what an oscilloscope recommended for a woman passionate about electronics?

[CharlotteSwiss]

work excluded of course, I am for social distancing to the bitter end ... 

[rstofer]

thanks rstofer, in the meantime I analyze d / dt (as soon as I archive this I read well your intervention 674 for the integrated)
I'm starting to understand something, and I also learned what slew rate is in op.amp datasheets
If I look at your test with the Rigol, I see that if you enable the diff function, the display shows you the value that is 1MV / s, and everything makes sense.
Now look at my attached image: usual compensation signal, as you can see the square wave values are 3.12v and a rise time of 984ns.
As you taught me, the d / dt should be: 3.12v / 0.000000984s = 3170731V / s
As you can see, the only data he shows me related to d / dt are the vertical divs (100kv / s) and the position on the display.
What good can this view be for me, if it doesn't even tell me the d / dt value? or maybe I'm missing something?
In your Rigol it makes sense, you enable the diff and it indicates the V / s relative to the ramp rise time of the signal ... but in my Siglent what do I do with this d / dt graph?

I thought your rise time on that compensation signal was a few microseconds, nowhere near nanoseconds.  Are you sure?

Notice how your dv/dt trace is broad at the bottom leading and trailing edge.  Assuming there is any basis in reality, that tells me that the leading and trailing edges of the source are rounded, the derivative is not yet very high.  This would be consistent with a slower rise time.  I would expect to see a spike just one pixel wide for a very fast rise time because the derivative dV/dt gets very high when dt approaches zero.  In fact, that's what this is all about.  dt  - how fast is something happening.

In the Integral post, I talk about taking the integral of sin(t) and getting -cos(t).  Here, with the derivative function, if you plug in sin(t), you should get cos(t) so you should see a couple of traces like my integral image except the cos(t) will be rightside up.  It will start at a high positive value and swing down to a matching low value.

I didn't get a useful display on my Rigol because the sine wave coming in isn't a pure sine, it is a synthesized sine.  As a result, there is some noise as it changes steps and the derivative function sees those steps as a change in voltage over time and posts it as a derivative.  Mathematically, this is correct and it is also the reason that analog computing doesn't use differentiators, they emphasize the noise.

Even though analog computing is used to solve (or model) differential equations, they don't use differentiators.  Lord Kelvin came up with the answer and his genius shows up in (1) and (2) in this paper.  Integrate until the derivatives disappear!  Then close the feedback loop.  This is so much easier than trying to come up with a solution using a slide rule!  How come they don't teach this stuff in college?

http://chalkdustmagazine.com/features/analogue-computing-fun-differential-equations/

Using the diff function on a scope is one place where I wish I had a 'pure' sine wave.

[CharlotteSwiss]

http://chalkdustmagazine.com/features/analogue-computing-fun-differential-equations/

while I see the compensation signal for a moment and its rise time .. I console myself with one of the first sentences of your link ...

When it comes to differential equations, things start to get pretty complicated—or at least that’s what it looks like. When I studied mathematics, lectures on differential equations were considered to be amongst the hardest and most abstract of all and, to be honest, I feared them because they really were incredibly formalistic and dry. This is a pity as differential equations make nature tick and there are few things more fascinating than them.

it is no coincidence that I have been stopped for 10 days between FFT, diff etc ... 

[rstofer]

It scared me when my grandson had to take Differential Equations last semester.  I struggled with it back in college because, among other things, we didn't have much beyond a slide rule to work with them.  Today it's all done in MATLAB using the ODE45 solver or even programming up a version of Euler's solution.  Its almost trivial.  When I took the course, I did have access to a computer and plotter.  What I didn't have, and what wasn't presented, was a computer method for solving.

I would recommend spending time with Desmos.com.  Attached is a plot of sin(x) and 2 * sin(x + pi/4).  In other words, the second trace has twice the amplitude and a 45 degree leading phase shift.  Sooner or later, this stuff comes up and pictures are helpful.

So, how did I decide that the green (leading) trace is leading?  It gets to a maximum 45 degrees sooner than the red trace.

