what an oscilloscope recommended for a woman passionate about electronics?

[CharlotteSwiss]

thanks for the help ristofer, I have to work at night for a few days, but tomorrow afternoon I analyze your suggestions
Good evening
 

[CharlotteSwiss]

You may find it interesting to plot some sqrt(sin(x)) equations at Desmos.

On the attached plot, the waveform around 0 is just sin(x) and sqrt(sin(x)).  Notice that the square root is always greater than or equal to the function.  That's because the square root of a number less than 1 is greater than the number itself.  Sqrt(0.5) = 0.707 and .707 is greater than 0.5

The second pair is offset by +3 and this complicates things (but saves on attachments and paper) so draw an X axis at 3 if you print this.  Here the square root is both larger and smaller than the function itself.

In both cases, the square root of a negative number is undefined.

Did I mention I love playing with Desmos?

oh yes, it's nice to flirt with desmos.
I have attached a sqrt test with desmos: ok i understood that the sqrt function does not draw the trace where the signal is negative (it's math)
Here are the few words of the manual:
Where the transform is undefined for a particular input, holes (zero values) appear in the function output.

Then I see that in the positive lobe, the sqrt trace touches the 3 common points (0, Pi / 2 and Pi): and connects them with semicircles.
But what can this view on the oscilloscope help me with? what can it tell me about the source signal? or in which cases could I use it?
thanks 

[CharlotteSwiss]

It actually works - the scope image is a lot like the Desmos graph:

Edit:  I changed the scope image

the desmos example is replicated, in your case we have a less acute arc sqrt (it does not touch the high positive intermediate point), but it will certainly be the mathematical calculation sqrt that plots it like this.
As a unit of measurement I see that in your use mU, in mine they speak of V raised to 1/2 (i.e. a value raised to 0.5 perhaps?)

[CharlotteSwiss]

A consideration for this siglent 1202, already made previously, but to keep remembered to the reader: unfortunately the Print function (save screen on key), too often (3 times out of 5?) Bugs the whole oscilloscope, you need to unplug it to turn it off: obviously when you press print, you have previously made some optimal settings, and on restart nothing is remembered! This is a big problem, which could make many people nervous!

-----

i did sqrt test with an online sine wave, Vp 400mV. The sqrt trace is ok, the negative parts of the signal marks them flat.
I see that the Vp value of sqrt is 624 mV high 1/2 ...
I have to understand how he does this calculation, and what all this teaches me ...
Note: for this function, is not disponible gate cursor

[rstofer]

0.624 squared is 0.389 - close enough.

Taking the square root of a number < 1 results in the root being larger than the number - that's why I put two plots on that Desmos page.  The upper plot is for values > 1.0.  My scope trace above is for a 4Vp-p sine (amplitude is 2V) and the square root is indeed 1.4 (a little less than 3 divisions at 0.5V/div).

Your observation that the square root and the sine wave contacted in 3 places works only when the amplitude is 1.0.  For more fun with Desmos, plot 0.5 sin(x) and sqrt(0.5 sin(x))

I have no idea what the square root function is used for.  It works and if I ever come up with an application, I'll probably be happy I have it.

You might check over in the Test Equipment forum for your Print problem.  I would expect your scope to have the latest firmware but you might find out whether other people are having the problem.  If nothing else, start a new thread and Tautech is likely to see it.  Maybe he will see it here as well.

One option I have with the Rigol is to save the settings on power-down.  That's good and bad.  It's good when I have spent a lot of time getting a particular display with cursors and measurements and I want to restart from that point.  It''s bad when I have put the scope in some strange mode (XY for example) and all I want to do is see a squiggly line.  It's a switchable option and I use it..

[CharlotteSwiss]

tomorrow I will read better, good evening
 

[CharlotteSwiss]

0.624 squared is 0.389 - close enough.

