and the Rigol actually only went down to something like 2mV/div (or perhaps 5mV/div) and then expanded the signal and decreased the resolution for the lower ranges, a sort of digital zoom.  They had a 500uV/div range if you hacked them, but it was pretty useless.

As for some previous responses, I disagree that the 10X probe will be noisier than the scope, that might apply to a very quiet scope but with either of these the scope front-end noise will be the issue when you are looking at small signals with a 10X probe.  And you'll almost always be using a 10X probe because 1X is very limited bandwidth, so a 'low' signal will be anything below 50mV.  So when I've used a 10X probe with scopes like this on small signals (the lower versions of Siglent and Rigol) the Siglent is clearly superior (3-4X better at least) but sometimes I wish it were even better.  Always remember the bandwidth limiter will help if your signals are below 20MHz.

There was a mention that the noise appears on all scales, not just the lowest volts/div settings.  There's some of each, I suppose you could call them the analog-front-end noise and the ADC noise.  The Siglent clearly has more noise on the lowest volt/div settings and above 2mV/div, the 'ADC' noise is pretty minimal.  The Rigol will be worse in this regard because of the digital expansion for the lowest ranges.

You asked about sample rate and bandwidth.  2GSa/s is just enough for 500MHz, but the Siglent only has that with two channels active and so the BW is limited with 4 channels.  The Rigol has more than adequate samples for its 350MHz bandwidth under any conditions.  I'm not sure how much this matters for your uses.

The memory configuration and lack of zoom-out on the Siglent is a baked in trait that I don't think is going to ever change.  I find it to not be problem, but it has annoyed some people who are used to a different configuration.  I'm also annoyed because I think its usability could be improved with trivial effort by reducing the whole-record display to a bar on the top.  This whole thing is less of a problem on the SDS2000X series because in zoom mode you are wasting 1/4 of a fairly large screen.  Earlier models were wasting 1/2 of a much smaller screen.

[bdunham7]

What circuits works with such a low signal?  This is my question from the beginning.
I understand that noise sucks but it is the practical application of a low noise oscilloscope that i cannot figure out.
For example, if I work with on an old valve radio am I going to encounter such kind of low signal? It Is only an example....

On an old valve radio you might want to use a 100X probe, in which case a 'small signal' that would merit using the lowest range of the scope might be hundreds of millivolts.  With a 10X probe, which is what you will almost always use, you might start caring about front-end noise at 50mV.  Low front end noise gives you better FFT performance as well.  There are all sorts of cases where noise is an issue and if you are just starting out,  I can't predict which ones you will run into.

[Fungus]

Ok, you all are giving me many usefull information, but now give a look to the attached photos.
The first image shows the apparent noise of CH1, It is set in AC mode, 1X attenuation, without bandwith limit

Never use 1x mode without the bandwidth limit.



nb. 10x mode is what you should be using almost all the time. 1x mode is only for very special cases.

[Fungus]

What circuits works with such a low signal?  This is my question from the beginning.

The example that's usually given is "power supply ripple", but unless you're designing a switched mode power supply then you probably don't need that.

Yes, and thick traces just suck. Try to make a cursor measurement on a trace that is 20% of a division even with V/div set to 1V/div. Needless to say that small variations of a signal are also lost in a noisy oscilloscope.

Put it in the middle of the trace.

Oscilloscopes aren't that accurate anyway - only about 5% even on a low-noise oscilloscope.

Edit: You can also turn on color gradient mode and the true signal will be highlighted for you. 



(I just learned that trick from the video below and you can be sure it will be repeated in all future "Rigol noise" threads...  )

[Fungus]

If the signal you're looking at is periodic (repeatiing) then you can turn on hires mode and look at the average of multiple waveforms:

[egonotto]

For example the noise floor:
In what situation really matter to have an oscilloscope with a "low" noise floor?

When you're measuring very small signals.

Actually not because the relative noise floor remains more or less constant!
I just tried on an R&S RTM3004. At 1V/div I get 17mV stdev. At 10V/div I get 170mv stdev (Stdev= RMS with DC removed). The noise floor scales with the V/div setting.

Hello,
this is much better than on RTA4004. R&S says: (50 Ohm 1GHz)
1 V/div  31.4 mV

Best regards
egonotto

[nctnico]

For example the noise floor:
In what situation really matter to have an oscilloscope with a "low" noise floor?

When you're measuring very small signals.

Actually not because the relative noise floor remains more or less constant!
I just tried on an R&S RTM3004. At 1V/div I get 17mV stdev. At 10V/div I get 170mv stdev (Stdev= RMS with DC removed). The noise floor scales with the V/div setting.

Hello,
this is much better than on RTA4004. R&S says: (50 Ohm 1GHz)
1 V/div  31.4 mV

I switched the 20MHz bandwidth limit on in order to allow making comparisons with other scopes. The example is just to show that the noise floor scales with the V/div setting.

[David Hess]

Noise level almost never matters because oscilloscopes are time domain instruments with two more significant limitations:

1. Noise is typically lower than the aberrations in the pulse response, and this is also a limit for instruments with resolution higher than 8 bits, so removing the noise just gives a clearer view of an inaccurate result where it counts.

2. Input stage offset and drift is usually greater than the noise.

And unlike the above errors, a digital storage oscilloscope can reduce noise with averaging.

Where noise does matter is spectrum analysis, but again, other limits like distortion have a greater effect.  Optimizing oscilloscope input stages for pulse response requires compromise in noise and distortion.

Good low noise performance for a 100 MHz instrument should be about 40 microvolts RMS based on a input JFET source follower with 3.2 nV/Sqrt(Hz) noise, and many old analog oscilloscopes and DSOs achieve this.  This is also why sensitivity below a couple millivolts/division is questionable except if bandwidth is limited.

For instance my 2232 DSO has 25 points per division so at 2 millivolts/division each point is 80 microvolts, and peak-to-peak noise is 3 points or 240 microvolts which comes out to 40 microvolts RMS.  I measured slight lower than this on some of my analog oscilloscopes under similar conditions, although my favorite analog oscilloscope has a noise level about 3 times higher than this.

[Fiorenzo]

So, with these settings:
1x probe, no BW limit, normal acquisition mode

I did other photos:
1) probe grounded to its tip
2) probe connected to a Power supply to check the ripple
3) no probe attached to the oscilloscope

What do you think about the noise floor?
It seem high to me....

[Fiorenzo]

Here same settings but hi-res mode.

[bdunham7]

Can you try again only set the vertical scale to 1mV/div and measure with no probe and with the shorted probe?  Then do the same at 100mV/div.

[Fungus]

So, with these settings:
1x probe, no BW limit

Again: You're not supposed to use 1x with no BW limit. Ever.

1x with no BW limit = noise.

2) probe connected to a Power supply to check the ripple

You can see the ripple, right? That's what counts.

Try the same thing with the bandwidth limiter on.