[CharlotteSwiss]

meanwhile, I confirm that the rise time of the compensation square wave ramp is around 1us.
See the image, I placed the cursors (10% and 90% of amplitude) as you can see the rise value of the measurements detects 1.01us, with the cursors I detect about 1.040, so I would say it is really around 1us

[rstofer]

Yes, that's the rise time as typically defined.  The thing is, the derivative function starts a 0V and you can see where the upper 10% is quite rounded.  As far as the derivative is concerned, things start changing earlier and end much later and dt is therefore a longer period and the MV/dt is less

[CharlotteSwiss]

It scared me when my grandson had to take Differential Equations last semester.  I struggled with it back in college because, among other things, we didn't have much beyond a slide rule to work with them.  Today it's all done in MATLAB using the ODE45 solver or even programming up a version of Euler's solution.  Its almost trivial.  When I took the course, I did have access to a computer and plotter.  What I didn't have, and what wasn't presented, was a computer method for solving.

your grandson is lucky to have a grandfather so that he follows him in his studies!
I also tried a graphic designer, I did not know this site, nice! 

[CharlotteSwiss]

Yes, that's the rise time as typically defined.  The thing is, the derivative function starts a 0V and you can see where the upper 10% is quite rounded.  As far as the derivative is concerned, things start changing earlier and end much later and dt is therefore a longer period and the MV/dt is less

ok I understand, but to recap:
your Rigol if you run diff, it makes you a d / dt value on the display, so it has its usefulness. But maybe you would prefer to calculate that d / dt manually based on the amplitude of the signal and a certain ramp time detectable with the oscilloscope?

My siglent apparently makes me just a d / dt chart, doesn't it seem to me that looking at this chart can help in any way? So should I proceed to manually calculate the d / dt value?

I downloaded the pdf of this review (siglent 1104, but the diff function seems the same to me)
https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-in-depth-review/
3° post /1* pdf
where for the differential it states:

This is a particularly cumbersome math function that just cannot work well on an 8 bit system. It has a
parameter dx, which determines the interval used for the difference computation. For an accurate and
detailed result, we would want dx to be as small as possible and the lowest value that can be set is 0.02
div. With this setting, we can get totally useless results, as we will see later.

then oh well, from the following examples it is not very clear what you mean, except that the graph on a square wave setting 0.02 right is nice to see ..

Go and make myself dinner, then I'll resume my homework ..
thanks rstofer

[borjam]

When it comes to differential equations, things start to get pretty complicated—or at least that’s what it looks like. When I studied mathematics, lectures on differential equations were considered to be amongst the hardest and most abstract of all and, to be honest, I feared them because they really were incredibly formalistic and dry. This is a pity as differential equations make nature tick and there are few things more fascinating than them.

it is no coincidence that I have been stopped for 10 days between FFT, diff etc ... 

Differential calculus is much easier to understand if you approach it from Physics. While it won't be so rigurous, it will become much more intuitive. That's why explaining the meaning of the Fourier transform using sound makes it much easier to understand.

[rstofer]

When it comes to differential equations, things start to get pretty complicated—or at least that’s what it looks like. When I studied mathematics, lectures on differential equations were considered to be amongst the hardest and most abstract of all and, to be honest, I feared them because they really were incredibly formalistic and dry. This is a pity as differential equations make nature tick and there are few things more fascinating than them.

it is no coincidence that I have been stopped for 10 days between FFT, diff etc ... 

Differential calculus is much easier to understand if you approach it from Physics. While it won't be so rigurous, it will become much more intuitive. That's why explaining the meaning of the Fourier transform using sound makes it much easier to understand.

That's probably why Calc I and Physics I are taught in parallel.  We can  intuitively understand displacement, velocity and acceleration from our daily livesm(especially if you were a teenager in the age of muscle cars).  We don't often talk about the 3rd and 4th derivatives, jerk and jounce (snap).  I had never heard of the 4th derivative being called jounce but it sounds way cooler than snap.  Following snap is crackle, pop, lock and drop (I have no idea what these represent other than the 5th through 8th derivatives but I suppose they need to be called something).

There's a table of integrals and derivatives of displacement at the bottom of this page:

http://wearcam.org/absement/Derivatives_of_displacement.htm

[rstofer]

I downloaded the pdf of this review (siglent 1104, but the diff function seems the same to me)
https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-in-depth-review/
3° post /1* pdf
where for the differential it states:

This is a particularly cumbersome math function that just cannot work well on an 8 bit system. It has a
parameter dx, which determines the interval used for the difference computation. For an accurate and
detailed result, we would want dx to be as small as possible and the lowest value that can be set is 0.02
div. With this setting, we can get totally useless results, as we will see later.