Taking the square root of a number < 1 results in the root being larger than the number - that's why I put two plots on that Desmos page.  The upper plot is for values > 1.0.  My scope trace above is for a 4Vp-p sine (amplitude is 2V) and the square root is indeed 1.4 (a little less than 3 divisions at 0.5V/div).

Your observation that the square root and the sine wave contacted in 3 places works only when the amplitude is 1.0.  For more fun with Desmos, plot 0.5 sin(x) and sqrt(0.5 sin(x))

I have no idea what the square root function is used for.  It works and if I ever come up with an application, I'll probably be happy I have it.

You might check over in the Test Equipment forum for your Print problem.  I would expect your scope to have the latest firmware but you might find out whether other people are having the problem.  If nothing else, start a new thread and Tautech is likely to see it.  Maybe he will see it here as well.

One option I have with the Rigol is to save the settings on power-down.  That's good and bad.  It's good when I have spent a lot of time getting a particular display with cursors and measurements and I want to restart from that point.  It''s bad when I have put the scope in some strange mode (XY for example) and all I want to do is see a squiggly line.  It's a switchable option and I use it..

this is exactly the case, the points of the trace square correspond to the square root of Vp.
Also verified on Desmos as suggested, if the wave has Vp 0.5, the square will have 0.70 at that point, all as expected!
In that 1104 review, the user wrote just this for the square function:
Might be useful to convert some sensor signal, that represents power, back into voltage/current.

For the USB key bug, at the end of the manual study I will investigate better; it could be that with a certain brand / size of key it doesn't go wrong! yes the firmware in July was the last one, then I'll check if they have a new one.
I tried to ask in this topic, even if they talk about 4-channel models ..
https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/1700/

The oscilloscope when I turn it off, it keeps the settings, but if I turn it off regularly; with this bug it freezes ... and forcing the shutdown does not save the configuration, but keeps the previous one ...
The trick now is to do a Print before making a configuration of which I want to save the screen .. if it works there it also works later .. but I don't always remember ..
 

[CharlotteSwiss]

in this narration of mine about a beginner's approach to an unknown digital oscilloscope (which I hope will be useful for future beginners), I came to the measurements; I was curious about the Stdev (standard deviation) measurement.
I saw this video from Dave:



But I'm still not a first level Nerd, not to mention that I followed the translated subtitles ...

Let's see if I understand: Stdev in practice would be a value that we can use when the signal we are measuring has a dc Offset: the RMS measurement is wrong (due to the offset) while the stdev value does not take into account the dc offset, and there is a more truthful RMS value.
I have attached an image:
-in the upper test the signal has no dc offset (stopped by the ac coupling) and we can see that the oscilloscope guesses the RMS value (in this obvious case it also coincides with the stdev value).
-in the lower test the signal has a dc offset of about 1.5v, we can see that the oscilloscope misses the RMS value (2.15v, therefore higher ..), while the detected stdev value is 1.51v, therefore very close to the right RMS value.
In conclusion we can say that it is better to use the stdev value as the RMS value?

 

I add: it seems that with a signal with dc offset, the stdev value indicated is the RMS one, while where the RMS value indicates it corresponds to the measurement AC + DC = [sqrt (ac * ac) + (dc * dc)]

[alsetalokin4017]

We need two ways to characterize a distribution of varying values. Some measure of central tendency (average, mean median mode) and some measure of variability, in other words how much does a particular value differ from the "average" and other values "on average" or how big is the spread of the data. Standard deviation is one such measure of variation. If the standard deviation is small, it means that all or most of the values in your data set are close to the "average" or arithmetic mean. If the standard deviation is large it means that there is a lot of variability in the data. It is also a way of characterising how much a particular value differs from the mean, in terms of the overall variability of the data.

The traces on a DSO are in some sense a display of the "average" of many cycles of the waveform of interest. The standard deviation is a measure of how much the individual values going into that "average" vary from the average itself.

