what an oscilloscope recommended for a woman passionate about electronics?

[tggzzz]

I am happy that DSO do many mathematical calculations: this, however, also has a sad dark side, that is, machines replace our brains.

Very true; you have the right attitude.

If you don't understand the mathematics and the concepts, then the arithmetical calculations are pointless - and probably misleading. Just using a convenient tool without understanding is cargo cult science or voodoo magic.

If we look at modern children, they no longer make any mental calculations, they use the cellphone calculator ...

Not necessarily true.

When my daughter was being taught arithmetic (~2000), I was impressed that they were no longer taught "the single true way" to do, for example, multiplication.

I was taught the times tables up to 12 (because there were  twelve pence in one shilling!), and beyond that it was the long multiplication algorithm.

My daughter was taught to visualise the numbers and how they fitted together, so that she could break a particular calculation into whichever simpler calculations were helpful. Thus 20*18 might be decomposed into 2*2*100-2*20. That gives a better "feel" for numbers than the simple algorithm, and enables mental "sanity checking".

OTOH, using calculators for normal operations is sensible; why do work that is better done by machines - provided you do understand what the machines are doing, and that what they are doing is sensible.

[CharlotteSwiss]

tutit 

tautech is shortened from my company name being the district Taupaki where I live in NZ and Technologies but if you insist call me tutit or even better just plain Rob.
When I have time I drop in here and offer screenshots from very similar DSO's so it's easier for you to understand.
I have been using and selling these X-E's for some years and know them very well so it's easy for me to provide good examples as your thread develops.

You must know this is already a well followed thread as you are not the only novice working hard to understand what you see on a scope display and why.
You like many others before you have come a long way from opening the box, pressing the power button and the Autoset and threads like this really help other novices too in their understanding of circuits and scopes.

Please carry on. 

it had to be ...you are all very kind and helpful, thank you 
it was not a nickname, but all that the translator has miserably wrong 

of course I continue, until I get to the last page of the manual. On the forum, however, I will continue even beyond this manual, after all I am passionate about electronics, even if I still understand too little about it.
Thank you
 

[CharlotteSwiss]

Very true; you have the right attitude.

If you don't understand the mathematics and the concepts, then the arithmetical calculations are pointless - and probably misleading. Just using a convenient tool without understanding is cargo cult science or voodoo magic.

If we look at modern children, they no longer make any mental calculations, they use the cellphone calculator ...

Not necessarily true.

When my daughter was being taught arithmetic (~2000), I was impressed that they were no longer taught "the single true way" to do, for example, multiplication.

I was taught the times tables up to 12 (because there were  twelve pence in one shilling!), and beyond that it was the long multiplication algorithm.

My daughter was taught to visualise the numbers and how they fitted together, so that she could break a particular calculation into whichever simpler calculations were helpful. Thus 20*18 might be decomposed into 2*2*100-2*20. That gives a better "feel" for numbers than the simple algorithm, and enables mental "sanity checking".

OTOH, using calculators for normal operations is sensible; why do work that is better done by machines - provided you do understand what the machines are doing, and that what they are doing is sensible.

to think well, are the older children who forget to have the brain and compensate with the calculator.
If the machine does the math for me, ok I accept it, but I have to understand the reason for those math. I hope it is evident in this discussion of mine, that I try to go beyond the calculation of the machine: otherwise some of my curiosities are not explained ...
 

[tggzzz]

If the machine does the math for me, ok I accept it, but I have to understand the reason for those math. I hope it is evident in this discussion of mine, that I try to go beyond the calculation of the machine: otherwise some of my curiosities are not explained ...
 

I suspect the reason people are helping you is because that is evident

[CharlotteSwiss]

Compensation signal siglent - DC coupling - ok the average value was about 1.50vdc (also measured with multimeter at the signal socket).
Since I ate bread and fox this morning, I thought that setting AV coupling made me an average value of zero volts on the signal socket. 



then I woke up my head, and I realized that the cut of the DC voltage happens then, in the circuit inside the oscilloscope. So it is right that the multimeter always measures me a dc value of 1.50v.
I see, however, the oscilloscope detects an average value of about -9mv: this perhaps means that the DC current is blocked, but not 100%

[tggzzz]

then I woke up my head, and I realized that the cut of the DC voltage happens then, in the circuit inside the oscilloscope. So it is right that the multimeter always measures me a dc value of 1.50v.
I see, however, the oscilloscope detects an average value of about -9mv: this perhaps means that the DC current is blocked, but not 100%

Scopes are not precision instruments; they are designed to show the shape of rapidly varying signals.

The capacitor inside the scope will block the DC completely.