Nice find!  It may turn out that some of the functions just work for very specific examples and are not generally useful.  Mostly I just use the scope for looking at squiggly lines - waveforms.  Maybe someday I'll have an application for the more exotic math functions but I'm not counting on it.

I'm sure there is a lot to learn from that review.  It is the longest running review I have ever seen.

[CharlotteSwiss]

Differential calculus is much easier to understand if you approach it from Physics. While it won't be so rigurous, it will become much more intuitive. That's why explaining the meaning of the Fourier transform using sound makes it much easier to understand.

I've been there 10 days on the FFT, but now I understand it pretty well .. obvious mainly thanks to your help here on the forum
thank you 

[CharlotteSwiss]

Nice find!  It may turn out that some of the functions just work for very specific examples and are not generally useful.  Mostly I just use the scope for looking at squiggly lines - waveforms.  Maybe someday I'll have an application for the more exotic math functions but I'm not counting on it.
I'm sure there is a lot to learn from that review.  It is the longest running review I have ever seen.

In fact I will probably never use this function, but for a personal nature I always like to learn all the functions 

did a great job with that review, I found it quite by accident, and I've already viewed various things about it; they are quite similar as menus, indeed almost identical ... But for the diff it doesn't add much beyond what I already know; I don't think they did it just to see a graph representing d / dt; in the review he puts some examples / screenshots, but only to highlight that the graph is not nice to see, but no hint of what calculation / value this function can communicate to us! mystery! in the meantime I look at your speech on the integrated
 

[CharlotteSwiss]

I add a screenshot of the manual on the d / dt function: the oscilloscope executes the formula reported in the manual, I can try to understand what value it makes, and if that value is in one of those reported in the math measurements

[CharlotteSwiss]

I found this topic concerning rigol 1054, it seems that the diff function is also unusable in rigol; curious that even there they say that if you change the sec / div of the source signal, the amplitude v / s of the diff signal changes.

https://www.eevblog.com/forum/testgear/rigol-ds1054z-how-to-get-_maximum_-positive-slew-rate/

ok, I'll put an x on this d / dt function, I won't use it; if I need to know the ramp speed of a square wave signal, I will use the formula: Vpp / rise time

Well, now I can go to bed relaxed, tomorrow I'll think about the integral
 

[rstofer]

The equation is used to find the slope (derivative) at point y(i) by drawing a line through points on either side spaced a delta-t away. Since these points are 2 delta-t apart, the result is divided by 2 and of course, delta-t itself.

[Mechatrommer]

At my age (74) mid afternoon naps are becoming more common.  Mid morning naps are popular as well.  Plus another nap while watching TV before bed time.

dont you regularly go to Wallmart/McDonalds anymore?

If it fits in between my naps...

I have been retired, and comfortably, for 17 years.  I get to play, and nap, as much as I want and never have to commute.
Social distancing hasn't been a problem, I've been doing it since I retired.

so you've been scamming us all these years? how dissapointing i've been saving money to buy plane's ticket to your place in hope i can get some good and joyful lesson, cheap wallet an all... now what i'm gonna do?  anyway i hope you are doing well..

[rstofer]

At my age (74) mid afternoon naps are becoming more common.  Mid morning naps are popular as well.  Plus another nap while watching TV before bed time.

dont you regularly go to Wallmart/McDonalds anymore?

If it fits in between my naps...

I have been retired, and comfortably, for 17 years.  I get to play, and nap, as much as I want and never have to commute.
Social distancing hasn't been a problem, I've been doing it since I retired.

so you've been scamming us all these years? how dissapointing i've been saving money to buy plane's ticket to your place in hope i can get some good and joyful lesson, cheap wallet an all... now what i'm gonna do?  anyway i hope you are doing well..

I just noticed your Malaysia flag.  Back in '88-'89, I was living and working in Singapore building a semiconductor plant.  I took up SCUBA diving and made 13 trips to Malaysia that year - several to the resort on Pulau Tioman but more often just sleeping on the beach at Pulau Aur and Pulau Dayang.

I also spent a week on a live-aboard sailboat diving off the Similan Islands in Thailand and another week living in a hut and diving at the Maldives.