[CharlotteSwiss]

We need two ways to characterize a distribution of varying values. Some measure of central tendency (average, mean median mode) and some measure of variability, in other words how much does a particular value differ from the "average" and other values "on average" or how big is the spread of the data. Standard deviation is one such measure of variation. If the standard deviation is small, it means that all or most of the values in your data set are close to the "average" or arithmetic mean. If the standard deviation is large it means that there is a lot of variability in the data. It is also a way of characterising how much a particular value differs from the mean, in terms of the overall variability of the data.

The traces on a DSO are in some sense a display of the "average" of many cycles of the waveform of interest. The standard deviation is a measure of how much the individual values going into that "average" vary from the average itself.

in fact I had also looked for standard deviation, and the definition is that: but if you look at the stdev measurements of the oscilloscope, the stdev value is equal to the RMS value in the case of a clean signal (or even if soiled by dc offset)
thanks for contributions
 

I add: we wrote together, while you added the final part: ok, but it doesn't seem like a variation value to me, but an effective value of the wave

[rstofer]

The standard deviation should be related to the variability in measurements.  For example, say your 1 kHz signal varied between 0.99 kHz and 1.01 kHz - there would be some standard deviation which would be a function of the variability.

The RMS value itself doesn't matter and shouldn't show up in anything other than the measurement itself.  But, if the signal generator allowed the voltage to flop all over the place, the standard deviation could be quite high.

I have attached a scope shot of a 1 kHz 4Vpp sine wave with a 1V DC offset.  Nothing much happens to Vpp or Vrms as a result of the offset.  The standard deviation is quite low. in the mV range.  Vpp seems to have less variability than Vrms.

I really hate statistics so I'm going to leave it as just the fact that the standard deviation shouldn't be anything like the Vrms - maybe there is a button to reset the statistics so it doesn't start with the beginning of time?

Notice that the standard deviation in frequency (middle block of numbers) is about 1 Hz in a 1 kHz signal.  I suspect that very accurate timed signals could do better but at some point the error is in the scope timebase and synchronizing the input.

[CharlotteSwiss]

I really hate statistics so I'm going to leave it as just the fact that the standard deviation shouldn't be anything like the Vrms - maybe there is a button to reset the statistics so it doesn't start with the beginning of time?

then there is something wrong, it is impossible that my Stdev value is close to the RMS value; it should be a negligible value, the signal is the 1Khz compensation signal, I don't think it fluctuates much, indeed almost nothing.
Poor even with an online sine wave.
If I had the generator, I would try to replicate the compensation signal, but without dc offset, with a Vp1,5v (vpp 3v) and I would like to see what stdev value the oscilloscope would indicate to me ..

But if my STdev is the same as your dev, I find it impossible for the value in mine to be 1.5v

 

[rstofer]

Switching from a sine wave to a square wave, I get terrible standard deviation in Vpp for a 1 kHz 4Vpp + 1V offset waveform - like around a volt.  Right up until I hit the Statistics Reset button.  Then I'm down around 20 mV on Vpp and 3.3 mV on Vrms.  I think that Reset button is a really big deal if the samples start before the generator even starts sending a signal.

ETA:  I think the Siglent User Manual does am award winning lousy job on statistics.  It is essentially worthless.

I suspect the reason it is so poor is that nobody uses these functions so there is no pushback on the manual.  The few people who actually use it probably figure it out by experimenting.

[CharlotteSwiss]

Switching from a sine wave to a square wave, I get terrible standard deviation in Vpp for a 1 kHz 4Vpp + 1V offset waveform - like around a volt.  Right up until I hit the Statistics Reset button.  Then I'm down around 20 mV on Vpp and 3.3 mV on Vrms.  I think that Reset button is a really big deal if the samples start before the generator even starts sending a signal.

ETA:  I think the Siglent User Manual does am award winning lousy job on statistics.  It is essentially worthless.

I suspect the reason it is so poor is that nobody uses these functions so there is no pushback on the manual.  The few people who actually use it probably figure it out by experimenting.

