Osciloscope Noise without probe

[DavidYork]

I have just received the micsig STO1152C oscilloscope about which I have some doubts.
Without a probe it detects noise as you can see in the images in the 2 channels.

   


With the square Test signal the noise persists.



I would like to know if this is normal or if I have some mismatch with the oscilloscope.
Do you know of any thread in the forum that talks about this?

[tautech]

Welcome to the forum.

Normal.
Hunt out Dave's video on DSO noise for further understanding.

[David Hess]

That is a high level of noise but not unusual for a modern DSO.  I do not think anything is wrong with it.

Hunt out Dave's video on DSO noise for further understanding.

Dave did not make any quantitative measurements of analog oscilloscope noise, which is fair since this measurement is often difficult to make.  He is correct in that digital displays make noise more apparent, but most or all modern "consumer grade" DSOs also have more noise. (1)

The best modern DSOs have comparable noise to the best analog and digital storage oscilloscopes of the past since they are both limited by semiconductor physics in the front end high impedance buffer and amplifier.

(1) For most analog oscilloscopes, the trace width sets a lower bound on observable noise which like a DSO, makes the noise look worse than it really is.  But even with this, observed noise is usually lower assuming you trust the DSOs ability to measure its own noise which is not always the case.

[Fungus]

That is a high level of noise but not unusual for a modern DSO.  I do not think anything is wrong with it.

The probe is set to "10x" so all the numbers shown on screen are 10x too big.

The noise on a Micsig is usually about 0.5mV in reality, which is right up there with the best of them.

[David Hess]

That is a high level of noise but not unusual for a modern DSO.  I do not think anything is wrong with it.

The probe is set to "10x" so all the numbers shown on screen are 10x too big.

The noise on a Micsig is usually about 0.5mV in reality, which is right up there with the best of them.

That would make it 90 microvolts RMS instead of 900, which is a little better than the worst "good" analog oscilloscopes that do things the STO1152C cannot, and 3 to 4 times worse than the best.

It is too bad Micsig's documentation and website are so horrid.

[Fungus]

That would make it 90 microvolts RMS instead of 900, which is a little better than the worst "good" analog oscilloscopes that do things the STO1152C cannot, and 3 to 4 times worse than the best.

That's a common mistake: The "RMS" value displayed there includes the DC offset so it's not surprising it seems high.

Dave's done a whole video on the topic:

https://www.eevblog.com/forum/blog/eevblog-1223-whats-all-this-ac-rms-stuff-anyhow/

[Fungus]

Here's mine with a fresh self-cal, probe set to 1x and 20MHz filter correctly applied. Noise is about 50uV.



It's a completely flat line at 10mV/div:

It is too bad Micsig's documentation and website are so horrid.

Can't argue with that, especially the website, but there's nothing wrong with the hardware.

(and firmware/user interface is outstanding even if it is missing a couple of features)

[2N3055]

That is a high level of noise but not unusual for a modern DSO.  I do not think anything is wrong with it.

The probe is set to "10x" so all the numbers shown on screen are 10x too big.

The noise on a Micsig is usually about 0.5mV in reality, which is right up there with the best of them.

That would make it 90 microvolts RMS instead of 900, which is a little better than the worst "good" analog oscilloscopes that do things the STO1152C cannot, and 3 to 4 times worse than the best.

It is too bad Micsig's documentation and website are so horrid.

Micsig STO1000 series has arround 85-90 uV RMS noise, for full 150MHz+ bandwith. It's measurements don't have AC RMS (stdev) so RMS shows AC+DC RMS.
This particular scope seems to have rather low offset, some 5 uV..

I would say those are very respectable numbers for a scope that has price of one Tektronix passive probe..

Could you please elaborate what would be "things that analog scope can do" and this Micsig cannot? It is usually very much the other way around......

[David Hess]

That's a common mistake: The "RMS" value displayed there includes the DC offset so it's not surprising it seems high.

Dave's done a whole video on the topic:

https://www.eevblog.com/forum/blog/eevblog-1223-whats-all-this-ac-rms-stuff-anyhow/

That is what I initially thought but the DC offset is shown in the measurements and insignificant, while the RMS value is consistent with the peak-to-peak value.

[David Hess]

Could you please elaborate what would be "things that analog scope can do" and this Micsig cannot? It is usually very much the other way around......

My benchmark for oscilloscope input noise is the Tektronix 7A13 which is high at about 120 microvolts RMS, but the reason it is high is that it has a full bootstrapped differential input with a common mode range of +/-10 volts.  The Micsig lacks true differential inputs which is hardly surprising, but also lacks that kind of common mode range which would be needed to have 10 volts of offset or position range at 1 millivolt/division sensitivity.

The 7A13 is a useful benchmark because it is a rare example where high sensitivity is combined with high noise making it easy to measure its own self noise.  Most analog oscilloscopes lack enough sensitivity to ignore the CRT trace width.

[2N3055]

Could you please elaborate what would be "things that analog scope can do" and this Micsig cannot? It is usually very much the other way around......

My benchmark for oscilloscope input noise is the Tektronix 7A13 which is high at about 120 microvolts RMS, but the reason it is high is that it has a full bootstrapped differential input with a common mode range of +/-10 volts.  The Micsig lacks true differential inputs which is hardly surprising, but also lacks that kind of common mode range which would be needed to have 10 volts of offset or position range at 1 millivolt/division sensitivity.

The 7A13 is a useful benchmark because it is a rare example where high sensitivity is combined with high noise making it easy to measure its own self noise.  Most analog oscilloscopes lack enough sensitivity to ignore the CRT trace width.

With all due respect, that has nothing to do with analog/ digital and has all to do with fact that it is differential input amplifier.  It is true though, like you said, that capability is not being offered nowadays in any scopes. It is reserved for active diff probes, active voltage rail probes or active diff preamps like LeCroy DA1855A...  Which, in a way is all the same as 7A13, because it is not a scope but a preamp module for 7000 series mainframes..  So it used to be special internal preamp module, now it's external preamp module.

[David Hess]

With all due respect, that has nothing to do with analog/ digital and has all to do with fact that it is differential input amplifier.  It is true though, like you said, that capability is not being offered nowadays in any scopes. It is reserved for active diff probes, active voltage rail probes or active diff preamps like LeCroy DA1855A...  Which, in a way is all the same as 7A13, because it is not a scope but a preamp module for 7000 series mainframes..  So it used to be special internal preamp module, now it's external preamp module.

Your question was what could my example "worst "good" analog oscilloscope" do that an STO1152C cannot.  The "worst "good" analog oscilloscope" for noise is a 7A13 in suitable 7000 mainframe.

For slightly worse noise than a STO1152C, it has a differential input and +/-10 volt input range.  These two things require the compromise of high input noise, although not nearly as high as a high voltage differential probe would have.  It was not used separate from the oscilloscope except in a way that is irrelevant here.  If I was more pedantic, I could have said 7904 with 7A13 vertical input but that does not add anything to the discussion.

What distinguishes it from your examples except the DA1855A, which is a descendant of the 7A13, is the bootstrapped input which adds considerably to the input noise.  I use the 7A13 as a benchmark because I can point to the 7A13 and say, "Your oscilloscope should never have input noise greater than that."  The 7A12 has a singled ended bootstrapped input leading to higher noise but lacks high sensitivity and is too rare to be useful for comparison purposes.

[Someone]

With all due respect, that has nothing to do with analog/ digital and has all to do with fact that it is differential input amplifier.  It is true though, like you said, that capability is not being offered nowadays in any scopes. It is reserved for active diff probes, active voltage rail probes or active diff preamps like LeCroy DA1855A...  Which, in a way is all the same as 7A13, because it is not a scope but a preamp module for 7000 series mainframes..  So it used to be special internal preamp module, now it's external preamp module.

Your question was what could my example "worst "good" analog oscilloscope" do that an STO1152C cannot.  The "worst "good" analog oscilloscope" for noise is a 7A13 in suitable 7000 mainframe.

Given the 7000 series mainframes included a digital scope too, your division of analog vs digital already fails.

Its pretty clear 2N3055 is pointing out the silliness of trying to compare a specialist preamp/differential optional plugin with the standard input of a scope. The correct comparison would be with the modern equivalents; active probes.

[David Hess]

Your question was what could my example "worst "good" analog oscilloscope" do that an STO1152C cannot.  The "worst "good" analog oscilloscope" for noise is a 7A13 in suitable 7000 mainframe.

Given the 7000 series mainframes included a digital scope too, your division of analog vs digital already fails.

The 7A13 has exactly the same noise in a 7854 as any other 7000 series mainframe.

Its pretty clear 2N3055 is pointing out the silliness of trying to compare a specialist preamp/differential optional plugin with the standard input of a scope. The correct comparison would be with the modern equivalents; active probes.

The comparison is between the circuit topologies.  The 7A13 has a bootstrapped differential input stage.  Almost all oscilloscopes whether old or new have a simple singled ended input.  So if your singled ended input has more noise than bootstrapped differential input, it has horrible noise performance.