I spent a LOT of time in the water but the Malaysia trips were the best.  The dive trip cost $25 for the boat ride and the compressed air.  I had a car and each additional diver paid $25 for the round-trip ride.  One extra diver and I dove for free!  If I carried two divers it would cover my gas.  It cost me absolutely nothing to spend the weekend diving.  The American School chef prepared the food and it was great!  The Australian and New Zealand embassies had a contest about wine and they provided 5 liter bags of wine - free.  Accommodations were simple:  pitch a tent on the beach.

No wonder I loved the place!  I will never forget my time in Singapore, Malaysia, Thailand and the Maldives.

[CharlotteSwiss]

Skipping ahead to the integral, the scope does the obvious calculus things.

The first image shows the integral of a constant and then what happens when the constant changes - a square wave.  The integral (c dt) is c*t where t is the time over which the integration occurs.  As time goes by, the product c*t increases linearly - this results in a ramp.  In the product, the only thing that is changing is linear time, the constant is, well, constant.  When the constant changes, the ramp heads in the other direction.  This is integration at its easiest and is the basis for analog computing.

Pour a constant 5 gallons per minute into a 50 gallon barrel and plot the height of the water over time up to 10 minutes when it overflows.

The second image shows the integral of the sin(x) function (waveform) and the integral(sin(x)) is -cos(x) plus a constant which I will ignore.  Note the middle of the screen where the yellow sin(x) crosses 0V going positive and the blue is at its minimum value (not 0 but some negative value).  In a perfect world where the sin(x) varied from 0 to +1 to 0 to -1 and back to 0, the cos(x) would vary from +1 to 0 to -1 to 0 and back to +1 and -cos(x) would vary from -1 to 0 to +1 to 0 and back to -1.  I don't know why my Rigol isn't displaying the magnitude of -cos(x) properly but the idea is right.  It shows a cosine but the upper and lower values are wrong.  I'll probably have to read the manual.

See the screen image from Desmos.com  The red trace is sin(x) and the green trace is -cos(x).

Here is a table of simple integrals from Calculus II:

https://www.mathsisfun.com/calculus/integration-rules.html

In any event, integration may be important.  The voltage on a capacitor is 1/C * integral(i dt).  This makes sense because a larger capacitor takes  more current (or more time) to charge to the applied voltage.  Double the size of the capacitor and it takes twice as long to charge to some voltage given an identical current.  Again, this makes sense.

You can think of an integrator (or capacitor) as a bucket that accumulates something - perhaps charge.  Or the pressure in an air compressor tank.  When empty it starts at 0 psi and as the pump runs, more air is accumulated in a fixed volume so the pressure increases.  Or the barrel of water described above.

I started studying the integral: in the meantime I started with the square wave compensation signal (AV coupling) to have an integral trace like your first example.
For now I leave out the offset and Gate settings in the integral menu ..
I see that the integral simulates the trend of the voltage in the source signal: the line is rising from when the edge rises and until it begins to fall .. and downwards from when it falls until it returns to rise ..
Can we say that this graph can help to check if a signal has a stable / smooth rise / fall voltage?
Based on the vertical div V * s, the integral ramp measures 820uV * s, so can we say that the charge strength of this signal is this value?
Then slowly I also see other examples and try to understand what offset and Gate mean (totally unnamed gate on the manual)
thanks 

[CharlotteSwiss]

The second image shows the integral of the sin(x) function (waveform) and the integral(sin(x)) is -cos(x) plus a constant which I will ignore.  Note the middle of the screen where the yellow sin(x) crosses 0V going positive and the blue is at its minimum value (not 0 but some negative value).  In a perfect world where the sin(x) varied from 0 to +1 to 0 to -1 and back to 0, the cos(x) would vary from +1 to 0 to -1 to 0 and back to +1 and -cos(x) would vary from -1 to 0 to +1 to 0 and back to -1.  I don't know why my Rigol isn't displaying the magnitude of -cos(x) properly but the idea is right.  It shows a cosine but the upper and lower values are wrong.  I'll probably have to read the manual.

See the screen image from Desmos.com  The red trace is sin(x) and the green trace is -cos(x).

in the meantime I take a step back: I am attaching a picture of Desmos.

sin (x): draws a perfect sinusoid, that is, at zero time it passes through zero and ramps up

cos (x): draws a sinusoid which at time zero starts from the max amplitude value and then falls

-cos (x): draws a sinusoid that at time zero starts from the minimum amplitude value and then rises

The question that comes to me would be: but in reality what does zero time represent in a circuit where a signal like these flows? Or when can I say, looking at a sine wave with the oscilloscope, if this is sin, cos or -cos?
Is the horizontal center of the display taken as zero time?
I imagine for example that the alternate of the home network is a sin (x) right?
thank you 

[indeterminatus]

The question that comes to me would be: but in reality what does zero time represent in a circuit where a signal like these flows? Or when can I say, looking at a sine wave with the oscilloscope, if this is sin, cos or -cos?