I highlighted the sentence in bold, because I fully agree with it; whoever buys an instrument, would have the right to have an exhaustive manual; I've gotten to 160 pages by now, but if it wasn't for the forum I wouldn't have understood almost anything from the manual. Now I do other tests to understand this Stdev value: however, I did not use statistics, but all measure enabled.
But I have a Clear button to use .. I'll try .. let's see what happens

note: ok, maybe users don't care about the Stdev value or what it means, but if it is displayed a better explanation would be useful.

[CharlotteSwiss]

It seems that in the all mesaure window (large window with all measurements) the Stdev value is approximately equal to the RSM one ... while if I enable Stdev in the statistics (see at the bottom of the display) the Stdev value is rightly low: in the screen in high is with dc offset, and the value is updated during a count (I guess you count every signal acquisition on the display; the value started from 300mV, while as you can see in the image after 290 counts it dropped to 87mV .. dropped again ..
Note in the statistics of the Stdev: current value they say 1.52v (which is the value reported in the measure), but then in the st dev column they put 87mV: maybe I don't understand, but it seems bizarre to me!

In the lower screen the signal is Av coupling without dc offset, it marks a Stdev value of 0 in the statistics

I add: Doing other tests, the Stdev value, on the same signal, gives me about 200uV ... however this seems the right value, certainly not what it reports in all measure, or as current.

[CharlotteSwiss]

always in the field of measurements, the values reported as "cycle .....", what period do the signal refer to?
From the manual they say:
Cycle mean: Average of data values in the first cycle.
Cycle RMS: Root mean square of all data values in the first cycle.

But first cycle, what period do they mean on the display? would I say that just after the trigger point?
 

[rstofer]

Check out the rise and fall times.  I was talking about this earlier in the thread but I don't remember the context.  In any event, the compensation signal has relatively poor rise and fall times hence limited harmonics.

On my Rigol, the rise time is 6 us and the fall time is 8 us - not especially fast!  In fact, a lot slower than yours...

I would think it would be the first cycle after the trigger.  I don't see how std dev would work on a single cycle, it probably doesn't.  But the measurements would.

[rstofer]

Try first cycle and single shot mode, see what happens.

[CharlotteSwiss]

Try first cycle and single shot mode, see what happens.

tomorrow I try, even the advice of the post just before; I saw that there are gate cursors also for measurements, so I can narrow down where I want the detection, it might be useful to understand what they mean ..
Now my eyes are closing, good evening rstofer 

[CharlotteSwiss]

rise time is 1us in my compensation signal, so it makes less harmonics: however, I don't understand the connection with the Stdev value.

I tried with single trigger, each press displays the Cstdev value (some us); but am I wrong or single it shows us an acquisition and stops .. and if I repeat the next acquisition and it stops and so on? what information could you give me about the cycle?
I tried to set the gate just beyond the period time, and move on the signal, it signals the cycle values also to the right of the trigger point.
Oh well but it doesn't matter come on, I keep in mind that the Stdev value only indicates it to me in the statistics, if ever I need it ..

Out of curiosity, I report the drawings it suggests for the type of measurement Stdev (a strange geometric shape, which is actually the symbol of standard deviation) and cycle (ok indicates the first period, but it is not clear whether it is the first period on the right display, or the one after the trigger point.
 

[CharlotteSwiss]

with these two great articles I understood what Stdev is; the lower the value, the more stable the signal amplitude; it is obvious that that value in measure of stdev about = the RMS value is completely wrong, it is a bug that inserts that value; while in statistics it is more realistic.

https://www.allaboutcircuits.com/technical-articles/average-deviation-standard-deviation-variance-signal-processing/

https://www.allaboutcircuits.com/technical-articles/how-standard-deviations-relates-rms-values/

[borjam]

I highlighted the sentence in bold, because I fully agree with it; whoever buys an instrument, would have the right to have an exhaustive manual; I've gotten to 160 pages by now, but if it wasn't for the forum I wouldn't have understood almost anything from the manual. Now I do other tests to understand this Stdev value: however, I did not use statistics, but all measure enabled.
But I have a Clear button to use .. I'll try .. let's see what happens

Talented technical writers are really scarce. Just look at the outstanding documentation produced by the big, classic manufacturers such as Tektronix, HP/Agilent/Keysight and you will notice the enormous difference.

