| Electronics > Beginners |
| Oscillo-confusion MHz GSa/s wfm/s Mpts |
| << < (11/13) > >> |
| tggzzz:
--- Quote from: james_s on January 02, 2019, 08:22:30 pm --- --- Quote from: rstofer on January 02, 2019, 08:19:44 pm --- It would seem to me that you need a lot more than 6 MHz of bandwidth to display a 6 MHz signal. At a minimum I would want to see the 5th harmonic (30 MHz) and preferably more (like 7th, 9th, 11th, 13th and 15th) with a 100 MHz scope. If you were to view a 6 MHz square wave on a 6 MHz scope (or 10 MHz scope) all you would see is a sine wave at the fundamental frequency. --- End quote --- But does that really matter? It depends on what you're trying to see here, whether it is the data/timing or signal integrity which is going to need more bandwidth. I actually do not recall ever using a scope to look at USB, signal integrity is rarely the issue there, normally USB problems are down to drivers and software issues. --- End quote --- Yes, it does matter, all too often. What you are trying to see is of key importance with any tool. Often it indicates that a different type of tool is possible or preferable. Of course you don't look at a USB line with a scope to verify signal integrity. If you haven't designed the hardware, if there are problems then it shouldn't have escaped the factory. Alternatively if you are the hardware designer, then you ought to know what the data sheets and app notes indicate. If you don't then signal integrity is the least of your problems! |
| james_s:
I just don't fully agree. I got by for many years with a 100MHz scope, before that I was using a 20MHz scope. It worked just fine for the stuff I was doing, microcontrollers and analog circuits. It simply did not matter that higher frequency square waves had the corners rounded off, I knew that was a limitation of the scope and it didn't matter, it was an irrelevant detail. I needed to see the data, and I could read rounded square waves just as easily as I could read nice square ones. It's just funny to see people going on like a hobbyist absolutely NEEDS things that no hobbyist could ever hope to afford until very recently. High bandwidth is a nice luxury, but outside of certain specialized use cases you don't NEED high bandwidth unless it's just to satisfy some kind of OCD. You bring up edge cases like looking at USB, why would someone even be looking at USB with a scope? That's using a hammer to drive a screw, it may work, sort of, but it's not the right tool for the job. When it comes to digital circuits, signal integrity is exactly what a scope is ideal for, otherwise a logic analyzer is often a more appropriate tool. The logic analyzer doesn't show you the true shape of the waveform either but that doesn't matter because unless you're looking at signal integrity the precise shape of the digital waveform is irrelevant so long as it is read as 1s and 0s and there is not something waaaaaaaaaaay out of whack. |
| tggzzz:
--- Quote from: james_s on January 02, 2019, 10:03:36 pm ---I just don't fully agree. I got by for many years with a 100MHz scope, before that I was using a 20MHz scope. It worked just fine for the stuff I was doing, microcontrollers and analog circuits. It simply did not matter that higher frequency square waves had the corners rounded off, I knew that was a limitation of the scope and it didn't matter, it was an irrelevant detail. I needed to see the data, and I could read rounded square waves just as easily as I could read nice square ones. It's just funny to see people going on like a hobbyist absolutely NEEDS things that no hobbyist could ever hope to afford until very recently. High bandwidth is a nice luxury, but outside of certain specialized use cases you don't NEED high bandwidth unless it's just to satisfy some kind of OCD. You bring up edge cases like looking at USB, why would someone even be looking at USB with a scope? That's using a hammer to drive a screw, it may work, sort of, but it's not the right tool for the job. When it comes to digital circuits, signal integrity is exactly what a scope is ideal for, otherwise a logic analyzer is often a more appropriate tool. The logic analyzer doesn't show you the true shape of the waveform either but that doesn't matter because unless you're looking at signal integrity the precise shape of the digital waveform is irrelevant so long as it is read as 1s and 0s and there is not something waaaaaaaaaaay out of whack. --- End quote --- I didn't bring up USB; I was responding to it in your post - which you have conveniently snipped. Your contentions are mutually contradictory and/or claiming the same points I have repeatedly advocated - without acknowledging that. For example, you claim a scope is ideal for signal integrity, but then claim that 20MHz scopes are adequate - which they demonstrably are not. |
| FriedMule:
Right now I am thinking that better data always give higher resolution and accuracy while more channels only are useful the times I do need to measure 3+ places, is that stupid thinking or good thinking? |
| ebastler:
--- Quote from: FriedMule on January 03, 2019, 07:59:03 pm ---Right now I am thinking that better data always give higher resolution and accuracy while more channels only are useful the times I do need to measure 3+ places, is that stupid thinking or good thinking? --- End quote --- Yes, that's right. Requiring more than two channels is relatively rare. But on the other hand: In most cases, having a lower bandwidth will only give you gradually reduced information. (The bandwidth rolls off at 3dB/octave or such, so you can still see signal above, although at limited resolution and maybe noticeably distorted.) In contrast, not having the two extra channels when you need them will often mean a "hard stop"; you just can't observe the required connection between the signals. Sorry, that argument doesn't make your decision any easier... I had it easier when the DS1054Z was a breakthrough in bandwidth and number of channels a few years ago. :P |
| Navigation |
| Message Index |
| Next page |
| Previous page |