You can shift any one of those function graphs to the left or to the right to basically fully overlap with another (their pattern is the same, just their starts are different). Unless you have a reference point that serves as "time 0", there is no way to distinguish these three from one another, so you could pick any one of them and be right.

AC power voltage that you could theoretically measure from your power outlets (disclaimer: please don't, if you do not know exactly what you're doing) closely follows a sine wave (or cosine wave, they're basically identical). "closely" because even though in theory the generation at the very start of the power line would be a true sine wave, much happens in between (until your outlet is reached): inductance and capacitance from all kinds of components and whatnot is super-imposed onto that 50 Hz base frequency. If you looked at it with a 'scope, I suspect you'd see some "wiggling" and high-frequency content.

[edit: mixed up power and voltage ]

[rstofer]

You are right!  Continuous time signals don't have a start or stop time nor is there something like t=0.  The thing is, we have problems analyzing signals that started with the big bang and worked forward.

The difference between sine and cosine is you pick one, call it sine and the other will be cosine.  It gets messy!  See 2d paragraph of page 5 here where an exponential form is converted to a complex trigonometric form using Euler's Formula

https://www.math.columbia.edu/~woit/eulerformula.pdf

I really and truly don't want to go there.  ALL of those brain cells were fried a long time ago.

I decided I wasn't happy with the previous integration examples so I made another one.  This time I used cursors to limit the area over which I would integrate (forming a definite integral) and I presented a 1 kHz square wave with 1Vp-p and 500 mV offset such that the signal varies between 0V and 1V.  The integration period is just 500 us so the area under the rectangle is 1V * 500 us or 500 uV s.  Reading from the right edge of the blue trace, we see it is at 5 divisions of 100 V us or 500 uV s.    The value at the far right on the integral does, in fact, give the area under the square wave over the limited period of time.  I love it when a plan comes together.

I think the Gate control on your scope does the same thing as my cursors and the only way you know I applied cursors (Start Time and End Time on the Rigol) is by the fact that the blue trace only exists during integration.

[rstofer]

I wasn't happy with the previous discussion on the integral of sin(x).  First, we know it should look like a negative cosine but, over the definite integral period of 0 to 2*pi, the integral is 0.  Take a symmetric sine wave and look at the area above the X axis and below the X axis.  They are equal and opposite sign so they cancel each other out.  At multiples of 2pi, the integral is 0.

So, here is a picture of a 1 kHz 1Vp-p sine wave and a definite integral over 1 period.  Note that the value at the far edge of the integral is 0 - just exactly what we expect.  The high point of the -cos(t) occurs when the sin(t) gets back down to 0V - that is because, up until that time, we were still pouring water in the bucket.  Once the sin(t) turns negative, the water starts getting drained out.

I like this "Start Time" and "Stop Time" on the integration function.  It makes everything match the math.

[CharlotteSwiss]

You can shift any one of those function graphs to the left or to the right to basically fully overlap with another (their pattern is the same, just their starts are different). Unless you have a reference point that serves as "time 0", there is no way to distinguish these three from one another, so you could pick any one of them and be right.

AC power voltage that you could theoretically measure from your power outlets (disclaimer: please don't, if you do not know exactly what you're doing) closely follows a sine wave (or cosine wave, they're basically identical). "closely" because even though in theory the generation at the very start of the power line would be a true sine wave, much happens in between (until your outlet is reached): inductance and capacitance from all kinds of components and whatnot is super-imposed onto that 50 Hz base frequency. If you looked at it with a 'scope, I suspect you'd see some "wiggling" and high-frequency content.

thanks for the explanation, very helpful 

Ok I will try not to be electrocuted on a home socket 
in general, however, I am very cautious with the current
but one more notice is always better
 

[CharlotteSwiss]

You are right!

it's not often that I'm right 
now 2 days of work await me, many people are in need.
I resume the study from Monday in the late evening (reading well your interventions 700 and 701)
In the meantime, thank you for everything and have a good weekend
 
Char