Chinese manufacturers such as Rigol and Siglent are new. Although they are maturing and their software design skills are improving there is still a long way to go. And a much longer way to go regarding proper documentation!

As for this thread, with editing it might become a sort of collaborative oscilloscope tutorial book. Your inquisitive questions have pushed everyone to do their best to write good explanations. Unfortunately most forum questions use to be a quick job. Not in this case.

[CharlotteSwiss]

Talented technical writers are really scarce. Just look at the outstanding documentation produced by the big, classic manufacturers such as Tektronix, HP/Agilent/Keysight and you will notice the enormous difference.

Chinese manufacturers such as Rigol and Siglent are new. Although they are maturing and their software design skills are improving there is still a long way to go. And a much longer way to go regarding proper documentation!

As for this thread, with editing it might become a sort of collaborative oscilloscope tutorial book. Your inquisitive questions have pushed everyone to do their best to write good explanations. Unfortunately most forum questions use to be a quick job. Not in this case.

I agree that the most famous brands have an edge; it must also be said that they have higher prices, and probably also assistance, more complete manuals ...
As soon as I get to the end of the manual, I update the first message of this thread, indicating that if the reader intends to take their first low-level DSO, they may be in the right place; the development of the discussion goes from the choice, to the first baseline experiments, to understanding all the functions of the oscilloscope; the things that I have easily learned are missing, but I believe that if I have understood them, I who have little experience, will be able to understand them all; I also believe that in the course of the discussion topics were dealt with on which many users (even experts) pass over; here, however, everything stopped, even for weeks, in search of an understanding of the subject; I think it's a plus for those who want to get inside an oscilloscope, without stopping only at the external aspect ...

a necessary note: for a beginner like me, learning a new world as the use of an oscilloscope can be was not easy; I gradually managed to understand this world thanks to many kind people who helped me; it is easy to mock a beginner who may surely ask trivial questions; but the internal values of a kind person are seen exactly in these cases, when they understand the difficulties and without rage they try to help you; it is to these people that I am infinitely grateful, and here in this discussion I should thank many (I do not mention them all because I would certainly forget a few ..)

 

[borjam]

a necessary note: for a beginner like me, learning a new world as the use of an oscilloscope can be was not easy; I gradually managed to understand this world thanks to many kind people who helped me;

Not a criticism, quite the contrary, but you have been learning upside down. 

It's much easier to learn about a complex instrument when you are trying to answer a question using it, like in the Physics classroom. "I need to measure the delay between these two events" or something like that.

it is easy to mock a beginner who may surely ask trivial questions;

Beware people who mocks trivial questions. You may find that they really ignore how to answer them. Trivial questions often require a deep understanding of the basic principles. That's why really talented teachers are so rare.

[tautech]

For the USB key bug, at the end of the manual study I will investigate better; it could be that with a certain brand / size of key it doesn't go wrong! yes the firmware in July was the last one, then I'll check if they have a new one.
I tried to ask in this topic, even if they talk about 4-channel models ..
https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/1700/

The oscilloscope when I turn it off, it keeps the settings, but if I turn it off regularly; with this bug it freezes ... and forcing the shutdown does not save the configuration, but keeps the previous one ...
The trick now is to do a Print before making a configuration of which I want to save the screen .. if it works there it also works later .. but I don't always remember ..
 

Just did 40, yes 40 screenshots with a new SDS1202X-E without issue. 
Fast presses without waiting for the saving message to disappear or slow presses made no difference. Several different usage modes and settings too and all without error.
It may be the USB stick you use, mine are all formatted in FAT32 and this one is a Strontuim 8GB that I've had for some years and I only bought them as they were very cheap, just a few $ ea IIRC.
All I can suggest is you try some other brands of USB sticks.