The 0.6% difference will be due to measurement errors inside the scope, and could be due to many causes.

There might be a DC offset inside the scope's analogue front end. Check that by turning off the signal generator, and see if the reading changes.
It might be that the duty cycle isn't exactly 50%.
The ADC in the scope has a limited vertical resolution (i.e. bits), and it will have a slope and offset and non-linearity errors. Compare 0.6% and 9mV with the scope's specification; you will need to read carefully (see my .sig ).
There can be arithmetic "errors" due to the limited number of samples and vertical resolution. Some scopes measure what they display on the screen, not all the samples in its internal memory.

Summary: it is worth doing a few quick tests to rule out gross errors, but that kind of "error" looks to be within normal behaviour.

[CharlotteSwiss]

Scopes are not precision instruments; they are designed to show the shape of rapidly varying signals.

The capacitor inside the scope will block the DC completely.

The 0.6% difference will be due to measurement errors inside the scope, and could be due to many causes.

yes I know that they are not precision instruments, their work is another. Measuring with the multimeter beyond that capacitor, it should mark 0vdc.
The purpose of my intervention was not those -9mv, but having understood that the compensation output signal always makes the same average value (1.51) both with dc and with AC.
 

[tautech]

2 things to do to improve accuracy.
Ensure probes are perfectly compensated with zero overshoot, Charlotte's are not perfect.
Run Autocal in the Utilities menu after some 30 minutes warm up.

Further, any accurate measurement requires the scope to be powered ON for a while so any/all internal component temperature drift has stabilized and matches the last run of Autocal.

[CharlotteSwiss]

2 things to do to improve accuracy.
Ensure probes are perfectly compensated with zero overshoot, Charlotte's are not perfect.
Run Autocal in the Utilities menu after some 30 minutes warm up.

Further, any accurate measurement requires the scope to be powered ON for a while so any/all internal component temperature drift has stabilized and matches the last run of Autocal.

I'm honest, when I did the calibration I acted after a few minutes that I had turned on the siglent.
In the first part of the manual I am studying, I have already come across these 2 items: Do self cal and quick-cal.
I have not delved into qaunto I will get there later when the manual will explain them better.
However, one question is spontaneous: the probe calibration procedure, reported in the very first pages of the manual, does not involve using these functions of the utily menu.
Are these two functions for further, more precise calibration?

[tautech]

2 things to do to improve accuracy.
Ensure probes are perfectly compensated with zero overshoot, Charlotte's are not perfect.
Run Autocal in the Utilities menu after some 30 minutes warm up.

Further, any accurate measurement requires the scope to be powered ON for a while so any/all internal component temperature drift has stabilized and matches the last run of Autocal.

I'm honest, when I did the calibration I acted after a few minutes that I had turned on the siglent.
In the first part of the manual I am studying, I have already come across these 2 items: Do self cal and quick-cal.
I have not delved into qaunto I will get there later when the manual will explain them better.
However, one question is spontaneous: the probe calibration procedure, reported in the very first pages of the manual, does not involve using these functions of the utily menu.
Are these two functions for further, more precise calibration?

Self Cal is a manually started internal self calibration and it's a good idea to run it after large temperature changes like winter to summer if you are not operating in A/C building.

Quick Cal is an automatic always ON running self adjustment to correct for temperature changes while the DSO is operating. It can result in brief pauses of responsiveness to controls however we can turn this feature OFF.

Probe compensation is a manual process we must undertake to give the best waveform step response on the display. (square edge with zero overshoot/undershoot.)
It does not matter the settings we use other than the probe compensation waveform must be completely on the display and I find 100us is the setting I use most to get accurate/best probe compensation adjustments.

The tiny changes of all/any of the above can impact on measurement accuracy.

Some example of how improper probe compensation can impact measurement accuracy.....note the P-P.

[dietert1]

I would check first  whether there is a specification for the calibration signal accuracy. There may be none and it's a bit misleading to assume an accuracy in the mV or uV.

Automatic characterization and compensation ot the scope signal chain as a whole is available in professional scopes. Separate probe and scope adjustment as used by simple scopes does not work very well at high frequencies above let's say 5 GHz. As someone wrote above: When you adjust the probe, you assume that the calibration signal remains constant, which is an approximation for low frequencies.

By the way: Does the calibration signal frequency change with timebase? I remember the Tektronix 2465 had that feature. The calibration signal could be used to check timebase calibration at all deflection speeds. Maybe for a digital scope that is less important.

Regards, Dieter

[CharlotteSwiss]

Self Cal is a manually started internal self calibration and it's a good idea to run it after large temperature changes like winter to summer if you are not operating in A/C building.

Quick Cal is an automatic always ON running self adjustment to correct for temperature changes while the DSO is operating. It can result in brief pauses of responsiveness to controls however we can turn this feature OFF.

Probe compensation is a manual process we must undertake to give the best waveform step response on the display. (square edge with zero overshoot/undershoot.)
It does not matter the settings we use other than the probe compensation waveform must be completely on the display and I find 100us is the setting I use most to get accurate/best probe compensation adjustments.

The tiny changes of all/any of the above can impact on measurement accuracy.

Some example of how improper probe compensation can impact measurement accuracy.....note the P-P.

taut you have attached two images? the second pngxxxx8 I don't see it.. 

thanks for the explanation, when I get to the self and quick paragraphs I will have less problems understanding: self is a manual calibration to be implemented if the environment changes temperature; quick is an auto calibration that adjusts itself based on temperature.
But when you advised me to do a self-adjustment, what function were you talking about? quick cal?
Or did you advise me to simply repeat the probe calibration as per the first pages of the manual? (10x - 1000Hz compensation signal - auto setup button - probe compensation via the small screw)

currently this is how my probe (yellow) is calibrated:



in your example Vp-p bounces about 0.34v, my signal bounces only 0.08v.
Should I reach a Vp-p equal to the amplitude value?
or can we also have a small range of wave angle deviation?

 

[tautech]

I would check first  whether there is a specification for the calibration signal accuracy.

There isn't.

There may be none and it's a bit misleading to assume an accuracy in the mV or uV.

Yes of course however the probe compensation signal and how well a probe is compensated is only given as an example of how an improperly compensated probe can effect measurement accuracy.

Automatic characterization and compensation ot the scope signal chain as a whole is available in professional scopes. Separate probe and scope adjustment as used by simple scopes does not work very well at high frequencies above let's say 5 GHz.

5 GHz  good 500 MHz probes have HF compensation adjustments.

As someone wrote above: When you adjust the probe, you assume that the calibration signal remains constant, which is an approximation for low frequencies.

Yes as most scope work is LF the common 1 KHz probe compensation signal is adequate.

By the way: Does the calibration signal frequency change with timebase?

Not on any low price DSO's that I'm aware of.

I remember the Tektronix 2465 had that feature. The calibration signal could be used to check timebase calibration at all deflection speeds.

Really, that's very cool. 
Still for it to be properly useful it too would need to be kept in calibration.

Of the good few CRO's I've fixed, used then flogged very few had provision to adjust the timing or amplitude of the probe compensation signal however we still used it as a sanity check.
I always yearned for one of the classic old Teks with the probe compensation loop that doubled as a current loop for setting/checking current probes.
Now we can buy these deskew fixtures the yearning has vaporized.

[tggzzz]

Some example of how improper probe compensation can impact measurement accuracy.....note the P-P.

The probe compensation can definitely affect the peak-peak value. If it didn't then there would be something seriously wrong with the measurement!

With any remotely competent calibration signal the positive edge and negative edge should be identical.
With any remotely competent scope compensation circuit, the various capacitances should not be voltage dependent. (If they are, then you have other major problems!)
Given those, the positive compensation error (i.e. area under the overshoot or undershoot) should be identical.
In that case the mean value of an AC coupled signal should still be zero.

The calibration signal's duty cycle will affect the mean value. That would not affect the utility of the calibration signal when compensating a probe. On an old analogue scope I wouldn't expect a calibration signal to have exactly 50% duty cycle, but on a modern scope with lots of digital logic it is easier to generate a 50% duty cycle than not.

[CharlotteSwiss]

I would check first  whether there is a specification for the calibration signal accuracy. There may be none and it's a bit misleading to assume an accuracy in the mV or uV.

Automatic characterization and compensation ot the scope signal chain as a whole is available in professional scopes. Separate probe and scope adjustment as used by simple scopes does not work very well at high frequencies above let's say 5 GHz. As someone wrote above: When you adjust the probe, you assume that the calibration signal remains constant, which is an approximation for low frequencies.

By the way: Does the calibration signal frequency change with timebase? I remember the Tektronix 2465 had that feature. The calibration signal could be used to check timebase calibration at all deflection speeds. Maybe for a digital scope that is less important.

Regards, Dieter

it seems to me that the calibration signal has a constant frequency of 1khz regardless of the time base.

I would like to point out that I do not demand a millesimal calibration, I am a beginner: I have to worry more in understanding the functioning of the signals (some MV will not change my life)
thanks Dieter 

[tautech]

taut you have attached two images? the second pngxxxx8 I don't see it.. 

