Products > Test Equipment

Help me choosing new Oscilloscope

<< < (14/16) > >>

smallfreak:
It's quite interesting how my half thought comment about the probably limited use of massive oversampling did result in a well of hardcore information about that topic  8)
Lot' to learn from here, once I picked it all apart.


--- Quote from: tggzzz on April 09, 2023, 03:31:34 pm ---Provided the sampling aperture is sufficiently small, you only need to capture at, say, 44kS/s. Downsampling is not required.
--- End quote ---

My limited knowledge about that is, that this only applies to a bandwidth limited signal of this maximum frequency. So you probably imply that this audio signal is passed through a 22KHz bandpass filter before sampling - which essentially gets rid of the 10MHz carrier/noise anyway.

I always like to refer to Monty Montgomerys great video about AD/DA and "stairsteps do not exist".

I did read quite a lot here the last view days and I stumbled across quite some pictures showing a really BAD representation of "actual waveforms" near the upper bandwidth. Likely random zigzag due to idiotically drawing straight lines between the measured points. Unfortunately I was unable to find them again quickly. This have been real screenshots of the scopes in the current 1k-1.5K range. I'm sure you all know what I mean. Maybe the scopes DO support better rendering and the people doing the screenshots simply didn't know or care.

Programmers rendering signal sample points that way should be - er - corrected. There are MUCH better ways to interpolate the values between these points that more likely show the "probably true signal within the given bandwith". Before someone renders straight lines, it's better to just display the points without any interpolation.  |O It's not that much better than the "stairsteps" representation.

Maybe that was a cheap way to overcome the problem that the vertical resolution of the screen is much greater than the vertical resolution of the common 8Bit scope. But this interpolation does not have to be done in real time at 500 MHz, but in "screen time" and just for the displayed samples. In case of hold-mode any cheap processor should be able to do this calculation without notable delay.  :-\

That made me think about the 8Bit sampling resolution again. I still have a tiny DSO112A pocket scope. This too has "8bit resolution" but that does fit well with the equally tiny screen. You don't expect much from such a device. Having to fill a 10" hi-res display with the same number of samples you certainly have to do some visual tricks to NOT make the scope look like a toy.

tggzzz:

--- Quote from: smallfreak on April 10, 2023, 05:32:30 pm ---Lot' to learn from here, once I picked it all apart.
...
So you probably imply that this audio signal is passed through a 22KHz bandpass filter before sampling - which essentially gets rid of the 10MHz carrier/noise anyway.

--- End quote ---

If you put a 20kHz audio signal on a 10MHz carrier, all the energy is between (roughly!) 9.98MHz and 10.02MHz. Putting that into a 10MHz+-20kHz bandpass filter changes nothing; the carrier and signal are still there. Put it into a 0-20kHz filter, and you will get nothing out :)


--- Quote ---That made me think about the 8Bit sampling resolution again. I still have a tiny DSO112A pocket scope. This too has "8bit resolution" but that does fit well with the equally tiny screen. You don't expect much from such a device. Having to fill a 10" hi-res display with the same number of samples you certainly have to do some visual tricks to NOT make the scope look like a toy.

--- End quote ---

Firstly beware of "raw" number-of-bits figures. ADCs are subtle beasts, and ENOB (effective number of bits) is a better measure of performance. Any decent scope specification will include that.

Even if you can't see all the bits on a screen, too few bits can still cause artifacts. That is particularly noticeable if you look at the frequency domain version of the captured waveform, by doing an FFT. Too few bits will be visible as worse harmonics and increased noise floor.

BTW, it is good to see someone reading and thinking - that's a necessary starting point :) Not everybody manages it :(

Once you have gone around the loops of conceptual understanding a few times, you will begin to see that samplers and mixers can perform similar functions - not only in terms of mathematic operations, but also in electronic circuits. Not everyone persists enough to reach that level of understanding and see the beauty.

Have fun.

Fungus:

--- Quote from: smallfreak on April 10, 2023, 05:32:30 pm ---I did read quite a lot here the last view days and I stumbled across quite some pictures showing a really BAD representation of "actual waveforms" near the upper bandwidth. Likely random zigzag due to idiotically drawing straight lines between the measured points. Unfortunately I was unable to find them again quickly.

--- End quote ---

Recent posts or old posts?

2N3055:

--- Quote from: tggzzz on April 10, 2023, 06:20:11 pm ---
--- Quote from: smallfreak on April 10, 2023, 05:32:30 pm ---Lot' to learn from here, once I picked it all apart.
...
So you probably imply that this audio signal is passed through a 22KHz bandpass filter before sampling - which essentially gets rid of the 10MHz carrier/noise anyway.