Fixed.

thanks for the explanation, when I get to the self and quick paragraphs I will have less problems understanding: self is a manual calibration to be implemented if the environment changes temperature; quick is an auto calibration that adjusts itself based on temperature.

Both help keep the instrument in top condition and providing best accuracy.
How you use them is entirely up to you however if we do a firmware update always run the SelfCal.

But when you advised me to do a self-adjustment, what function were you talking about? quick cal?

The manually started Self Cal. Maybe run it 2x/year or when you get results you think may be wrong. It takes a few minutes to complete so good to run when you need a coffee. 

Or did you advise me to simply repeat the probe calibration as per the first pages of the manual? (10x - 1000Hz compensation signal - auto setup button - probe compensation via the small screw)

currently this is how my probe (yellow) is calibrated:

Your probe compensation could be just slightly better without that tiny overshoot.
The adjustment changes very slightly as the scope warms up so do it after the scope has run some 10+ minutes.

Should I reach a Vp-p equal to the amplitude value?

The P-P includes noise (trace width) so adds to the Amplitude value.
It is of little concern other than to know why they are different. 

[tggzzz]

Automatic characterization and compensation ot the scope signal chain as a whole is available in professional scopes. Separate probe and scope adjustment as used by simple scopes does not work very well at high frequencies above let's say 5 GHz.

5 GHz  good 500 MHz probes have HF compensation adjustments.

My HP10020A 1.5GHz passive probes don't.

[CharlotteSwiss]

Fixed.

The manually started Self Cal. Maybe run it 2x/year or when you get results you think may be wrong. It takes a few minutes to complete so good to run when you need a coffee. 

Your probe compensation could be just slightly better without that tiny overshoot.
The adjustment changes very slightly as the scope warms up so do it after the scope has run some 10+ minutes.

not fixed, but not a problem 

Now I do this: I leave the oscilloscope on for at least 10 minutes, then I run self cal from utily ... and see if the compensation improves.
For now I do not intervene on the small screw of the probe.
Thanks taut 

[CharlotteSwiss]

I ran self cal (after 15 minutes of operation, the manual says to do after 30 minutes). In fact, on the signal of the yellow Ch1 something has improved (before Vp-p it was 3.10).
My surprise, however, is on the pink Ch2: it has been a long time since I used it, but I noticed that the probe is not well calibrated, at least the horizontal signal seems to be climbing. I believe I have to intervene on the small probe screw in this case.

note the difference between the yellow and pink signals

[tautech]

Now I do this: I leave the oscilloscope on for at least 10 minutes, then I run self cal from utily ... and see if the compensation improves.

It won't. It doesn't work like that. Self Cal adjusts the scope internally and the analog input that the probe is compensated to match remains unchanged.

For now I do not intervene on the small screw of the probe.

This ^ is what you must adjust.

[tautech]

note the difference between the yellow and pink signals

Yes the probe on ch2 needs better compensation adjustment.

[CharlotteSwiss]

It won't. It doesn't work like that. Self Cal adjusts the scope internally and the analog input that the probe is compensated to match remains unchanged.

This ^ is what you must adjust.

ok self cal intervenes on the internal circuit, but the final result is still to have a better signal (it certainly cannot self cal intervene on the physical calibration of the probe)
Yes, I have to intervene on the screw of probe ch2 (it has stood alone)
 

[FrostCone]

Sorry for hijacking this thread but I did not want to open a new one for a quick question.

I will also be purchasing my first scope next week and cannot decide to go with a rigol or a keysight.

Which one of these models should I go for? 
Rigol DS1074Z Plus
Rigol DS1104Z Plus
Keysight eduX1052A

EDIT: I have now decided to go for the Siglent SDS1104X-E, thank you to everyone who have replied. 

[tggzzz]

Sorry for hijacking this thread but I did not want to open a new one for a quick question.

I will also be purchasing my first scope next week and cannot decide to go with a rigol or a keysight.

Which one of these models should I go for? 
Rigol DS1074Z Plus
Rigol DS1104Z Plus
Keysight eduX1052A

Look around on this site and you will find there are already many many threads on that subject.

Before you find the answer there, I suggest you write down on a piece of paper, what you want to be able to do with the scope+probes. Hint: a scope without probes is useless (except for some high-end tasks), and using the wrong probes can be dangerous.

[CharlotteSwiss]

Sorry for hijacking this thread but I did not want to open a new one for a quick question.

I will also be purchasing my first scope next week and cannot decide to go with a rigol or a keysight.

Which one of these models should I go for? 
Rigol DS1074Z Plus
Rigol DS1104Z Plus
Keysight eduX1052A

I too had started with the idea of taking a Rigol, then on the way I chose a Siglent! Have you looked at the Siglent website and evaluated their products?