--- End quote ---

If you put a 20kHz audio signal on a 10MHz carrier, all the energy is between (roughly!) 9.98MHz and 10.02MHz. Putting that into a 10MHz+-20kHz bandpass filter changes nothing; the carrier and signal are still there. Put it into a 0-20kHz filter, and you will get nothing out :)


--- Quote ---That made me think about the 8Bit sampling resolution again. I still have a tiny DSO112A pocket scope. This too has "8bit resolution" but that does fit well with the equally tiny screen. You don't expect much from such a device. Having to fill a 10" hi-res display with the same number of samples you certainly have to do some visual tricks to NOT make the scope look like a toy.

--- End quote ---

Firstly beware of "raw" number-of-bits figures. ADCs are subtle beasts, and ENOB (effective number of bits) is a better measure of performance. Any decent scope specification will include that.

Even if you can't see all the bits on a screen, too few bits can still cause artifacts. That is particularly noticeable if you look at the frequency domain version of the captured waveform, by doing an FFT. Too few bits will be visible as worse harmonics and increased noise floor.

BTW, it is good to see someone reading and thinking - that's a necessary starting point :) Not everybody manages it :(

Once you have gone around the loops of conceptual understanding a few times, you will begin to see that samplers and mixers can perform similar functions - not only in terms of mathematic operations, but also in electronic circuits. Not everyone persists enough to reach that level of understanding and see the beauty.

Have fun.

--- End quote ---

All you said here is absolutely true. If you connect 10MHz carrier AM modulated with 20 kHz audio signal all your energy will be around carrier.  And you can see that nicely if you connect that signal to spectrum analyser, you will see that clearly... If you connect it to a CRT analog scope you will see  10 MHz signal varying in amplitude.....  And in order to see exactly the same image on digital scope you need to sample in such manner that your effective sampling rate should satisfy Nyquist  for 10Mhz and a bit more (to account for modulation), so roughly lets say a 25 MS/s or more would be OK number...

If audio signal was single 20Khz tone, you could use old repetitive sampling scope, that should have sampling aperture at faster than  40 ns (1/25M samples per second) and you could take single samples from thousands of separate trigger events by varying time where you take sample in time in regards to trigger point and reconstruct on screen same thing you would see on realtime digital scope sampling at real 25 MS/s because your effective sampling rate would also be 25 MS/s. Nyquist happy.

On the other hand if you where looking at the signal that had any variation in modulation signal ( a voice or music instead of single 20 kHz tone) than you need to use realtime sampling scope that has sufficient BW to capture full signal which is 10 MHz or analog scope.... Repetitive sampling scope will show crap..

There is no place whatsoever mentioning RF samplers and mixers in a discussions about realtime digital scopes. It is a out of topic, by a mile.. We might as well speak about gardening when someone asks a question about scopes....


And by this I don't mean RF samplers and mixers are not fascinating subject and that you obviously know quite a lot about it. Quite a lot. It's just it is off the topic here.

smallfreak:

--- Quote from: Fungus on April 10, 2023, 07:10:22 pm ---Recent posts or old posts?

--- End quote ---

If I find them again, I will take note of that. It could well be it was off site. I've seen a couple of YT videos on the scopes in my shortlist. Not everyone did make the impression to be a real scope specialist. But even if they just did not know how to configure it right, I still was amazed, that any reputable manufacturer would even allow for such a crappy display. I don't want to blame a particular model. But it was more than once, so I maybe wrongly assumed that it is more common.

It also might have been enhanced by a later firmware update however. Unfortunately many tests, videos or just "unboxing of ..." are not always kept up to date with "problems already fixed". It's even quite difficult to find such information here. It might be hidden in "4000+ folloups".  ???

Perhaps a different situation, but maybe I might refer to this pretty recent post of an obviously knowledgeable person. The screenshot probably shows "raw data samples connected by lines" and not what I would recognize as a "true signal reconstructed from data points within bandwidth". Although it is a single shot and no average. At 20ns/div I would expect overshoots on the edges and a much smoother top, even with 250MHz bandwidth and 1GS/s. The measurements certainly do have signal noise and at least 1Bit digitizing noise but that shouldn't necessarily get translated into a jaggy line showing edges and corners way beyond the bandwidth. There is no thing like "corners" in signals.

While these data points might have been measured just so, they most likely have to get correlated first to get a better representation of the underlying true signal they had sampled - which than may take advantage of a higher vertical resolution than the 256 different discrete values an 8Bit digitizer might collect. Maybe the scopes CAN do that, but nobody uses such a function for certain reasons. I know that it can be done on modern scopes for multiple waveforms in a greyscale or color manner to smooth out statistical variations.

I have not yet found a suitable discussion that/why this might be the only viable method of displaying the samples, but I still have a few thousands of Posts to read.  ::) But I do believe I have not seen a "reconstructed signal" so far on any DSO topic.

Being mostly analog with my measurements so far, I might have to get used to different expectations in the DSO world.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod